Potter Thesis 2009

THE PROPAGATION, CHARACTERISATION AND
OPTIMISATION OF CANNABIS SATIVA L
AS A PHYTOPHARMACEUTICAL
A thesis submitted by
David Potter JP
MIBiol CBiol FLS CMIOSH
In fulfilment of the requirement

for the degree of Doctor of Philosophy (PhD)
in Pharmaceutical Sciences
Department of Pharmaceutical Science Research
King’s College London
March 2009
Abstract
In response to known pharmacology, and an increasing weight of anecdotal evidence of
efficacy, clinical trials have been performed to support the licensing of cannabis-based
botanical medicines. The initial applications envisaged were the treatment of cancer
pain, neuropathic pain and various symptoms associated with multiple sclerosis. With
effective alternatives often unavailable, otherwise law-abiding UK patients have regularly
turned to illicit cannabis for medical relief. The main active ingredients in this are the
cannabinoids THC and CBD, but other pharmacologically active cannabinoids are also
present. One study reported here quantifies these cannabinoids and assesses the likely
implications for efficacy. Using light microscopy, studies are performed to expand current
knowledge of the form and function of trichomes in Cannabis sativa L. Supporting
chemical analyses ascertain what secondary metabolites are biosynthesised within
these trichomes, and determines where and when this occurs. To comply with the
demands of the pharmaceutical industry, and in marked contrast to illicit cannabis, a
phytopharmaceutical feedstock must meet high expectations regarding the minimum
and maximum content of a range of compounds. Specific studies are performed to
ascertain how growing methods affect the secondary metabolite content. They also aim
to find out how a tight specification can be met while satisfying commercial and
environmental expectations. This involves studying plant development and secondary
metabolite biosynthesis in both indoor and outdoor conditions. The first approved
cannabis-based botanical medicine supported by this research is Sativex®. This
became available in Canada in 2005 for the treatment of central neuropathic pain in
multiple sclerosis and in 2007 for intractable cancer pain. The medicine is also available
in the UK and many other countries on a ‘named patient basis’. This thesis has also
supported the production of a range of other cannabinoids which are undergoing in-vitro
and in-vivo testing. This could lead to the commercial production of an increasing range
of phytopharmaceuticals.
i
Cannabis sativa L cv Gayle
CBD Chemotype.
Awarded European Plant Variety Rights EU 16301, 10th Oct 2005
To my best mate John Pook

and those who cared for him at
St Leonard’s Hospice, York
ii
Acknowledgements
This PhD would not have been possible without the support of friends and colleagues, of
which there are too many to comprehensively mention. My deep appreciation first goes
to my dear friend and champion Dr Brian Whittle who, having convinced me to do the
PhD, guided me on my ‘long road to Ithaka’. Enormous thanks goes too to my supervisor
Professor Marc Brown for his faith in me and his support and patience throughout.
I express my sincerest gratitude to GW Pharmaceuticals Ltd, particularly Dr Geoffrey
Guy, for sponsoring this PhD. Special thanks goes to a fabulous team of botanical
colleagues – Dr Etienne de Meijer, Dr Tim Wilkinson, Barry and Chris Mears, Keith Hines
and Roger Phillips for supplying cannabis and moral support throughout the project. I am
especially grateful to Kathy Hammond for diligently analysing so many of my samples
and Heather North for administrative assistance and pastoral care.
Down on the farm I am indebted to great friends John and Vicky Clinch, over whose
kitchen table much of the PhD was planned. In addition to growing many of my crops,
they assisted in overnight monitoring as plants were dried. In the interest of science
they also tolerated the incredible mess as I made CBD-rich hashish in their kitchen.
Aren’t our police wonderful? Much of the work would not have been possible without help
from several constabularies, and I would like to thank officers Phil Painter, Jerry Prodger,
Howard Chandler, Bill Stupples and Steve Holme for assisting me with my enquiries.
Heartfelt thanks goes to an ex-colleague and fond friend Heather House, who was my
co-speaker at several inspirational Multiple Sclerosis Group Meetings. My thanks go to
Dr Peter Toomey for his services to St Leonards Hospice in York, and for the invitation to
give a presentation at a memorable Yorkshire Pain Management Group Meeting at
St Stephens College – an experience that opened my eyes to the dedicated work of
those involved in pain control. Warm thanks go too to good friends Peter Smith for
photographic assistance and Valerie Bolas for botanical illustrations, and to colleagues
at Kings College London - especially Darragh Murnane and Yanjun Xhao. Finally, I thank
my wife Jane for supporting my PhD application – and for her forebearance thereafter.
iii
List of Publications
Potter, D.J., 2004. Growth and Morphology of Medicinal Cannabis. In Guy, G.W.,
Whittle, B.A. and Robson, P.J. (Eds.). The Medicinal Uses of Cannabis and
Cannabinoids. Pharmaceutical Press, London. pp. 17-54.
Potter, D.J., Clark, P. and Brown M.B. 2008. Potency of Δ9-THC and other cannabinoids
in cannabis in England in 2005: Implications for psychoactivity and pharmacology.
J Forensic Sci 53 (1): 90-94.
Russo, E.B., Jiang, H-E., Li, X., Sutton, A., Carboni, A., del Bianco, F., Mandolino, G.,
Potter, D.J., Zhao, Y.X., Bera, S., Zhang, Y-B., Lü, E-G., Ferguson, D.K., Hueber, F.,
Zhei, L-C., Liu, C-J., Wang Y-F., Li C-S. 2008. Phytochemical and genetic analyses of
ancient cannabis from Central Asia. J. Exp. Bot. 59, 15, 4171-4182.
iv
Table of Contents
Abstract.....................................................................................................................i
Acknowlegements................................................................................................... iii
List of Publications .................................................................................................. iv
Table of Contents.....................................................................................................v
List of Figures ....................................................................................................... xiii
List of Tables........................................................................................................ xxii
Abbreviations ......................................................................................................xxvii
Chapter 1 INTRODUCTION.....................................................................................1
1.1 Plants as a source of medicines – past and presen .......................................1
1.2 Cannabis Botany ............................................................................................4
1.3 Cannabis Taxonomy ......................................................................................6
1.4 UK Medicinal Cannabis Use - History and Legal Complications. ................. 8
1.5 The influence of the BMA and the House on Lords Select Committee on
Science and Technology on UK Cannabis Research..................................10
1.6 International Legal Attitudes to Medicinal Cannabis.....................................11
1.6.1USA ........................................................................................................12
1.6.2 Canada ..................................................................................................14
1.6.3 Mainland Europe....................................................................................14
1.6.4 Ireland....................................................................................................15
1.6.5 Australia.................................................................................................15
1.6.6 Japan .....................................................................................................15
1.7 The choice of active pharmaceutical ingredients (APIs) ...............................15
1.8 Cannabinoid and terpene biosynthesis ........................................................17
1.9 Cannabinoid Receptors and Cannabinoid Pharmacology. ...........................21
1.10 Outline of Thesis ........................................................................................24
v
CHAPTER 2 CHARACTERISATION OF ILLICIT CANNABIS IN THE UK.............27
2.1 INTRODUCTION..........................................................................................27
2.2 AIM AND OBJECTIVES ...............................................................................29
2.3 MATERIALS .................................................................................................29
2.3.1 Cannabis samples .................................................................................29
2.3.2 Microscopy, Photography and other Apparatus .....................................29
2.4 METHODS ...................................................................................................30
2.4.1 Collection of Representative Samples ...................................................30
2.4.2 Storage of illicit cannabis samples .........................................................30
2.4.3 Categorisation of the form of each sample ............................................30
2.4.3 Categorisation of the form of each sample ............................................30
2.4.3.1 Cannabis resin ................................................................................30
2.4.3.2 Herbal cannabis ..............................................................................31
2.4.3.3 Sinsemilla ........................................................................................31
2.4.3.4 Cannabis powder ............................................................................32
2.4.3.5 Other categories not included .........................................................33
2.4.4 Measurement of cannabinoid potency and profile .................................33
2.4.5 Statistical Analysis .................................................................................33
2.5 RESULTS AND DISCUSSION .....................................................................34
2.5.1 Categorisation of cannabis type between regions .................................34
2.5.2 The range of cannabinoids in each cannabis category ..........................34
2.5.3 Comparison of cannabis potency and profile between regions..............38
2.5.4 Trends in Cannabis Potency..................................................................39
2.5.5 The efficacy of illicit cannabis ................................................................42
2.6 CONCLUSIONS ...........................................................................................45
Chapter 3 Cannabis trichome form, function, and distribution ...............................47

3.1 INTRODUCTION..........................................................................................47
vi
3.3 MATERIALS .................................................................................................50
3.3.1 Germplasm ............................................................................................50
3.3.2 Microscopy, Tissue Stains, Photography and other Apparatus..............51
3.4 METHODS ...................................................................................................51
3.4.1 Photomicrograph Studies.......................................................................51
3.4.1.1 Choice of Microscopes ....................................................................51
3.4.1.2 Staining ...........................................................................................52
3.4.1.3 Unmounted Sample Preparation .....................................................52
3.4.1.4 Mounted sample preparation...........................................................52
3.4.1.5 Illumination ......................................................................................53
3.4.1.6 Photography ....................................................................................53
3.4.1.7 Isolation and Observation of Detached Glandular Resin Heads .....54
3.4.2 Effect of glandular trichome array on the secondary metabolite content
of plant tissues.......................................................................................54
3.4.3 Organoleptic Assessment of the Effect of Trichome Colour and
Pubescence Density on cannabis potency. ...........................................55
3.4.4 Effect of photosynthetic ability, or lack of ability, on cannabinoid
biosynthesis in sessile trichomes. ..........................................................56
3.4.5 Statistical Methods.................................................................................57
3.5.1 Photomicrograph studies .......................................................................57
3.5.1.1 Simple unicellular trichomes............................................................57
3.5.1.2 Cystolythic trichomes ......................................................................58
3.5.1.3 Capitate sessile trichomes .............................................................59
3.5.1.4 Antherial Sessile Trichomes ............................................................60
3.5.1.5 Capitate Stalked Trichomes ............................................................61
3.5.1.6 Bulbous Trichomes..........................................................................73
3.5.1.7 Effect of age and storage on glandular trichome colour ..................73
3.5.2 Effect of glandular trichome array on the secondary metabolite content
of plant tissues .....................................................................................74
vii
3.5.3 Effect of capitate stalked trichome density and colour on cannabinoid
content and profile. ................................................................................76
biosynthesis in sessile trichomes on variegated leaf tissue. ..................80
3.6 CONCLUSIONS ...........................................................................................82
Chapter 4. The Function and Exploitation of Secondary Metabolites from

Glandular Trichomes of Cannabis sativa L. .........................................84
4.1 INTRODUCTION..........................................................................................84
4.3 MATERIALS .................................................................................................87
4.3.1 Germplasm ............................................................................................87
4.3.2 Apparatus ..............................................................................................87
4.4 METHODS ...................................................................................................87
4.4.1 Separation of Sessile and Capitate Stalked Trichomes Glandular
Heads from mature fresh floral material.................................................87
4.4.2 Bulk-Production of Pure Sessile Trichome Preparations. ......................88
4.4.3 Production of a cannabichromene-rich sessile trichome preparation.....88
4.4.4 Ontogenetic changes in Secondary Metabolite Content of Glandular
Trichome Contents.................................................................................88
Heads from mature fresh floral material..................................................90
4.5.2 Isolation of intact sessile glandular trichomes from vegetative material.91
4.5.3 The collection of sessile trichomes from foliage of a high CBC
chemotype as a means of isolating the minor cannabinoid CBC ...........94
4.5.4 Ontogenetic changes in glandular trichome secondary metabolite
content ...................................................................................................94
4.6. CONCLUSIONS ........................................................................................100
viii
Chapter 5 Indoor Propagation of Medicinal Cannabis..........................................102
5.1 INTRODUCTION........................................................................................102
5.2 AIM AND OBJECTIVES .............................................................................106
5.3 MATERIALS ...............................................................................................107
5.3.1 Plant Propagation and Drying Materials...............................................107
5.3.2 Germplasm Details ..............................................................................108
5.3.3 Light Measurement and Weighing Equipment .....................................108
5.3.4 Growth Medium ...................................................................................108
5.4 METHODS .................................................................................................109
5.4.1 Routine Propagation and Plant Production Methods ...........................109
5.4.1.1 Seed sowing and transplantation of seedlings ..............................109
5.4.1.2 Production of Cuttings (Clones).....................................................109
5.4.1.3 Nurturing Vegetative Growth of Seedlings and Cuttings. ..............110
5.4.1.4 Induction and Maintenance of Flowering.......................................110
5.4.1.5 Biological Pest Control ..................................................................110
5.4.1.6 Harvesting .....................................................................................110
5.4.1.7 Crop Drying ...................................................................................111
5.4.1.8 Stripping ........................................................................................112
5.4.1.9 Garbling.........................................................................................112
5.4.1.10 Storage........................................................................................112
5.4.1.11 Environmental Control System ....................................................113
5.4.2 Specific Methods .................................................................................113
5.4.2.1 Uniformity of Plants Grown from Cuttings or Seed ........................113
5.4.2.2 Effect of Duration of Flowering Period on Yield ..........................114
5.4.2.3 Effect of Daylength on Cannabinoid Profile (Part 1) Comparison of
12 and 13 hour daylength ..................................................................114
5.4.2.4 Effect of Daylength on Cannabinoid Profile (Part 2) Comparison of
11 and 12 hour daylength...............................................................115
5.4.2.5 Plant height assessment ...............................................................115
5.4.2.6 Stigma senescence assessment ...................................................116
ix
5.4.2.7 Plant Weight Assessment .............................................................116
5.4.2.8 Cannabinoid Content and Profile...................................................116
5.4.2.9 Effect of Irradiance Level on Plant and Cannabinoid Yield............117
5.4.2.10 Effect of the length of flowering period on the cannabinoid profile
of heterozygous plants of the mixed THC/CBD chemotype. ..............118
5.4.2.11 Statistical Analysis.......................................................................118
5.5 RESULTS AND DISCUSSION ...................................................................119
5.5.1 Comparison of the Yield and Uniformity of Plants Grown from
Cuttings or Seeds...............................................................................119
5.5.2 Effect of Irradiance Level on Plant and Cannabinoid Yield ..................120
5.5.3 Effect of Duration of Flowering Period on BRM and Cannabinoid
Yield ...................................................................................................126
5.5.3.1 Effect of Duration of Harvest Period on Ratio of THC and CBG in
THC Chemovars ................................................................................127
5.5.3.2 Effect of the length of flowering period on profile of heterozygous
chemotypes with mixed THC/CBD profiles ........................................129
5.5.4 Effect of Daylength on Plant Development and Cannabinoid Profile ...130
5.5.4.1 Comparison of Twelve and Thirteen Hour Daylength Regimes.....130
5.5.4.2 Comparison of Eleven and Twelve Hour Daylength Regimes.......136
5.5.4.3 Review of the comparisons of plants induced to flower on
daylengths 11, 12 and 13 hours. ........................................................140
5.6 CONCLUSIONS .........................................................................................141
Chapter 6 The Outdoor Propagation of Phytopharmaceutical Cannabis .............143

6.1 INTRODUCTION........................................................................................143
6.2 AIM and OBJECTIVES...............................................................................144
6.2.1 The effect of growing environment on female plant development........145
6.2.2 Comparison of the secondary metabolite yield and profile of fresh
plant material and enriched trichome preparations made from them ...145
x
6.2.3 The effects of harvest timing on secondary metabolite yield and profile
.............................................................................................................145
6.2.4 Comparison of the secondary metabolite profiles of glasshouse and
outdoor grown plants ...........................................................................145
6.2.5 Evaluation of outdoor pest and disease issues....................................145
6.2.6 Evaluation of Crop Drying Methods .....................................................145
6.3 MATERIALS ...............................................................................................146
6.4 GENERAL AGRONOMIC METHODS ........................................................146
6.4.1 Seedbed Location, Preparation and Crop Establishment ....................146
6.4.2 Field Trial Design .................................................................................147
6.4.3 Soil Nutrition ........................................................................................147
6.4.4 Pest and Disease Monitoring and Management ..................................147
6.4.5 Harvest ................................................................................................147
6.4.6 Crop Drying and Stripping....................................................................147
6.4.7 Assessment of Crop Development ......................................................148
6.5 Secondary Metabolite Purification and Analytical Methods ........................148
6.5.1 Production and Collection of Enriched Trichome Preparations............148
6.5.2 Steam distillation of trichome rich preparations ...................................148
6.5.3 Steam distillation fresh foliar and floral material...................................149
6.6 Statistical methods .....................................................................................149
6.7 RESULTS and DISCUSSION.....................................................................149
6.7.1 Observations on Crop Establishment and Plant Development ............149
plant materials and enriched trichome preparations made from them .154
6.7.3 The effects of harvest timing and growth stage on yield and
cannabinoid profile .............................................................................156
6.7.3.1 Botanical Raw Material Yield.........................................................156
6.7.3.2 Potency of CBD chemovars ..........................................................157
6.7.3.3 CBD Yield......................................................................................158
xi
6.7.3.4. Effect of Harvest Date and Growth Stage at Harvest on
Cannabinoid Profile.....................................................................159
6.7.4 Effect of Growth Stage and Harvest Date on Essential Oil Profile .......161
6.7.5 Comparison of the secondary metabolite content of glasshouse and
outdoor grown plants..........................................................................165
6.7.6 Summary of Pest and Disease Problems in the Field Trials ................167
6.7.7 Effect of Raised Temperatures on Crop Drying Time ..........................171
6.8 CONCLUSIONS .........................................................................................172
Chapter 7 GENERAL DISCUSSION....................................................................174
REFERENCES ...................................................................................................185
APPENDICES ..................................................................................................215
xii
List of Figures
Figure 1.1. Contrasting leaf morphology in three clones of Cannabis sativa L., (a) CBD
chemotype G5 M16 cv Gill, (b) THC chemotype G1 M3 cv Guinevere (c) Afghan
landrace clone M146 (Illustrations by Valerie Bolas, commissioned by GW
Pharmaceuticals. 4
Figure 1.2. (a) Male (left) and female cannabis (right) in later stage of flowering. (b)
Female cannabis inflorescence. (c) A cluster of male flowers with sepals split open and
reflexed to expose the anthers. 5
Figure 1.3. Biosynthetic pathway of THC and THCV, via CBG or CBGV. 1,
Geranylpyrophosphate; 2, Divarinic Acid (R1) or Olivetolic Acid (R2); 3,
Cannabigerovarin (CBGV) (R1) or Cannabigerol (CBG) (R2);. 4, Δ9-
tetrahydrocannabivarin (R1) or Δ9-tetrahydrocannabinol (R2). R1 (-C3H7) and
R2(-C5H11) indicate the propyl or pentyl forms of the metabolites; Enzyme I:
geranylpyrophosphate:olivetolate geranyltransferase(GOT); Enzyme II: THC(V)
synthase. 18
Figure 1.4. The two pathways of isopentyl diphosphate (IPP) biosynthesis in plants, as
found in the plastid and cytosol respectively. 20
Figure 1.5. The disassembly of an activated G-protein into two signalling components.
(Alberts et al., 2002) 22
Figure 2.1. Examples of cannabis resin samples (<1g up to 230g) seized by police in
2004/5. 30
Figure 2.2. (a) Loose herbal cannabis material showing separated seeds; (b)
Compressed herbal cannabis material with selection of removed seeds. 31
Figure 2.3. A typical confiscated sample of illicit sinsemilla cannabis consisting of three
separate packets, each containing approximately one gram. 32
xiii
Figure 2.4. (a) A herb grinder in closed position; (b) An open herb grinder revealing the
component parts. 32
Figure 2.5. (a) The balance of THC, CBD and CBN in sinsemilla (n = 256); (b) The
balance of THC, CBD and CBN in herbal cannabis (n = 35); (c) The balance of THC,
CBD and CBN resin (n = 169). 37
Figure 2.6 The correlation between THC and CBN content in resin samples seized in five
constabularies in 2004/5 (n = 169). 38
Figure 2.7. A comparison of the range and distribution patterns of THC content of seized
imported herbal cannabis samples in 1998, King et al. (2004) (n = 44) and 2005 Potter et
al. (2008) (n = 33). 39
Figure 2.8. A comparison of the ranges of THC contents of Sinsemilla seized in the UK
and analysed by the Forensic Science Service in 1996-8 (n = 145) and samples seized
by police in Derbyshire (n=15), Kent (n=58), London Metropolitan (n = 96), Merseyside (n
= 44) and Sussex (n = 34) in 2004/5 (total n = 247). 40
Figure 3.1. Upper surface of a bract within a cannabis inflorescence showing glandular
stalked trichomes to be present only within the proximal region. 55
Figure 3.9. A variegated leaf of clone M60, with 1cm diameter disks cut from
symmetrically opposite sides of the midrib. 57
Figure 3.3. (a) Unicellular non-glandular trichome. The sample is temporarily mounted
under hemp oil and viewed in transmitted light; (b) Cystolythic trichomes observed on the
leaf margin of a young leaf. The sample was temporarily dry-mounted and viewed in
transmitted light. Cystolyths (concretions of calcium carbonate) are visible at the base of
each trichome. 58
Figure 3.4. (a) A capitate sessile trichome observed on the edge of one of the first pair of
true leaves of a cannabis seedling. The specimen was temporarily dry-mounted and
viewed using both transmitted and incident light; (b) a sessile trichome on a leaf surface
xiv
stained with Fast Blue. The still-wet sample was temporarily dry-mounted and viewed
using incident light. 59
Figure 3.5. (a) a row of antherial sessile trichomes showing their normal distribution in
the furrow of a cannabis anther. These anthers were captured in incident light through a
low-power microscope. (b) closer view of antherial trichomes. 61
Figure 3.6. A capitate stalked trichome (centre) between two cystolythic trichomes. The
specimen is temporarily dry-mounted and illuminated from below. The secretory cells
are out-of-focus due to the optical distortion within the glandular head. 62
Figure 3.7. Two dry-mounted capitate stalked trichomes viewed in transmitted light.
Most of the features are out-of-focus. In the right-hand trichome, a crisp view of cells
within the secretory cell disk appears as an in-focus image. However these appear to
be located outside of the trichome structure, due to the refractive properties of the resin
head. 63
Figure 3.8. (a) A temporarily dry-mounted capitate stalked trichome viewed in
transmitted light. An irregular arrangement of poorly-defined secretory cells is visible at
the base of the glandular head; (b) A capitate stalked trichome, temporarily mounted in
glycerol and viewed in transmitted light, and (c) an illustration of a capitate stalked
trichome on Cannabis sativa by Briosi and Tognini (1894). 64
Figure 3.9. (a and b). Similar sized capitate stalked trichomes temporarily mounted
under hemp oil. The samples are viewed in transmitted light. Possibly because of a
similarity in the refractive index of the oil and the secretory cell contents, these cells
appear clear. The outer membrane at the base of the glandular head appears dark and
opaque. 65
Figure 3.10. A contrasting pair of resin heads on capitate stalked glandular trichomes,
naturally orientated to allow sideways-on (left) and overhead views (right). The
specimen is temporarily mounted in a 70% v/v aqueous solution of glycerol and
illuminated from below. 66
xv
Figure 3.11. (a) Secretory cells stained red, within the glandular head, after thirty minutes
in 1% tetrazolium red; (b) A capitate stalked trichome with glandular head removed by
slight abrasion. After thirty minutes in 1% tetrazolium the disk of secretory cells is
stained bright red; (c) A mature capitate stalked trichome between two non-glandular
cystolythic trichomes, viewed after twelve hours in 0.1% tetrazolium solution. 67
Figure 3.12. (a) A capitate stalked trichome temporarily dry-mounted and viewed in
transmitted and incident light. The glandular head has become partly detached from the
stalk to expose the stipe cells, which connect the disk of secretory cells to the
hypodermal cells within the stalk; (b) The stalk of a capitate stalked trichome after
detachment of the resin head. The specimen was dry-mounted and illuminated with
incident light. The stipe cells can just be seen protruding from the top of the stalk. 68
Figure 3.13. Separation of the glandular head during (left) and after (right) the
appearance of a fissure above the secretory cells. The example shown was observed
on (M280). The specimen was viewed in incident light. 68
Figure 3.14. (a) An intact glandular stalked trichome of a naturally pigmented clone
M186, with coloured cells visible within the resin head. The sample was temporarily
mounted in hemp oil; (b) A direct overhead-view of the stalk of a capitate stalked
trichome on clone M186 after resin head detachment. The resin head has become
detached leaving thr base of the secretory head attached to the stalk. The sample was
temporarily mounted in oil. 70
Figure 3.15. A detached resin head (approximately 100µm diameter) from a capitate
stalked trichome, viewed from below to gain a clear view of the scar where the stipe cells
were originally attached. The head had been trapped on the surface of clear adhesive
tape. 71
Figure 3.16. (a) A dense pubescence of glandular stalked trichomes on a bract within a
cannabis female inflorescence. The specimen was illuminated from behind and
photographed with a tripod-mounted camera incorporating a macro lens. The
orange/brown structures are senesced stigmas; (b) two young cotton-melon aphids
xvi
Aphis gossypii. All six legs on each specimen are irreversibly adhered to the resin heads
of capitate stalked trichomes. 72
Figure 3.17. (a) A small bulbous trichome (left) alongside a fully developed glandular
stalked trichome. The contrast in resin head diameter (10 µm v 100 µm) is clear; (b) a
simple bulbous trichome and (c) a complex bulbous trichome. These are 10-15 µm in
diameter. These samples were temporarily dry-mounted and viewed in mixed
transmitted and incident light. 73
Figure 3.18. (a) Clear glandular stalked trichomes on freshly harvested young cannabis
floral tissue; (b) Brown trichomes on three-year old stored cannabis. 74
Figure 3.19. The mean density of capitate stalked and sessile trichomes (± 1SE) (upper
and lower surfaces combined), on each of three high-THC cultivars M3, M6 and M7. On
each clone twenty randomly selected fields were counted on both the upper and lower
surface in the proximal and distal areas. 74
Figure 3.20. The proportion of CBC expressed as % of THC+CBC in the cannabinoid
profile of proximal and distal tissue of bracts from each of three high-THC clones. (Error
bars on clones represent ± 1SD, and on the mean ± SE). In all three the difference was
highly significant (ANOVA, *** denotes p < 0.001). 75
Figure 3.21. The correlation between capitate stalked trichome density (visually
assessed 1-9 scale) and the overall THC content of the sample. Values shown are the
mean % w/w THC content (± SE only where n > 5) of all samples for each density score.
Regression line is shown in red. The regression model is:- % THC = 18.051 – 1.629 *
Density Score. (p < 0.001, R2 = 0.17). 76
Figure 3.22. The correlation between capitate stalked trichome colour (visually assessed
using a 1-9 scale) and the overall THC content of the sample. Values shown are the
mean % w/w THC content (±SE) of all samples for each colour score. Regression line is
shown in red. The model for this is: - Percent THC = 15.990 – 0.587 * Trichome Colour
Score (p < 0.001, R2 = 0.053). 77
xvii
Figure 3.23. The mean CBN content (± SD) of populations of sinsemilla samples
awarded each of the 1-9 ratings for trichome colour. All samples were seized by police in
2004/2005. 78
Figure 3.24. The variability in degree of THC catabolism to CBN as related to their
colour. Of 249 original samples 6 were rejected as outliers and not included here.
Trichome colour was assessed visually and scored on a 1-9 scale where 1 represents
totally clear and successively higher scores denote an increasing opacity and darkening
in colour. 79
Figure 3.25. Aged brown sessile glandular trichomes on 2700 year old cannabis. The
sample is illuminated with incident light and photographed digitally. The pubescence of
unicellular non-glandular trichomes is also clearly visible. 80
Figure 4.1. A smeared sample of sessile glandular trichome resin heads prepared from
vegetative foliage of high-THC clone G1 M3. The freshly captures specimens were
collected from the surface of a 25 µm sieve and are free-floating in water. A few pieces of
leaf fragment are also present as a minor contaminant. 92
Figure 5.1. Production of cannabis cuttings. A vegetative cannabis branch (a) is cut into
sections (b). Each cutting has been cut leaving approximately five centimetre of stem
below a single axial bud and up to one centimetre above (c). The base of the cutting is
dipped in rooting powder (d) and then placed in moist peat plugs (e). After two weeks
roots are protruding from the peat plugs and the cutting is ready to be planted. 111
Figure 5.2. A newly harvested crop hung to dry on wires. 112
Figure 5.3. An experiment to compare plant development and cannabinoid content when
flowered in 11 and 12 hour daylengths. Plants are maintained on ebb-and-flood
benches and lighting provided by high pressure sodium lamps. Duplicate batches of
plants are maintained either side of the curtain with plants on the right receiving the
longer daylength regime. 115
xviii
Figure 5.4. A close-up view of part of an unpolllinated cannabis inflorescence, showing
viable and older non-viable stigmas. 116
Figure 5.5. Average monthly BRM yield (two to four crops per month) (± SD) of THC and
CBD chemovars during the first full year of propagation. (No THC chemovar was
harvested in April and no CBD in February and November.) 120
Figure 5.6. The seasonal variation in cannabinoid yield of THC and CBD chemovars
during the first full year of propagation. Values shown were estimated by combining the
average monthly Botanical Raw Material yield (two to four crops per month) and the
average THC or CBD content (w/w). 121
Figure 5.7. The average yield of the THC chemovar before and after the replacement of
mercury vapour lamps (17 W m-2) with high pressure sodium lamps (55 W m-2) of
improved supplementary lighting (± SD). The mean is typically for four crops per month.
No crop was harvested in April of the first year. 122
Figure 5.8. Pattern of irradiance level in the glasshouse between 7 am and 7 pm (prior
to the improvements in supplementary lighting) and the pattern of average monthly
yields of THC chemovar raw material ± SD (n = 4). 124
Figure 5.9. The average monthly yield as a function of the glasshouse light level at the
beginning of flowering. On both axes the data was expressed as a percentage of the
maximum observed (r2 = 0.92, p < 0.001). 125
Figure 5.10. The yield of THC achieved by each of the clones (n=5) after six, eight and
ten weeks in flower. For clarity, the clone lines have been sorted in order of descending
THC yields after ten weeks in flower. 127
Figure 5.11. A comparison of the mean relative proportions (±SD) of THC and CBG in
twenty five clones at three harvest dates. Analyses of variance (one-way) compared the
proportion of THC in each clone to that in the Sativex-dependent clone G1 (shown in
red). (* p < 0.05, ** p < 0.01, *** p < 0.001). 128
xix
Figure 5.12. Effect of Daylength on Plant Height ± SD (n = 20) ten weeks after induction
of flowering (* p < 0.05, *** p < 0.001, ANOVA for individual clones and paired t-test for
the overall mean). 133
Figure 5.13. Effect of Daylength on Yield of Botanical Raw Material ± SD (n = 5 plants)
ten weeks after induction of flowering (* p < 0.05, *** p < 0.001, ANOVA). 134
Figure 5.14. Effect of Daylength on Plant Height ± SD (n=20) ten weeks after induction of
flowering. 138
Figure 5.15. Effect of Daylength on Yield of Botanical Raw Material ± SD (n = 20 plants)
ten weeks after induction of flowering. In the Analyses of Variance, the levels of
significance were shown as * p < 0.05 and *** p < 0.001. 138
Figure 5.16. Effect of Daylength on cannabinoid yield ten weeks after induction of
flowering. (Paired t-test, two tail ** p < 0.01). 139
Figure 6.1. The mean height (± SD) of G5 M16 crop as observed at weekly intervals in
2005. Thirty plants were measured on each occasion. Data points are shown as square
symbols during the establishment and vegetative phase. Data points are shown as
triangle during the flowering (generative) phase. 150
Figure 6.2. A comparison of the pattern of stigma senescence (%, ± SD, n = 7) in 2006
field trial plants between 10th September and 15th October with that observed in five
consecutively grown routine indoor crops of the same variety (± SD, n = 5). 153
Figure 6.3. Yield of Botanical Raw Material in the 2006 trial showing the effect of
planting date and harvest date. The results are the mean dry weights (gm-2 ± sd)
harvested from seven replicates. 156
Figure 6.4. Potency of Botanical Raw Material in the 2006 trial showing the effect of
harvest date. The results are the mean % CBD (± SD) content of samples from each of
the seven replicates, as measured by GC. Regression model, n = 7, p = 0.0028, R2 =
0.983. 157
xx
Figure 6.5. The yield of CBD in the 2006 trial, showing the effect of harvest date. The
results are the mean CBD yields (gm-2 ± SD) produced in each of the seven replicates.
Regression model, p = 0.0011, R2 = 0.994 157
Figure 6.6. A comparison of the terpene profiles, as a proportion of the total peak area, of
enriched trichome preparations prepared from glasshouse and outdoor crops († =
Caryophyllene Oxide). The results are the mean of five samples produced at weekly
intervals towards the end of flowering (± SD). Glasshouse plants had been in a 12 hour
day length for 6 to 10 weeks. Field grown crops were sampled between September 17th
and October 15th. (ANOVA, Glasshouse v Field, * p < 0.05, ** p < 0.01, *** p < 0.001).166
Figure 6.7. Fungal damage of a cannabis inflorescence due to Botrytis cinerea. 168
Figure 6.8. A cannabis plant at the late flowering stage. Resinous bracts are unaffected
but leaves below the inflorescence are heavily grazed. In some cases little more than
the midrib of the leaf remains. 169
Figure 6.9. The rate of moisture loss of field grown cannabis, when dried at three
temperatures (30, 40 and 50ºC). The results are the mean of three crops dried in
2004-2006 (± SD) and show the pattern of moisture loss until mean moisture content
was <15%. 171
xxi
List of Tables
Table 1.1. Examples of plant derived drugs and modern semi-synthetic drugs made from
the secondary metabolites of outdoor grown plants. 3
Table 1.2. A suggested classification of Cannabis sativa L (Sytsma et al., 2002) 8
Table 1.3. The predominant cannabinoids found in Cannabis sativa and their main
catabolites. 9
Table 2.1. Photographic, microscopy and other miscellaneous items and commercial
sources. 29
Table 2.2. Number of each type of sample received from each constabulary. (In addition,
one sample of cannabis powder was received from Kent). *The low number of resin
samples from Merseyside was due to the late inclusion of such samples from this
constabulary. 34
Table 2.3. The median and the range of potencies of five cannabinoids (% w/w) in resin,
herbal cannabis, sinsemilla and cannabis powder, seized in five constabularies in
England in 2004/5. 35
Table 3.1 The names and suppliers of the cultivars used. 50
Table 3.2 Photographic, microscopy and other miscellaneous items and commercial
sources. 51
Table 3.3 The 1-9 scale for overall capitate stalked trichome resin head colour. Within
each sample some variation would occur. 56
Table 3.4. The potency (THC content) of yellow and green leaf tissue of the variegated
cultivar G60-M55 assessed in each of two tests (± SD). The potency in the second test
is also shown as a weight of THC per unit area. 81
Table 4.1 Name, source and chemotype of clones used. 87
Table 4.2 Miscellaneous items and commercial sources. 87
xxii
Table 4.3. A comparison of the cannabinoid profile and content of fresh cannabis floral
material (clone G1 M3) and sieved trichome filtrates made there from. One sample of
each fraction was produced. Analyses show mean analytical results of four
subsamples (±SD). Also shown is the cannabinoid content of the floral material before
trichome extraction and the spent pulp after extraction. 90
Table 4.4. The relative proportions of CBC and THC in sessile trichome rich preparations
made by sieving dislodged trichomes from vegetative foliage of high-THC clone G1 M3.
93
Table 4.5. The proportions of principal cannabinoids (±SD) found in a trichome-rich
filtrate clone M240, containing only sessile trichomes. One bulk sample was prepared
and four sub-samples analysed. Also shown is the purity of the CBC (expressed as a
%w/w of total cannabinoids detected) within the foliage from which these trichomes were
collected. 94
Table 4.6. The terpene profile of essential oils produced by steam distillation of glandular
trichomes extracted from high-THC clone G2 M6 at various stages in the plant’s
development. The results are the relative peak area after analysis of the essential oil by
GC. Missing values occur where individual terpene contents were below the detectable
limits. 96
trichomes extracted from high-CBD clone G5 M13 at various stages in the plants
development. The results are the relative peak area after analysis of the essential oil by
GC. Missing values occur where individual terpene contents were below the detectable
limits. 97
Table 4.8. The cannabinoid profile of the original plant material (high-THC clone G2 M6)
from which the trichome rich preparations were made. Six subsamples were combined
and milled to produce one sample for analysis by GC. A missing value denotes that
cannabinoid level was below the detectable threshold. 98
xxiii
Table 4.9. The cannabinoid profile of the original plant material (high-CBD clone G5
M13) from which the trichome rich preparations were made. Six subsamples were
combined and milled to produce one sample for analysis by GC. A missing value
denotes that cannabinoid level was below the detectable threshold. 99
Table 5.1 Plant Propagation Materials 107
Table 5.2 Germplasm Details 108
Table 5.3 Light Measurement Equipment 109
Table 5.4 A comparison of Botanical Raw Material yield and potency of a THC and a
CBD chemovar when grown in two irradiance levels during winter. 121
Table 5.5 The relative proportions of CBD and THC during plant development in five
clones derived from variety G159. Results are shown as the proportion of CBD
expressed as % of CBD+THC (± SD). The regression calculations test the significance of
the changing proportion of CBD and THC in each clone, between the 4th and 10th week
in 12 h daylength. 130
Table 5.6. The falling late-summer daylength (hours.minutes) in a range northern
hemisphere cannabis growing areas. * Contrasting locations at same latitude in
Afghanistan, USA and Morocco. 131
Table 5.7. A comparison of the proportion of senesced stigmas observed on ten clone
when induced to flower in daylengths of twelve or thirteen hours. Assessments were
made eight and ten weeks after the plants were placed in short daylength. ** Significant
difference (p < 0.01, paired t-test). 132
Table 5.8. Effect of Daylength on cannabinoid yield (g m-2), eight and ten weeks after
induction of flowering. 135
Table 5.9. The effect of day length during flowering on cannabinoid profile. Results
shown are the proportion of CBG and THCV, expressed as a % of the CBG+THC or
THCV+THC total, in ten clones after eight and ten weeks in short daylength. 136
xxiv
Table 5.10 A comparison of the proportion of senesced stigmas on ten clones in twelve
and eleven hour daylength regimes when assessed eight and ten weeks after the
induction of flowering. Just one overall visual assessment was made for each clone. 137
Table 5.11 The effect of day length during flowering on cannabinoid profile. Results
shown are the proportion of CBG expressed as a % of the CBG+THC+CBD total, in six
clones after eight weeks in short daylength. 140
Table 6.1. Propagation Materials and Equipment used in the field trials program to
evaluate the outdoor propagation of Cannabis sativa L. 146
Table 6.2. Summary of Agronomic and Yield Data from field trials performed between
2000 and 2006. *The 2000 trial was performed before commencement of this thesis, and
the data is included for comparison. 150
Table 6.3. A comparison of the pattern of inflorescence development and stigma
senescence in indoor and outdoor crops of cannabis chemovar G5. The stages of
development of the glasshouse crop are shown from the point at which the plants are
moved into a 12hour light/12hour dark until they are routinely harvested eight weeks
later. The field crop development is shown from mid-August, just before stigma
formation commenced. 153
Table 6.4. A comparison of the terpene profile of freshly harvested fully mature field
grown cannabis leaf and flower material (cultivar G5 M16) and ETP made from the same
fresh material (2005 Field Trial). Also included is ETP made from similar mature
glasshouse grown material. The data show the relative peak areas when assessed by
GC at Botanix Ltd. In each case one batch of material was analysed. 155
Table 6.5. The changing proportions of CBG and THC with respect to CBD (Mean ± sd)
in enriched trichome preparations produced from plants harvested at weekly intervals
between 17th September and 15th October. Just one ETP sample was made on each
date after bulking together one plant from each of the seven replicates. Three
subsamples of each preparation were analysed. 160
xxv
Table 6.6. A comparison of the terpene profile of steam-distillates of enriched trichome
preparations made from freshly harvested plants on five dates between 17th September
and 15th October 2006. One bulked sample was analysed on each date. 163
Table 6.7. Ratio of eight terpenes in steam distilled enriched trichome preparations made
from freshly harvested field grown plants of cultivar G5 M16. The result for each
terpene is expressed as a weight percentage (% w/w) of the total within each column.
The table also shows the ratio of myrcene (the dominant monoterpene) and
trans-caryophylene (the dominant sesquiterpene). 164
Table 6.8. Comparison of myrcene/trans-caryophylene ratios when calculated from
Relative Peak Areavalues and w/w data. 165
Table 6.9. The relative proportions of THC and CBD synthesised in heterozygous BTBD
clones derived from variety G159. The proportion of THC produced is shown as a % of
the THC+CBD total. Just one sample of dry inflorescence material was analysed from
each plant. 165
Table 6.10. Summary of pest problems experienced, in decreasing order of magnitude
using a subjective 1-5 score. * Botrytis initially minor but extremely severe in late
harvested plots. † Symptoms not recognized as pest damage until 2005. ‡ Aphids
absent in trial plots but moderate infestation of black bean aphid Aphis fabae observed
on neighbouring cannabis seed-crop with lower secondary metabolite content. 168
Table 6.11. The level of infection with Botyrytis cinerea observed on plants harvested on
five dates between 17th September and 15th October 2006. Scores are the mean %
infection rates in the seven plots of each treatment. 169
xxvi
Abbreviations
ACMD Advisory Council on the Misuse of Drugs
ANOVA Analysis of Variance
API Active pharmaceutical ingredient
BD/BD Homozygous CBD chemotype
BD/BT Heterozygous mixed THC/CBD chemotype
BDS Botanical Drug Substance
BMA British Medical Association
BRM Botanical Raw Material
BT/BT Homozygous THC chemotype
cAMP cyclic Adenosine 5’- monophosphate
CB1 and CB2 Cannabinoid Receptors 1 and 2
CBC Cannabichromene
CBCA Cannabichromenic acid
CBCV Cannabichromevarin
CBCVA Cannabichromevarinic acid
CBD Cannabindiol
xxvii
CBDA Cannabidiolic acid
CBG Cannabigerol
CBGA Cannabigerolic Acid
CBN Cannabinol
CI Confidence Interval
CNS Central Nervous System
EMEA European Medicines Agency
ERK Extra-cellular regulated kinase
ETP Enriched trichome preparation
GC Gas chromatography
GAP Good Agricultural Practise
GDP Guanosine 5’-diphosphate
GMP Good Manufacturing Practice
GTP Guanosine 5’-triphosphate
GWP Good Wild crafting Practice
HPLC High Performance Liquid Chromatography
HPS High Pressure Sodium
xxviii
MBFU Mercury Vapour Lamp
MH Metal Halide
MHRA Medicines and Healthcare Products Regulatory Agency
MS Multiple Sclerosis
NIAB National Institute of Agricultural Botany
PAR Photosynthetically Active Ration
SD Standard Deviation
SE Standard Error
TEM Transmission Electron Microscope
THC Tetrahydrocannabinol
THCA Tetrahydrocannabinolic acid
THCV Tetrahydrocannabivarin
THCVA Tetrahydrocannabivarinic acid
United Nations Childrens’s Fund formally United Nations

UNICEF
International Emergency Fund
UV-B Ultra violet (Band B)
W m-2 Watts per square meter
WHO World Health Organisation
xxix
xxx
Chapter 1 Introduction
Chapter 1 INTRODUCTION
1.1 Plants as a source of medicines – past and present

Since modern man (Homo sapiens L) first walked this earth he has probably always
exploited plants as medicines. Evidence from burial sites suggests that the
Neanderthals (Homo neanderthalis) also shared this practice. Man’s present day
closest relative the chimpanzee (Pan troglodyte L), medicates with many plant species.
Observations suggest that when the chimp purposefully swallows unchewed leaves of
one of these, Aspilia mossambicensis, it is making use of the dense foliar trichomes to
ensnare gut parasites (Huffman et al. 1996). Conceivably, common ancestors of Homo
and Pan also shared similar practices, since many less developed animal species are
frequently observed consuming plants for what appear to be medicinal, rather than
nutritional, purposes (Sumner, 2000).
There are written records dating back several millennia BCE, from early civilisations on
most continents, which describe man’s use of plants as medicines. These include
American Indian, early European, Middle Eastern, Ayurvedic (Indian sub-continent),
Chinese, Korean and Japanese, and Aboriginal cultures. Cannabis features in many of
these, and in oriental and Middle Eastern countries its use can be traced back many
thousands of years. Evidence suggests that around 3000 BCE Cannabis sativa L was
used as an Ayurvedic (Russo, 2004) and Chinese medicine (Mechoulam, 1986). In
Egypt, mention of medicinal uses of cannabis was written in the Papyrus Ramesseum
III (circa 1700 BCE). More detailed uses were recorded in the Ebers Papyrus (circa
1600 BCE) which describes the use of cannabis as a decoction in enemas,
applications to the eye and topically in the form of medicated bandages (Mannische
1989). An archaeological discovery of cannabis in China dating back 2700 years is
also supportive evidence of its early medicinal use (Russo et al., 2008). In Hebrew,
Greek and Roman texts there are references to the sedative hypnotic uses of cannabis.
These sources refer, inter alia to its use in obstetrics and gynaecological products.
There are also ancient references to the inhalation of smoked cannabis. Most ancient
Middle Eastern and Asian civilisations record it being smoked for medical and ritual
purposes. An archaeological find in Jerusalem, from the fourth century AD, indicates
the use of cannabis vapour in an enclosed environment by women during labour (Zias
et al., 1993). In museums around the Mediterranean (e.g. Empurius North of Barcelona)
there are collections of surgical artefacts, which include pipes for smoking drugs. This
is at least a millennium before the introduction of smoking tobacco from the New World.
1
According to a joint UNICEF and WHO report (UNICEF, 1992) of the 80% of the
world’s population living in developing countries, only 15% had access to modern
scientific medicine; the rest depended on traditional indigenous systems of health care
in which herbal medicines played a part. For about three-quarters of the world
population there is therefore complete reliance on plant-derived medicines. In 1976, in
modern western medicine, plant derived active ingredients were still represented in up
to 25% of prescription-drugs (Farnsworth and Morris, 1976). In developed countries
plant-derived pharmaceuticals contribute more than $30 billion of sales revenue to the
industry. Over 60% of anti-cancer drugs and 50% of cardiovascular and analgesic
treatments are derived from plants (Fowler and Law, 2006).
The World Health Organisation (WHO) reported that more than 21000 plant species
are regularly harvested for the production of medicines. The vast majority of these are
harvested in their natural environment. This is especially the case in areas with low
technological and economical development (Europam, 2006). The collection of wild
plants – so called wild crafting – has many drawbacks. Over zealous collection of wild
plants for medicinal use has actually threatened the existence of some species e.g.
Galanthus woronowi for the production of galanthamine, and Coleus forskohlii for the
supply of forskolin (Evans, 2002). Consequently this type of collection has led to some
plant species receiving CITES protection to help prevent their extinction. To maximise
plant quality and to minimise the impact of wild-crafting on species survival and the
environment, the modern pharmaceutical industry stipulates that the harvest should
comply with the Good Wild-crafting Practice (GWP) Guidelines (Europam, 2006). To
maximise the sustainable yield from some threatened wild species, research has
established strategies for regulating and improving harvest techniques. An example is
the Chinese tree Camptotheca acuminata (Vincent et al. 1997), which is the source of
the major anticancer drug camptothecin, and had a market value of $5500 million in
2006 (Fowler and Law, 2006). Wild plants are generally highly variable in their
secondary metabolite content and they may at times also be in short supply. These
concerns, and the threat to species survival, can at least be partly overcome by
commercially growing crops for the pharmaceutical industry. Surprisingly perhaps, only
about a hundred plant species are specifically cultivated for this purpose (Europam,
2006). Growing plants for medicinal use goes back many millennia. By 660 BCE,
Assyrian herbalists were cultivating many plants (Baker, 2002). Pliny the Elder, in his
Naturalis Historia of 77AD, described the medicinal qualities of cannabis, and detailed
the recommended planting and harvest timings (Pliny, 1951).
2
A vast number of plant secondary metabolites form the active ingredients of modern
drugs. By definition, these are organic compounds which are not directly involved in the
normal growth, development or reproduction of organisms. However they add to the
plant’s survival chances by improving its resistance to predators, parasites and a range
of environmental stresses. Some secondary metabolites are highly purified before
being formulated as medicines. Others provide a starter material from which semi-
synthetic drugs are produced. Examples of both types are shown in Table 1.1.
Plant-derived Semi-synthetic Chemical Type Medical Use Plant Source

ingredient/ drug Product
precursor
Codeine, _ Opiate alkaloid Analgesic Papaver
Morphine somniferum
Taxol _ Diterpene ester Antineoplastic Taxus
Brevifolia
Vinblastine, _ Bis-indole Antineoplastic Catharanthus
Vincristine alkaloid roseus
Podophyllotoxin Etoposide Lignan Antineoplastic Podophyllum
peltatum
Diosgenin Progesterone Aglycone Birth control Dioscoria
Steroid sylvatica
Camptothecin Topotecan Indole alkaloid Antineoplastic Camptotheca
acuminata
Physostigmine Neostigmine Alkaloid Cholinergic Physostigma
venenosum
Table 1.1 Examples of plant derived drugs and modern semi-synthetic drugs made from the
secondary metabolites of outdoor grown plants.
The plant kingdom has also enabled the production of so called phytopharmaceutical
or ‘botanical drugs’. These are defined as well characterised, multi-component
standardised drugs extracted from plant sources. The medicine VeregenTM, derived
from green tea Camellia sinensis, and approved for the topical treatment of warts
(Medigene Inc.) is such an example. In 2004 the United States Food and Drug
Administration issued the Botanical Drug Guidance which made it possible to bring to
market a complex mixture for which evidence of adequate safety and efficacy had been
established (FDA, 2004). This could result in the successful company being awarded a
period of exclusivity. By 2006 VeregenTM was the only medicine to have been
successful (NDA 21-902). Other botanical medicines are prescribed in some European
countries but as of 2008 none were dispensed in the UK. GW Pharmaceuticals
3
commenced clinical trials in the US in 2008, to evaluate the efficacy of the cannabis-
based botanical medicine Sativex® for the control of pain in terminal cancer patients. If
trials were successful, a New Drug Application (NDA) would be sought.
1.2 Cannabis Botany

Cannabis is a tall upright annual herb. It is generally dioecious i.e. producing separate
male and female plants (Figure 1.2a), but fibre hemp varieties have been specifically
bred to be monoecious (hermaphrodite) (Small and Cronquist, 1976). The leaves are
palmate, and in the iconic image of a cannabis leaf there are seven lobes, the lowest
pair showing as backwards-facing spurs. However this number and shape is not fixed.
On seedlings the first pair of leaves is typically monophylous (single lobed), the second
pair having three lobes and the next pair five. In many plants, especially of central
Asian origin, the number does not extend beyond five while in others the number can
extend to around thirteen. Leaf size and shape differs markedly according to genetic
origin, three contrasting examples being illustrated in Figures 1.1a – c.
Figure 1.1. Contrasting leaf morphology in three clones of Cannabis sativa L., (a) CBD
chemotype G5 M16 cv Gill, (b) THC chemotype G1 M3 cv Guinevere (c) Afghan landrace
clone M146 (Illustrations by Valerie Bolas, commissioned by GW Pharmaceuticals).
Cannabis is a wind pollinated species. The males, which are generally taller than the
females (Figure 1.2a), commence flowering first. The specimen in Figure 1.2a was
grown in still conditions and leaves appear yellow under the deep covering of pollen.
When mature, the sepals on the male flowers open to expose the anthers, which hang
freely on fine filaments (Figure 1.2c). The exposed anthers soon dehisce to shed pollen
onto any passing air current. Shortly after the cessation of pollen production the male
dies, but females from the same population will continue to mature for up to several
weeks. Females produce inflorescences containing vast numbers of florets over a
4
period of several weeks. This period is extended if pollen is not received. An example
of a well-developed inflorescence is shown at the front of this thesis (Page ii). Just as
leaf shape varies according to provenance, the shape of the inflorescence does also.
The inflorescences of the female can become very sticky due to a covering of resinous
glandular trichomes. These are the main source of the cannabinoids, a group of
terpenoid compounds unique to Cannabis. Consequently, the female inflorescence is
the most importance plant part to those exploiting it for its recreational or medicinal
properties. Part of a fertile female inflorescence is shown in Figure 1.2b, and the fertile
white stigmas are clearly visible against the bracts of this purple variety.
As a result of pollination the female develops copious numbers of seeds, or more

correctly achenes. These have been collected by man more several millennia for their
great nutritional value, the oil produced from crushed seeds also being used for
culinary purposes and to light lamps. The stems of some varieties are a rich source of
fibre which has also been used by man for several millennia for the production of paper,
rope, textiles and more latterly building materials and automotive parts (Wills, 1998).
During the reign of Henry VIII and Elizabeth I, farmers in this country were legally
required to grow hemp to ensure that the navy had sufficient rope and sail cloth
(Hansard (Australia), 1996). So common was Cannabis sativa L in the English
countryside that in his herbal of 1653 Nicholas Culpeper wrote, “This is so well known
in this country that I shall not need to write any description of it”.
5mm
(a) (b) (c)
Figure 1.2. (a) Male (left) and female cannabis (right) in later stage of flowering. (b) Female
cannabis inflorescence. (c) A cluster of male flowers with sepals split open and reflexed to
expose the anthers.
As a consequence of its various uses, growing of Cannabis sativa L spread to all

continents, apart perhaps from Antarctica. The original source of the species is heavily
debated but is commonly thought to have evolved in central Asia in a region
approximately 30° - 35°N with the Himalayas to the south, Turkestan to the west,
5
Pakistan to the East and Southern China as its probable northernmost extreme (Wills,
1998). This is just a few degrees south west of Western China where its closest relative
the hop Humulus lupulus L is believed to have originated (Neve, 1991a). From these
regions the two species spread and, heavily influenced by man, adapted to a range of
latitudes, habitats and growing methods.
1.3 Cannabis Taxonomy

The binomial Cannabis sativa L. carries the suffix L to record that the taxonomist Carl
Linneaus adopted this name in his Species Plantarum of 1753. However, the binomial
had been used much prior to this, by Leonardt Fuchs in his Kreuterbuch of 1543. Just
as Linnaeus recognised one species, most modern day taxonomists also regard
Cannabis as monotypic, with the species as one isolated gene pool (Harlan and de
Wet, 1971). Within that species several subspecies are sometimes identified (Small
and Cronquist, 1976). The biological/reproductive definition of a species states that all
specimens of a population are of a single species if they are naturally able to sexually
reproduce, generating fertile offspring. This is the case in the genus Cannabis, and by
this definition therefore there are no clear biological grounds to separate it into different
species (Schultes et al., 1974). However modern Cannabis taxonomy remains
confused, as some scientists and commentators prefer to define species according to
typological or morphological characteristics. Chemotaxonomy has also been used to
catagorise cannabis populations according to their terpenoid content (Hillig and
Mahlberg, 2004).
In 1785 Lamark described the genus as polytypic and introduced the separate name
Cannabis indica for plants grown in India. Such plants he regarded as being a different
species to the European ‘Cannabis sativa type’ based upon their different morphology,
geographic range, pronounced smell and greater narcotic potency. In the twentieth
century Lamark’s name Cannabis indica came to be widely used to describe the short
wide-leaved plants indigenous to Afghanistan, like the example in Figure 1c. However,
re-examination of the Lamark’s original Cannabis indica herbarium samples shows his
plants to have been narrow leaved. Those modern day taxonomists who adhere to a
belief that Cannabis sativa L and Canabis indica Lam were truly separate species
would more likely have identified Lamark’s samples as Cannabis sativa L. Many other
species of Cannabis have been proposed of which just Cannabis ruderalis Janisch has
met wide use. This name was used to identify weak low-potency ruderal (road side)
plants from eastern Europe which produced small marbled achenes with a strongly
constricted abscission layer (de Meijer, 1994).
6
The argument that Cannabis is polytypic gained legal significance from 1972 onwards,
when an increasing number of court cases occurred in the USA, with defence lawyers
challenging the taxonomy in convictions involving marijuana. United States law
attributed the illegal recreational marijuana solely to the species Cannabis sativa L.
Defence lawyers, claiming that their defendants were involved with Cannabis indica or
other suggested species, argued that there was no case to answer (Small, 1976).
Perhaps partly stemming from this challenge to the law, a large proportion of the
commercial suppliers, advisors and commentators currently in the recreational
cannabis industry still commonly refer to the ‘species’ Cannabis indica and Cannabis
ruderalis in addition to Cannabis sativa (Snoeijer, 2002). Within this field, many
appear to feel some empathy or romantic association with (or a perceived financial
dependence upon) the important part that Cannabis has held within the anti-
establishment movement. For the remainder of this thesis, the species is regarded as
monotypic with the name Cannabis sativa L.
Within the last hundred years the taxonomy of Cannabis above species level has also
been cause of much debate. The genus Cannabis has most commonly been regarded
as belonging to the family Cannabaceae of the order Urticales (Raman, 1998).
Taxonomists originally placed Cannabis in Urticaceae (nettle family) but in the early
20th century some moved it to the Moraceae (fig family). However, the number of
dissimilarities between Cannabis and the nettles or figs led to this genus being
allocated, in the 1960s, to the new family – the Cannabaceae – along with just one
other genus Humulus (the hops). The existence of the family Cannabaceae has gained
widespread support. However, its place in the order Urticales (of the superorder
Dillenidae) has been challenged in recent years. Some taxonomists now place the
family in the order in Rosales along with the Urticaceae and the Moraceae, as shown in
Table 1.2.
The various suggested sub-species are described elsewhere (Raman, 1998). As

stated earlier, chemotaxonomy has been used to split Cannabis populations between
putative species and subspecies (Hillig and Mahlberg, 2004). European sources of
Cannabis, usually grown for fibre or seed, typically contain a much higher cannabidiolic
acid (CBDA) to delta-9-tetracannabidiolic acid (THCA) ratio than more tropical plants
where the majority of the field grown illicit crops are raised. (Small and Beckstead,
1973a, b). It has been postulated that the biosynthesis of the cannabinoids THCA or
CBDA from the common precursor cannabigerolic acid (CBGA) is controlled by the
presence of co-dominant alleles at a common locus. These give rise to homozygous
chemotypes with a high THCA or CBDA purity and a heterozygous chemotype
7
containing a THCA/CBDA mixture. All heterogenous cannabis populations will contain

a mixture of all three chemotypes (de Meijer et al., 2003). Differences in the
THCA/CBDA balance of geographically contrasting populations may be due to
environmental pressures e.g. prevailing temperature (Boucher et al., 1974). However,
the greatest effect on chemotype will be the selective pressure exerted man as a result
of a local need of Cannabis for fibre, seed or drug purposes. The genotypes used in
the research for this thesis were bred from varying putative sub-species, which
originated from a wide range of provenances. Many of these genetic sources were
landrace materials and others specifically-bred agricultural varieties.
Kingdom Plantae – Plants

Subkingdom Tracheobionta – Vascular plants
Superdivision Spermatophyta – Seed plants
Division Magnoliophyta – Flowering plants
Class Magnoliopsida – Dicotyledons
Subclass Hamamelidae
Order Rosales
Family Cannabaceae – Hemp family
Genus Cannabis L. – hemp
Species Cannabis sativa L.
Table 1.2 A suggested classification of Cannabis sativa L (Sytsma et al., 2002)
1.4 UK Medicinal Cannabis Use - History and Legal Complications.

The medicinal properties for cannabis were claimed in the UK many centuries ago.
Interest expanded greatly in the nineteenth century following the research of British
expatriate surgeon O’Shaughnessy, who used ethanolic tinctures of cannabis for the
treatment of pain (Robson, 1999a). While working for the East India Company in
Calcutta O’Shaughnessy described the methods of use of cannabis by the native
population where It had a variety of applications, particularly as a sedative and
analgesic. In a series of careful experiments, O’Shaughnessy defined the method of
extraction and useful doses of the preparation so produced. In his monograph he
described using alcohol to produce tinctures, which were increasingly imported into the
UK. In addition to a description of the useful properties of gallenical preparations
based on cannabis, O’Shaughnessy described some of the adverse events which
8
followed its use and really established the place of cannabis tincture and cannabis
extract in Victorian medicine.
By the end of the century the medicinal use of cannabis was well enough established
for it to be the subject of a reference in Merck’s Manual (1899). This described
cannabis as a hypnotic sedative and very useful for the treatment of hysteria, delirium,
epilepsy, nervous insomnia, migraine, pain and dysmenorrhoea. General practitioners
continued to prescribe complex cannabis formulations up until the middle of the
twentieth century. Many of the central nervous system indications for which cannabis
was used (analgesic, hypnotic, sedative and antiepileptic) were by then met by the
benzodiazepine group of drugs and analgesics like paracetamol and codeine. While
these guaranteed reliable dose control, it was difficult to obtain a consistent response
to cannabis because of the variable cannabinoid content of the plant material available
(Notcutt, 2004). There were also growing international concerns of social problems
caused by recreational cannabis use. This, and awareness of the adverse effects of
cannabis, finally led to its prohibition in the UK when the UK Government ratified the
1925 Geneva Convention on the manufacture, sale and movement of dangerous drugs.
It did however remain available through pharmacies, until it was completely outlawed
as a medicine by being declared a Schedule 1 substance in the Misuse of Drugs Act
1971. This act also outlawed the production and possession of cannabis for
recreational purposes. Amongst three categories of decreasing seriousness described
in Schedule 1 (A, B and C) cannabis was initially placed in Class B. This indicated that
those caught in possession of a small or moderate quantity for personal use would
likely attract a court fine. Repeat offenders and those supplying moderate quantities of
cannabis would be more likely to be sentenced to a community (i.e. non-custodial
penalty), while those producing cannabis and/or supplying large quantities could be
imprisoned for up to fourteen years. In 2004 the UK Government reclassified cannabis
as a Class C drug, which meant that possession could attract a much smaller
maximum prison sentence. Possession of small quantities would be more likely to
attract a police fine, without the offender needing to attend a Magistrates Court. In
2009, contrary to scientific advice from the Advisory Council on Misuse of Drugs,
cannabis was reclassified as Class B (ACMD, 2008).
Despite its continuing prohibition, in the last twenty years in the UK, an increasing
number of patients with severely debilitating diseases such as multiple sclerosis have
used illicit cannabis to obtain symptom relief (Whittle, 2004). Smoking cannabis for
recreational or medicinal reasons in the UK was almost unknown until the 1950s,
although recreational cannabis smoking had become common a decade later (Robson,
9
1999a). Most medicinal users of illicit cannabis would have also smoked the material,
although some would have ingested it in cooked form. A number of commercially
available books described recipes for foodstuffs containing the drug. The organized
production of such foodstuffs for alleged medicinal use resulted in two well publicized
convictions in 2006. A small survey carried out by the UK newspaper Disability Now in
1997 reported that amongst two hundred medicinal users of illicit cannabis, 20% were
using the drug to treat symptoms of multiple sclerosis and similar numbers for spinal
injury and back pain (HLSCST, 1998). A more extensive UK survey amongst those
using illicit cannabis medicinally received 2969 replies. It suggested that 136 diseases
were being treated with cannabis, the most predominant being Chronic Pain 25%,
Multiple Sclerosis 22%, Depression 22%, Arthritis 21% and Neuropathy 19% (Ware et
al., 2005). As these numbers suggest, some respondents were using cannabis for the
relief of more than one disease. The use of cannabis for the treatment of symptoms of
multiple sclerosis was well known in the UK. The high incidence of usage in treatment
of migraine and the pain of rheumatoid arthritis was perhaps less expected. In the UK
there was relatively little use for stimulation of appetite in sero-converted patients with
HIV and AIDS, although in the USA this was a significant indication for use of both
smoked cannabis and the synthetic-cannabinoid product Marinol®.
1.5 The influence of the BMA and the House on Lords Select
Committee on Science and Technology on UK Cannabis Research.
In 1997 the British Medical Association published a highly-influential report on the
therapeutic uses of cannabis (BMA, 1997). The report acknowledged the weight of
evidence for the drug’s efficacy in treating spasticity, nocturia and central pain in those
with spastic conditions such as multiple sclerosis, and stated that clinical trials in this
field merited a high priority. The report acknowledged that cannabinoids were
undoubtedly effective as anti-emetic agents in vomiting induced by chemotherapy and
anti-cancer drugs. More research was recommended to identify which cannabinoids
had the optimal therapeutic profile. The undoubted analgesic effects of some
cannabinoids were also acknowledged. The report cited research with cannabinoids in
cancer patients which showed that the pain control and cannabimimetic effects were
not inseparable (Evans, 1991). The term cannabimimetic in this context is defined as
pertaining to the pharmacological properties of Δ-9 tetrahydrocannabinol, which was a
cannabinoid identified as the main psychotropic ingredient of cannabis in 1964 (Gaoni
and Mechoulam, 1966). Cannabinoids were envisaged as useful adjuncts to standard
analgesics and hospices, pain control clinics and post-operative wards were suggested
10
as ideal settings for such research. This option was further supported by the
observation that mixtures of the cannabinoid THC and the opiate morphine had
exhibited synergistic analgesic activity (Cichewicz, 2004). To facilitate trials it was
stated that the World Health Organisation should advise the United Nations
Commission on Narcotic Drugs to reschedule certain cannabinoids under the United
Nations Convention on Psychotropic Substances, and in response the Home Office
should alter the Misuse of Drugs Act accordingly. In the absence of such action from
the WHO, the government was advised to consider changing the Misuse of Drugs Act
to allow prescription of cannabinoids to patients.
In the light of heightened interest in cannabis, and particularly the report by the BMA
and the House of Lords Select Committee on Science and Technology (HLSCST)
under the chairmanship of Lord Perry, was requested to examine the scientific and
medical evidence to determine if there was a case for relaxing some of the existing
restrictions on the medical uses of cannabis (HLSCST, 1998a, b). The Committee
considered was that if there is a clinical benefit accruing from the use of cannabis or
one of its constituents, then it should be regarded as a medicinal substance. Legally,
under the Medicines Act, medicinal substances could only be supplied properly if they
have been assessed by the Medicines Control Agency (now the Medicines Health
Regulatory Agency – MHRA) and a license issued by the regulatory authority. It very
strongly recommended that government should provide an environment in which
interested parties could research and develop cannabis-based medicines. If their
quality, safety and efficacy were adequately demonstrated, the developers should be
allowed to apply for a Medicines Act License and supply the product on prescription.
One stated criticism of an otherwise excellent report was that the BMA had focused on
single molecules. As pointed out, herbal cannabis contains a mixture of active
compounds, more than one of which possibly contributed to the therapeutic action.
Sensible patients who appeared to be benefitting from herbal cannabis did not report
an equal benefit when given just one of the active ingredients. Possible synergy
between active compounds was a possible explanation. The action of herbal cannabis
itself still needed definition (Wall, 1998).
1.6 International Legal Attitudes to Medicinal Cannabis

The United Nations Single Convention on Narcotic Drugs Schedule IV made cannabis
the subject of special restrictions 1961. Article 2 stated that individual countries should
protect their public using whatever means were regarded appropriate according to
conditions prevailing in their country. Protection of the public could necessitate
11
prohibition of the production, manufacture, export or import, trade in or possession of

any such drug - except for amounts which may be necessary for medical and scientific
research only (including clinical trails). In some countries it could be felt appropriate to
tolerate the possession of small quantities of the drug for recreational use.
In 1998, the UK legal regime regarding medicinal use of cannabis could be described
as one of the most restrictive in the world (HLSCST, 1998a). The following review of
the international legal status is not intended to be comprehensive, but describes the
history and current situation in a number of major countries.
1.6.1 USA
In the early 19th Century physicians in the USA (the world’s largest pharmaceutical
market) prescribed cannabis freely and several preparations were widely available. In
1894 the Indian Hemp Commission Report talked favourably of cannabis drugs and
recommended that cannabis should be controlled through taxation and regulation
rather than prohibition. However, as in Europe, newly discovered opiate-based drugs
were eclipsing cannabis as a medicine and its use declined.
Through the early 20th Century recreational cannabis use in the States was commonly
linked with hard-drug addiction, and it came to be increasingly regarded as an
unwelcome practise favoured by undesirable elements in society. (Bonnie and
Whitebread, 1970). In the 1930s Harry J Anslinger, the first commissioner of the newly
formed Federal Bureau of Narcotics, cultivated society’s belief that cannabis
threatened to destroy the country’s moral fabric. One result of this was the introduction
in 1937 of the Marijuana Tax Act, which placed such draconian burdens on those
attempting to use cannabis for medicinal purposes that its use almost ceased. During
the 1960s opinions in the States were beginning to change. President Nixon appointed
The Shafer National Commission on Marijuana and Drug Abuse to review US policy
and their subsequent report, entitled Marijuana, a Signal of Misunderstanding 1972,
recommended a relaxation in the laws controlling cannabis use. Little changed
however. President Nixon had declared a ‘War on Drugs’ and he rejected the report
before even reading it (Russo, 2003).
A number of patients with a firm belief in medicinal cannabis remained outspoken and
one of these Robert Randall successfully brought a suit against the Federal
Government seeking medically supervised access to cannabis. Citing ‘Compassionate
Use’ he and others were granted legal access to ‘research grade material’ grown under
government supervision in Mississippi. In 1983 thirty four States had enacted
legislation making cancer and glaucoma patients eligible to apply for medicinal
12
cannabis through this scheme. However, in 1992 the Secretary for Health and Human
Services closed the program for new patients saying that it ‘sent the wrong message’
(Mead, 2004). Seeking to drive patients away from herbal cannabis as part of its War
on Drugs the federal government diverted them towards synthetic THC (dronabinol).
First formulated and launched on prescription in 1985 as Marinol the federal
government judged that this product had rendered smoked medicinal cannabis
unnecessary. However, absorption of this orally administered drug by the gastro-
intestinal tract is highly variable. In contrast to smoked cannabis, patients commonly
found it difficult to titrate the dose against their symptoms (Ohlsson et al., 1980).
Recognising this, in 1996 the states of California and Arizona authorized seriously ill
patients to use or cultivate, possess or use cannabis if recommended by a medical
doctor. However, the Federal government Controlled Substances Act prohibited
cannabis cultivation for any purpose and high-ranking Federal officials threatened that
physicians as well as growers could face criminal prosecution. The backlash from
doctors to this threat led the Drug Czar Barry McCaffrey to ask for review of the
medical evidence supporting cannabis as a medicine. The report of this review
Marijuana and Medicine: Assessing the Science Base was published in 1999 and
included many recommendations in favour of cannabis (Joy et al., 1999). By 2008
twelve states, covering approximately 20% of the population, authorised the use of
cannabis for medicinal purposes. To assist these patients, a large number Cannabis
Growers Clubs were formed. However, despite this, the federal government still
pursued the closure of these clubs and similar outlets with vigour. In California in 2008
over two hundred thousand patients had a written recommendation from a medical
doctor supporting their medicinal marijuana use. Of these, 40% could be regarded as
having a serious illness (Room et al., 2008). In this state the Federal Authorities
appeared to be losing their battle.
In the USA, the vehicle for regulation of research into cannabinoids is the National
Institute for Drug Abuse (NIDA). Most of the research, which is supported in the USA,
is directed towards mode of action studies and the cataloguing of adverse effects
produced by cannabis. The majority of the research is concerned with preclinical
studies and very little clinical work is supported in the USA other than investigations of
adverse effects on the psychological profile of recreational users. The net effect of
prohibition of cannabis in the USA has been that little or no clinical research under
therapeutic benefit has been carried out to date. In 2009, in the last days of the Bush
Administration, there appeared to be minimal appetite within the US government for
any type of cannabis reform. Against this difficult background, as stated earlier, in 2008
13
GW Pharmaceuticals commenced clinical trials in the US to evaluate the efficacy of the

cannabis-based botanical medicine Sativex® for the control of pain in terminal cancer
patients.
1.6.2 Canada
In 2001, Health Canada defined categories of patient eligible to receive access to
medicinal cannabis. These included individuals with acute pain, violent nausea or
other serious symptoms caused by multiple sclerosis, cancer, spinal injury or disease,
AIDS/HIV, severe arthritis and epilepsy. With doctor’s approval, these patients could
receive cannabis grown under the name CanniMed by the company Prairie Plant
Systems. It was also legal for individuals to grow Cannabis for personal medical use.
In 2005 Sativex® received provisional approval with conditions for the treatment of
central neuropathic pain in multiple sclerosis, and in 2007 for intractable cancer pain.
1.6.3 Mainland Europe

For decades, the Dutch Government has had the most relaxed attitude to cannabis use
within Europe. Possession of small quantities of cannabis for personal recreational use
has been tolerated. Many people used cannabis for medicinal purposes, often growing
their own plants. Much of the cannabis used would have been bought from the large
number of regulated ‘coffee shops’ that offered the sale and consumption of cannabis
on their premises. Under the Guidance of the Ministry for Health, Welfare and Sport, in
2003 a new source of ‘medical cannabis’ became available through pharmacies in the
Netherlands. This material simply consisted of unformulated dried cannabis floral
material. It contained high levels of THC and minimum amounts of other cannabinoids.
The cannabinoid content was stated as being within a tightly specified range and the
product was sterilised using γ-radiation. The product was much more expensive than
similar, albeit less-well regulated, floral material on sale in coffee shops. Perhaps
because of this, consumption was much lower than government predictions and the
continued supply of this material was threatened.
In November 2006, the Ministry stated that the German and Italian governments were
interested in accessing the Dutch medicinal cannabis. On 9th July 2008, the Austrian
Parliament approved Cannabis cultivation for scientific purposes under the Health
Ministry’s control. Possession of cannabis for recreational purposes remained an
imprisonable offence. Spain has undergone a progressive decrimilisation with regard to
cannabis possession. In 2001 the Catalonia region permitted the possession of
cannabis for medicinal purposes and Sativex® became available in that area.
14
Early in 2009 the results of a Phase III clinical trial showed that Sativex® was
significantly efficacious in treating spasticity in multiple sclerosis. This triggered an
application for the regulatory approval for Sativex in the UK and mainland Europe.
1.6.4 Ireland
Cannabis is not recognised as having any medical benefits according to Irish law
(Misuse of Drugs (Designation) Order (S.I. 69/1998). By European standards, Irish
Courts have treated medicinal users of illicit cannabis harshly. In 2003 however, the
Irish Medicines Board permitted clinical trials to be performed to evaluate Sativex®.
1.6.5 Australia
In Australia, laws differ between states. Those caught in possession of cannabis of up
in New South Wales would only receive a fine if found with less than 15 g. Other states
are more lenient, up to 50 g justifying a fine in Queensland, Tasmania and Victoria
(Lenton, 2004). Cultivation of cannabis plants similarly attracts differing sentences,
and in the Northern Territory is decriminalised. In January 2009, a four year trial
commenced to evaluate the medical use of cannabis to treat chronically or terminally ill
patients.
1.6.6 Japan
Cannabis possession in Japan is illegal for both recreational and medicinal use, and
hefty fines and imprisonment are imposed. However, the Otsuka Pharmaceutical
Company Ltd is collaborating with GW Pharmaceuticals Ltd to research a range of the
cannabinoids for use in oncology and CNS ailments. The production of some of these
cannabinoids is described in this thesis.
1.7 The choice of active pharmaceutical ingredients (APIs)

At the commencement of this thesis in autumn 2003 the sponsoring company
(GW Pharmaceuticals Ltd) was propagating two feedstocks to produce the medicine
Sativex®, which was undergoing Phase III clinical trials. One of these studies showed
that this medicine significantly reduced spasticity symptoms in patients with multiple
sclerosis (Wade et al., 2004). Following further studies (Barnes et al., 2006), the
medicine became available in Canada in 2005 for the treatment of central neuropathic
pain in multiple sclerosis and in 2007 for intractable cancer pain. These separate
feedstocks were the dried floral and foliar material of two very different chemotypes,
which contained predominantly Δ-9 tetrahydrocannabinolic acid (THCA) or cannabidiolic
Acid (CBDA) as the active ingredients. These terpenophenolic compounds are so
called cannabinoids and are unique to cannabis (Turner et al., 1980). A large
15
proportion of this thesis concerns the propagation, characterisation and optimization of

these two chemotypes. At the completion of this thesis in 2009, Sativex® had received
conditional regulatory approval for use against neuropathic pain and cancer pain in
Canada. The medicine was also available in parts of the European Union as an
unlicensed medicine. In addition to Δ-9 tetrahydroxycannabinol and cannabidiol,
Sativex® also contains other cannabis derived ingredients, including additional
cannabinoids and terpenes. The acceptance of VeregenTM as a botanical medicine in
the US opened the door to the testing there of cannabis-based Sativex®, and the
product was subsequently approved for testing as a treatment for cancer pain, under
the US Adopted Name (USAN) Nabiximols.
The decision to formulate a medicine based on these two cannabinoids was due to
several factors. Numerous in vitro and in vivo studies had shown that THC and CBD
exhibited high levels of pharmacological activity, although the mode of action of the two
differed markedly. (The activities of the two were compared and contrasted in a review
by McPartland and Russo (2001)). The aim of mixing the two cannabinoids was not
simply to draw additive benefit from the individual properties of the two molecules.
CBD was suspected of being able to attenuate the psychoactive effects of THC, which
were undesirable in a medicine (BMA, 1997). A growing weight of evidence also
showed that mixtures of THC and CBD offered improved efficacy over THC-alone.
Phytomedicines could be produced which contained both these cannabinoids, and
these would potentially contain other additional active natural ingredients, including the
terpenes and flavonoids (Musty, 2004).
Due to the predilections of smokers, drinkers, tea-totallers and coffee lovers the three
most commonly used legal drugs are nicotine, alcohol and caffeine (Robson, 1999). In
each case the flavour of the plant-derived source is part of the enjoyment of these
drugs. Cannabis is perhaps unique amongst illicit drugs in that, for some users, the
wide range of tastes produced by the various sources is a similarly important part of the
‘cannabis experience’ (Rosenthal, 2001). The ingredients having the greatest effects
on the cannabis taste would most probably be the fragrant terpenes within the essential
oils. Some of these have their own pharmacology and have been cited as likely
synergists in mixtures with cannabinoids (McPartland and Russo, 2001). The potential
benefit of these ingredients was demonstrated in a test measuring pain relief in mice, in
which unknown powerful synergists produced a 330% increase in activity compared to
THC alone (Fairbairn and Pickens, 1981). Synergistically improved efficacy of cannabis
extracts over THC-alone was also demonstrated in a mouse model which assessed
their antispacticity effects (Williamson, 2001). In subsequent research cannabis
16
extracts also showed a significantly increased antihyperalgesic effects compared to

CBD-alone when tested in a rat model of neuropathic pain (Comelli et al., 2008). The
potential benefits for mankind were supported by the observation that patients taking
synthetic derivative nabilone for neurogenic pain actually preferred cannabis herb and
reported that it relieved not only pain but the associated depression and anxiety
(Williamson and Evans, 2000). Reasons suggested included the more rapid absorption
through the lung than the gut; the presence of other ingredients in plant-derived
cannabis which might give additive or synergistic effects; and the ability of smokers to
self-titrate their dose (Grinspoon and Bakalar, 1995).
The decision to evaluate THC/CBD mixtures as a potential medicine was further

supported by the knowledge that the majority of cannabis used in the UK at the time –
at least for recreational purposes – was in the form of cannabis resin. It was suspected,
but not confirmed, that the majority of anecdotal reports for the efficacy of cannabis in
the UK would be based upon this material. Studies in the USA had shown that
cannabis resin (hashish) typically contained approximately equal quantities of both
cannabinoids, whereas herbal cannabis (marijuana) contained predominantly THC
(ElSohly et al, 1984).
A series of clinical trials was subsequently planned which would require the
propagation chemotypes dominant in cannabinoids other than THC and CBD. Some of
these are also discussed in this thesis.
1.8 Cannabinoid and terpene biosynthesis

The cannabinoids of greatest initial interest in this research, Δ-9 tetrahydrocannabinol
and cannabidiol are referred to as ‘pentyl’ cannabinoids, due the presence of a five-
carbon chain attached to the aromatic moiety (Figure 1.3). Other pentyl cannabinoids
frequently discussed in this thesis are cannabichromene (CBC) and cannabigerol
(CBG). All four have important propyl analogues. The first step in the synthesis of
pentyl cannabinoids is the condensation reaction of geranylpyrophosphate (GPP) with
olivetolic acid. The resulting product CBGA is the direct precursor for the three major
pentyl cannabinoids tetrahydrocannabinolic acid (THCA) (Fellermeier et al., 2001),
cannabidiolic acid (CBDA) (Taura et al. 1996) and cannabichromenic acid (CBCA)
(Gaoni and Mechoulam, 1966). The enzymes affecting the synthesis of these three
cannabinoids are named THC synthase, CBD synthase and CBC synthase
respectively. The first step in the synthesis of propyl cannabinoids is the condensation
reaction of geranylpyrophosphate (GPP) with diverinic acid. The resultant product
17
cannabigerovaric acid (CBGVA) is the direct precursor of the propyl analogues of the
three aforementioned cannabinoids ie THCVA, CBDVA and CBCVA.
When herbal cannabis is dried stored and heated, these cannabinoid acids
decarboxilize gradually or completely to the neutral forms (e.g. THCA → THC) (de
Meijer et al. 2003). Complete decarboxilation of the acid form of the cannabinoids
occurs during the analytical process, when using gas chromatography (Brown, 1998).
To varying extents, and according to storage conditions, these cannabinoids undergo
oxidative catabolism with the production of a range of additional cannabinoids. Some of
these (eg CBN) have their own reported pharmacological activity and in some cases
interact with other cannabinoids (Pertwee, 1998, Wilkinson et al, 2003). The principle
active cannabinoids, their precursors and catabolites are listed in Table 1.3.
Figure 1.3. Biosynthetic pathway of THC and THCV, via CBG or CBGV. 1,
Geranylpyrophosphate; 2, Divarinic Acid (R1) or Olivetolic Acid (R2); 3, Cannabigerovarin
(CBGV) (R1) or Cannabigerol (CBG) (R2);. 4, Δ9- tetrahydrocannabivarin (R1) or Δ9-
tetrahydrocannabinol (R2). R1 (-C3H7) and R2 (-C5H11) indicate the propyl or pentyl forms of
the metabolites; Enzyme I: geranylpyrophosphate:olivetolate geranyltransferase(GOT);
Enzyme II: THC(V) synthase.
18
Although the cannabinoids in fresh plant material exist in the acid form (THCA, CBDA
etc.,) it is common practice to simply refer to these cannabinoids in their neutral form
(THC, CBD etc.). In this thesis, that convention applies unless the acidity of the
cannabinoid needs to be highlighted.
Initial Principle Initial

Second Biosynthetic Decarboxylated
Biosynthetic Oxidative
Product Product
Product Catabolite
CBGA
Cannabigerolic CBG
_
Cannabigerol
Acid
THCA
THC CBN
Tetrahydrocannabinolic
Tetrahydrocannabinol Cannabinol
acid
CBDA CBD
CBS
Cannabidiolic Acid Cannabidiol
CBCA CBC CBL
Cannabichromenic Acid Cannabichromene Cannabicyclol
CBGVA
Cannabigero- CBGV
-
Cannabigerovarin
varic Acid
THCVA THCV
CBNV
Tetrahydrocannabivarinic Tetrahydro-
Cannabivarin
acid cannabivarin
CBDVA CBDV
CBSV
Cannabidivanic acid Cannabidivarin
CBCVA
CBCV
Cannabichromevarinic CBLV
Cannabichromevarin
acid
Table 1.3 The predominant cannabinoids found in Cannabis sativa and their main catabolites.
The ingredients in cannabis suspected of contributing to the synergistic

pharmacological effects include the main constituents of the essential oil – the
monoterpenes and sesquiterpenes. The precursor of cannabinoid biosynthesis GPP
also reacts with a range of other structures to produce the monoterpenes. These are a
very diverse chemical group, which includes cyclic and acyclic structures. They are
relatively volatile and the main source of essence within the essential oil. In addition to
the ‘true’ monoterpenes (C10H16) there are also a series of increasingly more oxidized
families within the group. These include the alcohols (e.g. linalool C10H18O), ethers (e.g.
1,8-cineole C10H18O), esters (e.g. bornyl acetate C12H20O2), aldehydes (e.g. citral
C10H16O) and ketones (e.g. pulegone C10H16O). All of these were amongst fifty eight
19
monoterpenes identified in cannabis in a detailed study by Turner et al. (1980), and

these were accompanied by thirty eight sesquiterpenes. The latter are reported to be
formed from a dominant precursor farnesyl pyrophosphate (FPP). Both FPP and GPP
are derived from isopentenyl pyrophosphate (IPP). The cannabinoids, sesquiterpenes
and monoterpenes therefore all share a common precursor in IPP. As with the
monoterpenes, the true sesquiterpenes (C15H24) are also accompanied by series of
more oxidized related structures. Higher molecular weight terpenes are also present,
the next group within the ‘terpenoid skeleton family’ being the diterpenes. These are
not volatile and not generally considered active constituents of an essential oil. The
only significant example in cannabis is phytol which is a precursor and catabolite of
chlorophyll and tocopherol Vitamin E, both of which are abundant in cannabis. Phytol
has been shown to exhibit antispasmodic activity (Pongprayoon et al., 1992) and to
elevate the neurotransmitter GABA levels in central nervous system (Bang et al., 2002).
It too, may contribute to the pharmacological properties of cannabis based medicines.
Two core pathways of IPP biosynthesis have been identified. The first of these, the
acetate/mevanolate pathway, operates within the cytosol (the fluid component of
cytoplasm) while the second and more recently discovered pyrovate/glyceraldehyde-3-
phosphate pathway takes place within the specialized membrane-bound subcellular
organelles called plastids (or more specifically leucoplasts) (Hallahan and Gray, 2000).
These pathways are outlined in Figure 1.4. Terpene biosynthesis has been reviewed in
detail by Croteau and Johnson (1984) and is not repeated here.
Glucose Glucose
Glyceraldehyde 3-Phosphate (GAP) Triosephosphate
Pyrovate Acetyl-CoA
GAP
CO2 3-hydroxy-3-methylgluterate-CoA
1-Deoxyzylulose-5-P
Mevalonate
Isopentyl Diphosphate (IPP) Isopentyl Diphosphate (IPP)
PLASTID CYTOSOL
Figure 1.4. The two pathways of isopentyl diphosphate (IPP) biosynthesis in plants, as found in
the plastid and cytosol respectively.
20
1.9 Cannabinoid Receptors and Cannabinoid Pharmacology.

The understanding of the pharmacodynamics of the cannabinoids was greatly
enhanced when specific cannabinoid receptors were identified in mammalian brain
(Devane et al., 1988) and subsequently cloned (Matsuda et al., 1990). Named CB1
receptors (Appendix 3a), these were identified in the central nervous system (CNS)
and certain peripheral tissues. These were identified as belonging to a so-called
superfamily of G-protein-linked receptors (Matsuda et al., 1990). G-protein-linked
receptors characteristically exhibit a seven-folded structure which spans the cell’s
plasma membrane. In some such receptors the ligand attaches to a domain (a section
of the polypeptide chain) that is entirely exposed on the outer surface of the cell.
Others ligands will have domains which include parts of the chain deep within the cell
membrane. The binding of a ligand signals any one of several activities within the cell.
Over half of all known drugs work through G-protein receptors (Alberts et al., 2002).
These receptors include three protein sub-units – α, β and γ. In the unstimulated state
of the so-called ‘two state model of constitutive activity’ no ligand is bound to the
receptor ‘R’ and guanosine diphosphate (GDP) is bound to the α-subunit. This
conformation the G protein is termed RGGDP. Upon stimulation, this sub-unit releases
its GDP allowing Guanosine 5’- triphosphate (GTP) to bind in its place (Figure 1.5).
This conformation of the G protein is termed R*GGTP (Ross, 2007). As a consequence,
the αβγ complex dissociates into an α subunit and a βγ complex. The now-free α
subunit is now able to change its shape and interact with target proteins. The α
subunits have intrinsic slow hydrolase activity, and as a result the GTP spontaneously
hydrolyses to GDP with the release of one phosphate moiety. In doing so, the G-
protein resets itself back to the rest position. Within the cell this ‘two state’ model of
constitutive activity suggests that an equilibrium exists: -
RGGDP ↔ R*GGTP.
The balance of these two forms signals a series of separate reactions within the cell.
These include alterations to the cyclic adenosine 5’- monophosphate (cAMP)
concentration and to the activity of potassium and calcium channels.
Δ-9 THC interacts with the CB1 receptors as a ligand, causing psychoactive effects
(BMA, 1997). A second type, the so called CB2 receptor (Appendix 3b), was
subsequently discovered in spleen macrophages and found not to be present in the
CNS (BMA, 1997; Pertwee, 1997). The natural function of the cannabinoid was more
easily explained with the discovery of endogenous substances within mammalian
tissue which interacted with these receptors. These have been termed as
21
endocannabinoids to differentiate them from the plant-derived cannabinoids, which are

sometimes called phytocannabinoids (Pate, 1994).
Figure 1.5. The disassembly of an activated G-protein into two signalling components. (Alberts
et al., 2002)
The first of the endocannabinoids was discovered to be the arachidonic acid derivative
arachidonylethanolamide in 1992, and subsequently named anandamide (Devane et al.
1992). In addition to the phytocannabinoids and endocannabinoids a number of
synthetic cannabinoid ligands have also been manufactured. These include chemicals
closely related to the phytocannabinoids such as the Δ-9 THC analogue nabilone
(Cesamet) and the nonclassical cannabinoid group which includes the compound
CP 50556 (levonantradol). The pharmacological activity of many of these cannabinoids
has been assessed in vitro and in vivo. In the former case, this has included an
assessment of the test cannabinoid’s ability to block or reverse the effects of another
22
standard receptor antagonist (Pertwee, 1997). This method has been used to assess
extracts from many of the chemotype studied for this thesis.
The interaction of phytocannabinoids or endocannabininoids with cannabinoid

receptors is illustrated schematically in Figure 1.6. In addition to being present in the
central nervous system and throughout the brain, CB1 receptors are also found in the
immune cells and the gastrointestinal, reproductive, adrenal, heart, lung and bladder
tissues. By altering the cell activity within such a wide range of tissues, several
prominent pharmacological responses result from the agonism of CB1 receptors,
including both the control of nociceptive and neuropathic pain. CB2 receptors are
thought to have immunomodulatory effects and to regulate cytokine activity and
thereby altering cell-to-cell communication. In combination with CB1 receptors, they
show pain control effects (Pertwee, 2004).
Figure 1.6. The interaction of phytocannabinoids and/or phytocannabinoids with the CB1
receptor in the eukaryote cell and the consequent effects on cAMP and ERK activity and ion
channelling. Modified from Ross (2007).
While many of the pharmacological properties of cannabis can be explained by

interactions of its ingredients with the CB1 and CB2 receptors, not all of the effects can
be explained this way. Cannabidiol has shown significant levels of antipsychotic
activity, which are seen as advantageous in medicines containing Δ-9 THC, but this
cannot be totally explained by interactions with the cannabinoid receptors as CBD is a
weak ligand (Iversen, 2008). Both CBD and Δ-9 THC have been shown to have equally
potent neuroprotective antioxidant properties in rat cortical neuron cultures, where
23
glutamate otherwise reached toxic levels (Hampson et al., 1998). Indeed, the
antioxidant abilities perhaps indicate a natural function of these cannabinoids in plant
tissues.
In 1998 an important trial was performed by the Royal Pharmaceutical Society of Great
Britain, in collaboration with the UK Medical Research Council. This CAMS study
(cannabis in MS) involved 630 patients and explored the effects of synthetic THC
(Marinol) and a cannabis extract “Cannador” given orally on spasticity and other
symptoms related to multiple sclerosis (Zajicek et al., 2003). The results of the study
were mixed, and a large placebo effect was noted, but both active treatments
demonstrated significant improvements in subjective measures of spasticity, muscle
spasms, pain and sleep, and also in an objective measure of mobility. No effect was
apparent on irritability, depression, tiredness, tremor or loss of energy. The authors
noted an unexpected reduction in hospital admissions for relapse in the two active
treatment groups. The known interaction of cannabinoids with the immune system, and
the fact that MS was still regarded as an auto-immune condition, led them to comment
that this finding was worthy of further investigation.
A highly significant recent finding was the observation that β-caryophyllene, a major
sesquiterpene in cannabis, selectively binds to the CB2 receptor and is a CB2 receptor
agonist with anti-inflammatory activity in vivo (Gertsch et al., 2008).
1.10 Aims and Outline of Thesis

Supported by the GW Pharmaceuticals plc, this thesis was performed to improve the
reliable production of phytopharmaceutical feedstock for the production of botanical
medicines from Cannabis sativa L. As the title of this thesis suggests, the challenging
research discussed here was multifaceted and covered a broad range of scientific
disciplines. The majority of the work was divided between a pharmaceutical research
laboratory and a glasshouse, but it also necessitated much involvement with the
agricultural industry and law enforcement agencies. The research had five broad aims,
as covered in Chapters Two to Six.
The first botanical medicine from GW Pharmaceuticals to be widely tested in clinical

trials contained the two cannabinoids THC and CBD in approximately equal quantities.
A significant reason for choosing an equal mixture of these two cannabinoids was the
supposition that this balance existed in most of the illicit cannabis in the UK. Chapter
Two describes research that investigates what cannabinoid profile actually existed in
illicit material. For many patients in the UK, such material was the only form available.
Indeed, the anecdotal evidence that many multiple sclerosis patients were finding this
24
illicit material to be efficacious was a major consideration in allowing GW

Pharmaceuticals a license to evaluate cannabis based medicines. Despite this, little
was known of its cannabinoid profile. As stated at the 1998 House of Lords Select
Committee on Science and Technology enquiry into Cannabis, the action of herbal
cannabis needed defining (Wall, 1998). The suspected 1:1 THC:CBD ratio in UK
cannabis was based on research into the cannabinoid profile of illicit material in the
USA. No such large scale research had been performed here in the UK. Seeking in
part to check if a one to one ratio of the two cannabinoids was indeed the typical
cannabinoid profile of ‘street cannabis’, Chapter Two describes work to evaluate the
cannabinoid content of this material in what was the first such large scale UK study. In
addition to supporting the business aims of GW Pharmaceuticals plc, the research also
provided an insight into the apparently escalating potency of UK cannabis, which was
being increasingly linked with psychotic disorders amongst Britain’s youth. This data
was eagerly welcomed by the scientific, medical and law enforcement communities.
The co-ordination of the collection of the illicit samples from a number of constabularies
necessitated much police time.
The cannabinoids and most terpenes in cannabis are synthesised in small structures
called glandular trichomes. A pharmacognosist could reasonably argue that they are
the most important part of the plant. Chapter Three looks in depth at the differing
structures of Cannabis trichomes, with the aim of improving the knowledge of their form
and function. Although not all types formed cannabinoids, the inactive non-glandular
types would frequently occur as contaminants when collecting the glandular types for
research purposes, and they needed to be routinely identified. The aim of the work in
Chapter Four was to compare and contrast the different glandular trichome forms and
examine their secondary metabolite profiles. The effect of the state of trichome maturity
on the cannabinoid content and profiles was also investigated. Techniques are
evaluated in which trichomes are separated so as to exert some control over the
cannabinoid profile within a phytopharmaceutical feedstock.
From a purely botanical perspective it was essential to gain a knowledge of how the
plant develops. From a horticultural perspective it was imperative to learn how to
propagate plants that were healthy and high yielding. This latter aim is shared by
growers of illicit cannabis, but the grower for the pharmaceutical industry has the
additional aim of learning how to repeatedly propagate crops that were uniform in
secondary metabolite content. Chapter Five describes the research to address these
aims in an indoor environment. Chapter Six addresses the same aims but in an
agricultural setting. Although the research is reported in sequence in Chapters Two to
25
Six, the differing areas of study were performed in parallel. Growing a crop through to
harvest takes time, especially in an outdoor environment. The horticultural and
agricultural investigations took several years to complete.
Addressing a seminal cannabis symposium in 1969, organised by the Institute for the
Study of Drug Dependence, the internationally renowned expert Dr. R.E. Schultes of
Harvard University stated that “A thorough understanding of Cannabis sativa L as a
plant must be basic to progress in studies of its derivatives and their significance to
man and their effects on life and social evolution” (Schultes, 1970). The author of this
thesis very much shares that view and this motivated and influenced the studies
reported here.
26
Chapter 2 Characterisation of Illicit Cannabis in the UK
2.1 INTRODUCTION
It is often stated that a drug is a substance with the ability to interact with the
metabolism of an animal (usually man). A medicine however is a beneficial drug
formulated in such a way as to optimise its absorption and performance. The cannabis-
based medicine Sativex® is a medicine formulated as a sub-lingual spray. In ancient
civilisations the medicinal benefits of cannabis are achieved by smoking the dried plant
material or resin. Much of the ‘medicinal cannabis’ used around the world is supplied
as an unformulated drug. To achieve the desired effect, this material would normally be
smoked, but it could sometimes be vaporised or ingested.
Smoking cannabis for recreational or medicinal reasons in the UK was almost unknown
until the 1950s, but recreational cannabis smoking had become common a decade
later (Robson, 1999a). Most medicinal users of illicit cannabis would have also smoked
the material, although some would have ingested it in cooked form. Organized
production of foodstuffs containing cannabis for alleged medicinal use resulted in
several criminal convictions in the UK in recent years. Although thought to be
widespread, until now the true extent and pattern of medicinal use of illicit cannabis in
the UK was not known. A number of surveys have been performed to gain a better
knowledge of this activity. Consroe et al. (1997) asked people with MS to describe
which of their symptoms were relieved by smoking cannabis. Spasticity and muscle
pain were reported to show the greatest improvement (97% and 95% of responders).
The published results did not report what form of cannabis was typically used.
In what was thought to be the most extensive survey of illicit cannabis use for medicinal
use among chronically ill patients, Ware et al. (2005) reported the effects of cannabis in
2969 candidates. The patients were not selected randomly, or by any systematic
procedure, and the study is therefore skewed towards highly motivated responders.
82% of users smoked the cannabis and 43% reported eating it. The survey invited
respondents’ to report their preference for cannabis resin or herbal cannabis (including
sinsemilla). The replies were not included in the published report. However, a random
sub-sample of 500 unnamed reply forms was acquired from the authors. Just 72 of the
500 responders had answered this question. Of these, 50 (69%) expressed a
preference for ‘herbal cannabis’, 16 preferred resin (22%) and 5 (7%) were satisfied
with both. One respondent reported a preference for ‘cannabis oil’. The term ‘herbal
cannabis’ in this context was used to collectively categorize materials described by
27
respondents as ‘herbal’, ‘skunk’, ‘bud‘, ‘leaf’, ‘marijuana’ and similar terms. It was plain
from these replies that several different forms of cannabis were being used. It can be
assumed that these would have varied greatly in cannabinoid content and profile.
Without further research it could not be fully explained why some users preferred one
form of cannabis over another. The British Medical Association publication Therapeutic
uses of cannabis (1997) listed the pharmacological properties of the phytocannabinoids
and recognised that the illicit cannabis circulating in the UK was a very inconsistent
product and its THC content varied widely. This was confirmed by King et al. (2004) in
the most detailed study of UK cannabis potency to date. However, the content of CBD
and other cannabinoids in illicit UK cannabis remained little studied.
In recent years Cannabis has been by far the most commonly used illicit recreational
drug in the UK (Szendrei, 1997). In the early 1970s, resin consistently accounted for
about 70% of cannabis seizures in England and Wales but from 1997 to 2002 the
proportion fell yearly from 72% to 53% (Mwenda et al., 2005). By the following year,
resin appeared to have been surpassed by locally grown cannabis as the main form of
used (Hough et al., 2003). This change would be expected to affect the balance of
THC, CBD and other cannabinoids in those cannabis materials circulating. In addition
to having significant implications for the pharmacological properties of street cannabis,
changes in CBD also had the potential to affect its psychoactive potential.
It was highlighted by Smith (2005) that future studies of potency in the UK cannabis
should include an assessment of CBD as well as THC levels. This was reiterated by
the United Nations Office on Drugs and Crime in the World Drug Report 2006
(UNODC, 2006). Both publications stated that CBD was a cannabinoid with
antipsychotic activity, having the potential to alter the potential harm attributable to the
recreational use of THC. Neither publication however was concerned how CBD might
affect the pharmacological properties of illicit cannabis. The term ‘potency’ in this
context refers entirely to the concentration of cannabinoid in a specified sample. This
meaning of potency is widely accepted in forensic science (e.g. ElSohly et al 1984).
However, in pharmacology potency is a measure of drug activity expressed in terms of
the amount required to produce an effect of given intensity (Page et al 2006). In this
chapter potency is defined as a concentration of cannabinoid.
In the absence of a prescribable efficacious cannabis-based medicine in the UK, illicit

cannabis continues to be used widely for medicinal purposes. However, the
cannabinoid content and pattern of use of this material has not been comprehensively
studied.
28
2.2 AIM and OBJECTIVES

The aim of this part of the study was to gain an understanding of the cannabinoid
content and variability of illicit cannabis circulating in England. Within this study there
were a number of objectives.
2.2.1. The study would initially assess how the market was split between cannabis
resin, herbal cannabis and sinsemilla.
2.2.2. The cannabinoid profile of these different types would be measured and
compared, to gain an understanding of the potential efficacy and safety of illicit
cannabis – when used for medicinal or recreational purposes.
2.2.3. The potency of samples would be compared to previous data, generated by the
Forensic Science Service. The existence of any trends in THC content would
thus be identified.
2.3 MATERIALS
2.3.1 Cannabis samples

The samples analysed from this study had been seized by police in Derbyshire, Kent,
London Metropolitan (SE1 area), Merseyside and Sussex between 2004 and 2005.
2.3.2 Microscopy, Photography and other Apparatus

The microscopes and associated apparatus for this study are shown in Table 2.1.
Apparatus Source
Brunel Microscopes
Photonic PL2000 - double arm cold light source.
Unit 12 Enterprise Centre,
MX3 Low Power Light Microscope
Bumpers Industrial Estate
Bumpers Way, Chippenham,
Wiltshire. SN14 6QA
High Power Stereo Light Microscope with STE UK Ltd., Staplehurst Rd

Trinocular Head for camera attachment. Sittingbourne, ME10 2NH
Colour Charts RHS Enterprises Ltd, RHS Garden,

Wisley, Woking, Surrey, GU23 6QB
Table 2.1. Photographic, microscopy and other miscellaneous items and commercial sources.
29
2.4 METHODS
2.4.1 Collection of Representative Samples

The collection of samples was coordinated with the assistance of the South East
Government Home Office. Authorization was gained from the Chief Constables of
Derbyshire, Kent, London Metropolitan, Merseyside and Sussex. This spread of
constabularies would capture populations of differing socio-economic conditions and
varying influences of local ports. To obtain a pool of samples which most accurately
represented the illicit material circulating amongst cannabis users, police were
requested to forward materials that had been seized within the last year, during street
arrests or while raiding the property of minor drug suppliers. To help prevent any
skewing of the data towards imported material, the study excluded large seizures from
those arrested for smuggling cannabis into the country. Such materials would have
likely entered the cannabis market higher up the supply chain, and not been offered
directly to individuals purchasing cannabis for their own use.
2.4.2 Storage of illicit cannabis samples

Upon arrival, samples were individually numbered and stored in darkness at ambient
temperature in 30-35% RH. All were analysed within 28 days of arrival.
2.4.3 Categorisation of the form of each sample

The materials were assessed visually, using a Brunel MX3 low power light microscope
where necessary, and the form of the cannabis sample was established. Four
categories were identified:
2.4.3.1 Cannabis resin
Figure 2.1. Examples of cannabis resin samples (<1g up to 230g) seized by police in 2004/5.
30
Cannabis resin consists of the glandular trichomes and other fine particles collected
from the inflorescences and upper leaves. The material is compressed into hard
blocks prior to importation (Raman, 1998). All samples were dark brown in colour
(Figure 2.1). These varied in shape and size and, when present in sufficient quantity,
generally had a light characteristic odour.
2.4.3.2 Herbal cannabis
As adopted by King et al. (2004), the term ‘herbal cannabis’ was used only to include
imported dried plant material collected from outdoor grown plants. The material was
light to dark brown in colour. The glandular trichomes were always brown, due to
ageing (Mahlberg et al., 1984). Seeds were frequently present. The material was
sometimes in loose form (Figure 2.2a), but was also frequently encountered in hard
blocks (Figure 2.2b) where it had been compressed to reduce volume during
importation. The material had a light fragrant odour. Fungal mycelium was
occasionally visible, suggesting that decay had occurred at some point during
importation or storage.
(a) (b)
Figure 2.2. (a) Loose herbal cannabis material showing separated seeds; (b) Compressed herbal
cannabis material with selection of removed seeds.
2.4.3.3 Sinsemilla
The sinsemilla form of cannabis was light green or grey-green in colour. The material
consisted of resinous female floral material only. Close examination often revealed
where bracts and leaves had been physically removed. Large intact sections of
inflorescence, up to several grams in weight, were sometimes present. More
commonly, the material had been broken into smaller pieces and packaged into small
packets for marketing (Figure 2.3). Seeds were always absent as a result of the all-
female crops being grown without exposure to pollen. The glandular trichome colour
varied between crystal clear, white and light brown. The odour was clearly stronger
31
than that of resin and herbal cannabis. There was no visible sign of fungal
deterioration. This pungent, light material was generally regarded as having been
grown in the UK, but some could have entered the UK from mainland Europe.
Figure 2.3. A typical confiscated sample of illicit sinsemilla cannabis consisting of three
separate packets, each containing approximately one gram.
2.4.3.4 Cannabis powder
Some herbal cannabis and sinsemilla samples were recovered from portable cannabis
grinders. These are used to break herbal cannabis and sinsemilla into a suitably fine
texture for smoking. More complex grinders included a fine metal mesh within the
construction, as shown in Figure 2.4b.
Interlocking
grinding
components
Sieve
Lid
Base and
collected
powder
Resin
(a) (b)
Figure 2.4. (a) A herb grinder in closed position; (b) An open herb grinder revealing the
component parts.
In this example dry cannabis would be placed between the top two interlocking
sections. As these were contra-rotated by hand, sharp projections on the face of these
sections would abrade the cannabis and break it into small portions. These would fall
32
through 4 mm diameter holes in the top right hand section onto the sieve section
below. Any dislodged glandular resin heads would fall through this mesh into the base.
Glandular trichomes dislodged from the plant during grinding could fall through this
sieve and be collected in a separate chamber within the device. One grinder was found
with approximately 1 cm-3 of separated yellow powder. This consisted almost entirely
of glandular trichomes. Some of these have become trapped and crushed between
walls of the interlocking sections forming a ring of black resin.
2.4.3.5 Other categories not included
Previous studies on cannabis in the UK described a hash oil. This preparation is made
by dissolving, and subsequently concentrating, cannabis extracts in an organic solvent
(Barber et al., 1996, Hough et al., 2003). No such samples were identified during this
study. Many samples were seized which consisted of a mixture of cannabis and
tobacco; all were excluded. Three seized suspected-cannabis samples were also
analysed and found to be plant material other than cannabis or tobacco.
2.4.4 Measurement of cannabinoid potency and profile

Where detectable the THC, CBD, CBC, CBN, CBG and THCV content of each sample
was measured, using gas chromatography, as described in Appendix 1.
2.4.5 Statistical Analysis

Analyses of variance (ANOVA), regression and F-tests were used as appropriate,
utilising Microsoft Excel 2003 related software. Kolmogorov-Smirnov and Wilcoxon
Rank Sum tests and Hodges-Lehman Estimates were performed using SAS software,
with the assistance of colleagues in the Statistical Analysis Department, GW
Pharmaceuticals Ltd.
33
2.5 RESULTS and DISCUSSION
2.5.1 Categorisation of cannabis type between regions

Four hundred and sixty samples were analysed in this study. The proportion of
sinsemilla, herbal cannabis or resin in each area is shown in Table 2.1. Sinsemilla was
the most common form found overall, accounting for 55% of the samples seized.
However, differences were found between regions. In Kent, resin accounted for 85 of
the 146 samples (59%), possibly due to importation through the major ports in this
region. Herbal cannabis is presently the most common form of cannabis in the USA
(ElSohly et al., 2000). In contrast, this study revealed this type to be the least common
in England. Little or no herbal cannabis was identified in four of the regions, but it did
account for 30 of the 159 samples (19%) seized in the South East London Metropolitan
area.
Constabulary Sinsemilla Herbal Resin

Derbyshire 15 1 28
Kent 61 0 85
London Metropolitan 97 30 32
Merseyside 49 1 9*
Sussex 34 3 15
Total 256 35 169
Table 2.2. Number of each type of sample received from each constabulary. (In addition, one
sample of cannabis powder was received from Kent). *The low number of resin samples from
Merseyside was due to the late inclusion of such samples from this constabulary.
2.5.2 The range of cannabinoids in each cannabis category

The range of cannabinoids (the cannabinoid profile) in each type of cannabis is shown
in Table 2.2. The distribution of potencies of sinsemilla, herbal cannabis and resin
samples were all non-parametric and the median is therefore shown as a more
appropriate average than the arithmetic mean. The potency of resin samples varied
widely from almost 0% THC up to nearly 11% (Table 2.2). The majority of the samples
were at the weaker end of this range. 40% had a THC content of <2% THC, and more
than 80% had <6% THC. The range of potencies in herbal cannabis was similar. The
maximum THC content found was nearly 12% but approximately 90% had a content of
<6%.
34
Sinsemilla potency ranged from about 1% to 23%, the majority being toward the high
end of this range. The cannabis powder, retrieved from the herb grinder, was the most
potent of the samples analyzed (40.6% THC, Table 2.2). This material had been
prepared using a very simple piece of equipment, and this illustrates that extremely
potent cannabis preparations are readily available.
Type THC CBD CBC THCV CBG CBN

Sinsemilla Median 13.98 <0.10 0.20 <0.03 0.41 0.16
(n = 256) Minimum 1.15 <0.10 <0.10 <0.10 <0.10 <0.10
Maximum 23.17 0.56 1.41 2.74 2.16 2.98
Herbal Median 2.14 <0.10 0.22 0.17 0.21 0.55
(n = 35) Minimum 0.28 <0.10 <0.10 <0.10 <0.10 <0.10
Maximum 11.81 1.97 0.42 0.43 0.76 3.62
Resin Median 3.54 4.17 0.34 0.10 0.29 1.55
(n = 169) Minimum 0.44 0.36 <0.10 <0.10 <0.10 0.38
Maximum 10.76 6.97 0.66 0.29 1.05 4.30
Powder
40.63 0.18 0.41 0.29 1.59 0.57
(n = 1)
Table 2.3. The median and the range of potencies of five cannabinoids (% w/w) in resin, herbal
cannabis, sinsemilla and cannabis powder, seized in five constabularies in England in 2004/5.
In sinsemilla samples, THC typically accounted for approximately 94% of the total
detectable cannabinoid content. The THCV, CBG and CBN content of individual
samples occasionally exceeded 2% w/w, but this was approximately one tenth of the
maximum potency recorded for THC. The CBD content of sinsemilla was typically very
low and fell below detectable levels (0.1%) in the majority of samples. The lack of the
THC catabolite CBN, suggested that samples were comparatively fresh when seized,
and had remained in good condition in police stores.
THC was also the dominant cannabinoid in herbal cannabis and CBD levels were
similarly mostly below the detectable threshold (0.1%). CBN levels were much higher
in herbal cannabis than in sinsemilla. The ratio of THC and CBN in these samples
varied greatly. This was at least partly due to the varying lengths of time that herbal
cannabis encounters on its route to the UK, the majority coming overseas from South
Africa (UNODC, 2006). A long transport period would favour the breakdown of THC to
the catabolite CBN (Ross et al., 1997). As a result of THC catabolism, and CBD, CBG,
35
CBC being more stable, THC only accounted the 65% of the median total cannabinoid
content of herbal cannabis.
The balance of THC and CBD in the different forms of cannabis was clearly affected by
their contrasting genetics. Research suggests that the production of THC or CBD, from
the common precursor CBG, is closely controlled by two co-dominant alleles at a single
locus (de Meijer et al., 2003). As a result, cannabis plants can be identified as
belonging to any one of three chemotypes; i.e. THC dominant, CBD dominant or an
approximately equal mixture of the two (de Meijer et al., 2003, Small et al., 1973a).
Sinsemilla appeared to be entirely derived from the THC dominant chemotype. Two of
the herbal cannabis samples had a substantial CBD content due to the presence of the
Bd gene. The resultant effect of these contrasting genetics is illustrated in Figures 2.4a
to 2.4c.
Cannabis resin had a very different cannabinoid profile to that of herbal cannabis and
sinsemilla. The majority of the cannabis resin would appear to be prepared from
landrace populations of plants which contain all three chemotypes. CBD appeared
slightly more dominant (mean content 4.3%) than THC (mean content 3.5%) in this
material. CBN was present in much higher quantities than in herbal cannabis or
sinsemilla. Cannabis resin samples showed very variable contents of THC (0.4 –
10.8 %) and CBD (0.4 – 7.0%). Compared to sinsemilla, the THC content of resin was
significantly much more variable (two-sided f-test, p < 0.001). The ratio of these
cannabinoids within individual samples also varied widely, as shown in Figure 2.4c.
36
% CBN % CBD
% THC
(a)
% CBN % CBD
% THC
(b)
% CBN % CBD
% THC
(c)
Figure 2.5. (a) The balance of THC, CBD and CBN in sinsemilla (n = 256); (b) The balance of
THC, CBD and CBN in herbal cannabis (n = 35); (c) The balance of THC, CBD and CBN resin
(n = 169).
37
It was hypothesized that much of the variation in THC:CBD ratios within the resin
samples was due to THC being a less stable product than CBD. Indeed, in Figure 2.4c
resin samples exhibited a wide range of ratios between THC and its main catabolite
CBN. To test the hypothesis, an estimate was made of the original THC content of
each sample before catabolism to CBN had commenced. A simple arithmetic
calculation formula was not available because the breakdown of THC to CBN is not
quantitative (Phillips, 1998). Studies with herbal cannabis (Ross and ElSohly, 1999)
and resin (Martone and Della Casa, 1990) suggested that a concentration of one mole
of CBN implied the original presence of four to six moles of THC. To estimate the likely
correlation between the CBN content and original THC content in these samples, the
relative contents of the two cannabinoids were plotted (Figure 2.5).
9
8
7 y = -2.2401x + 7.3255
Percent THC
6
5
4
3
2
1
0
0 0.5 1 1.5 2 2.5 3 3.5
Percent CBN
Figure 2.6 The correlation between THC and CBN content in resin samples seized in five
constabularies in 2004/5 (n = 169).
The coefficient value of -2.24 suggested that the extrapolated original median THC
content of these samples was 7.3% THC, compared to the actual median THC content
of 3.5% at the time of analysis. Previous studies have suggested that the half-life of
CBD in resin is approximately three times that of THC (Martone and Della Casa, 1990).
As a result, in the plants used to produce this resin, the original CBD content was
estimated to have been approximately 5.5% compared to 4.2% at seizure. Figure 2.4c
also shows that of the one hundred and sixty nine resin samples, four were almost
devoid of CBD. These were therefore made from the THC dominant chemovar and
probably from a very different source.
2.5.3 Comparison of cannabis potency and profile between regions

Sinsemilla potency ranged from about 1% to 23%, the majority being toward the high
end of this range. There were small but statistically significant differences in mean THC
38
content of sinsemilla seized in different areas. The highest mean THC content was
found in the Derbyshire region (16.3%). This material was significantly more potent
(p < 0.05) than that seized in London (mean 12.9%). The remaining counties returned
mean sinsemilla potency values between these two extremes, the differences in
potency not being statistically significant (p > 0.05). Due to the small number or
absence of herbal cannabis samples in most regions, a meaningful comparison of
herbal cannabis potency levels between regions was not possible.
There were proportionally larger differences in the mean potencies of resin between
regions. Resin seized in Sussex (6.6%) and Derbyshire (5.4%) had significantly higher
mean THC contents (p < 0.05, ANOVA) than those seized in Kent (4.2%), London
(3.6%) or Merseyside (2.8%). Conversely, Sussex resin was notable for having a
significantly lower mean CBN content than that from the other counties (p < 0.05). The
mean CBN content of the Derbyshire resin was also significantly lower than that from
London and Merseyside. These data suggest that the Sussex and Derbyshire samples
were less aged.
2.5.4 Trends in Cannabis Potency

The mean THC content of imported herbal cannabis in this study (3.1%) was lower
than the 4.1% observed by the UK Forensic Service in 1998 (King et al., 2004). Little
inference can be drawn regarding potency trends from this because of the small
number of samples in both studies, the lack of a reported median in 1998 and the
highly non-parametric data (Figure 2.7). However, the potency ranges and distribution
patterns are broadly similar, and no major trend is apparent.
45
40
35
% of samples
30
25 1998
20 2005
15
10
5
0
<2 2-4 4-6 6-8 8 - 10 10 - 12 >12
%THC
Figure 2.7. A comparison of the range and distribution patterns of THC content of seized
imported herbal cannabis samples in 1998, King et al. (2004) (n = 44) and 2005 Potter et al.
(2008) (n = 33).
39
The mean THC content of resin (3.7%) was typical of that found previously, where
potency levels varied from approximately 3 to 6% THC between 1998 and 2002 (King
et al., 2004). The mean sinsemilla potency value of 13.3% (sd 4.21%) is higher than
that reported over the period 1995 to 2002 when the Forensic Science Service reported
that THC content of seized cannabis rose annually from approximately 6.0 to 12.5%
(King et al., 2004). This supports the belief that sinsemilla potency in the UK is
potentially increasing but, due to the large standard deviation in the 2005 data, and the
lack of reported detail in earlier data, the 2002-2005 rise cannot claimed to be
significant.
In both this study and in those analyses performed by the Forensic Science Service
from 1996-98, the THC content of samples ranged from <2% THC to >20%. The range
of sinsemilla potency levels observed in each study is compared directly in Figure 2.8.
To assess the significance of this change in distribution of sinsemilla potencies, the
data was analyzed using the Kolmogorov-Smirnov test for different distributions. This
showed that the two distribution curves were significantly different (p < 0.001). A
Wilcoxon Rank Sum test showed that the potency levels in 2005 were significantly
higher than those in 1996/8 (p < 0.001). A Hodges-Lehman Estimate of median
difference suggests an increase of 4.86% THC in the median potency between the
1996/8 data and the 2005 data (C.I. = 3.77, 5.54) (Potter et al., 2008).
25
20
% of Samples
15
1996/8
10 2005
-2 -4 -6 -8 0 2 4 6 8 0 2 4
0 2 4 6 -1 -1 -1 -1 -1 -2 -2 -2
8 10 12 1 4 16 18 20 22
% THC
Figure 2.8. A comparison of the ranges of THC contents of Sinsemilla seized in the UK and
analysed by the Forensic Science Service in 1996-8 (n = 145) and samples seized by police in
Derbyshire (n=15), Kent (n=58), London Metropolitan (n = 96), Merseyside (n = 44) and
Sussex (n = 34) in 2004/5 (total n = 247).
40
Large increases in cannabis potency achieved in the 1970s were largely attributed to
the achievements of cannabis breeders (Clarke, 2001). Many named cultivars
produced in the 1970s and 1980s are still widely marketed. Seeds of new cannabis
cultivars are continually produced in large numbers (Snoijer, 2002) but these do not
appear to produce plants of significantly higher THC content (confidential GW
Pharmaceutical data). However, cannabis seed production is unregulated and there is
no guarantee that seeds currently marketed under established variety names are truly
identical to those circulating thirty years previously. The rise in reported potency is
more likely due to increasing expertise amongst the illicit UK growers in recent years,
more of whom are able to push THC levels closer to the possible maximum. In the UK
during this period there has been a large increase in the number of retail outlets selling
cannabis seeds and sophisticated growing equipment. Many books, videos and DVDs
have been produced, advising growers how to maximise cannabis potency. The
internet has facilitated on-line purchasing of these items and has also provided easier
access to advice on cannabis growing and processing through expert web pages and
focused ‘chat-rooms’. During this period the UK Government and Police express the
opinion that the production and dealing of cannabis have not always been targeted
sufficiently vigorously (Clarke, 2006).
A recognised weakness in this study is that all the samples tested were seized by
police from users or dealers on the street. It is not known how representative this was
of the cannabis being consumed across the whole population. In ten years personal
experience as a magistrate, serving both Adult and Youth Courts, a vast majority of
offenders charged with using cannabis were seen to be from Social Grades C2 to F, as
defined by the NRS Survey (Appendix 5), and a large proportion were young. A
skewing in the average age of charged offenders arose because police were instructed
to charge offenders for cannabis use if they were less than eighteen. In contrast, they
were allowed some discretion and could issue a warning to adults. Of three million
estimated cannabis users in England and Wales, one million were over thirty years old
but much less likely to be charged (May et al., 2002). Research suggests that the
decision to charge an offender, rather than issue a warning, depended partly upon the
attitude of the offender (May et al., 2007). The more mature, circumspect and
sophisticated user is less likely to attract police attention. The cannabis used by this
type of user may differ in provenance, potency and price. During informal
conversations at Multiple Sclerosis Support Group meetings, attendees frequently
revealed that they grew their own cannabis to ensure supply, to remove contact with
drug suppliers and to reduce costs. Hough et al. (2003) reported the same observation.
41
This avoidance tactic might have significantly reduced the chances a ‘medicinal’
cannabis sample being seized.
2.5.5 The efficacy of illicit cannabis

The three main forms of illicit cannabis circulating in England in 2005 were very
variable in both their potency and their cannabinoid profile. Herbal cannabis showed
the greatest variability in THC content, this being significantly more variable than resin
(two-sided F-test p < 0.05). Both were significantly more variable in THC content than
sinsemilla (two-sided F-test, resin v sinsemilla, p < 0.001). Sinsemilla generally
contained high concentrations of THC and relatively few other cannabinoids. Herbal
cannabis was dominated by THC and CBN, the ratio of these two varying greatly
between samples. Resin contained substantial quantities of three cannabinoids –
THC, CBD and CBN – and the ratio of these varied greatly between samples. THC and
CBD have been demonstrated to act together synergistically in mammalian systems
and small differences in relative content of each could have proportionally greater
effects on the user (Williamson, 2001).
From its appearance, sinsemilla gave the greatest hint of its cannabinoid content.
Although the range of THC levels in sinsemilla ranged from < 2% up to > 22%, 85% of
samples had a THC content of between 8% and 24% THC (Figure 2.7), with the
majority being within a closer margin of 12% to 18%. It is possible that the lowest
potency material was also poor in appearance. Experienced users may thus have had
an increased likelihood of being able to judge when a sample’s THC content fell within
this 12–18% band. This would have enabled a more accurate judgment to be made of
the likely potency of the material. Being more variable in their cannabinoid profile as
well as their potency, herbal cannabis and cannabis resin would have potentially
demonstrated greater variations in pharmacological effect. The appearance of these
products gave little indication of their cannabinoid content.
Experienced users of this cannabis, who self-titrated their dose by smoking or

vapourising, may have had the ability to adjust their intake according to the potency of
the material. Potentially suitable blood plasma cannabinoid level could have been
reached within a few minutes (Huestis et al., 1992). Users could thus respond quickly
to a perceived under-dosing of cannabinoid. The variability in THC content would have
presented greater problems for those who relied on eating cannabis-based foodstuff for
medicinal purposes. Cannabis that is ingested would experience first-pass metabolism.
As a result, a large proportion of the THC would be metabolised to 11-OH-THC. A
number of other lesser metabolites would also form and plasma levels of delta9-THC
would not peak until one to six hours after ingestion Consequently, those users
42
ingesting cannabis would have less ability to accurately self-titrate to achieve their
optimum dose. The final bioavailability of ingested THC is estimated to be just 6%
compared to 10-27% during smoking (Hawksworth et al., 2004). Obviously, those who
were supplied with low-potency cannabis may indeed find that they were unable to
achieve sufficient THC plasma-levels, especially if ingesting the material.
From 1994 to 1998 resin accounted for 69-72% of all cannabis seizures in England and
Wales. From 1997 onwards the proportion of resin seizures fell annually, reaching 53%
in 2002 (Mwenda et al., 2003). However, although resin still appeared to be the
dominant form of illicit cannabis in the recreational market at the time, the study by
Hough et al. (2003) and an unpublished subset of data from the patient survey by Ware
et al. (2005), suggested that only a minority of medicinal cannabis users were choosing
resin despite the fact that this was the way they could gain access to substantial doses
of CBD. It has often been suggested that medicinal users appreciate the ‘high’ that can
be achieved from cannabis, perhaps welcoming it as a pleasant distraction from their
symptoms. In a published personal testimony, presented on behalf of the Alliance for
Cannabis Therapeutics (HLSCST, 1998b), Clare Hodges stated that when treating her
MS symptoms, she did not have to get ‘’high’’ for cannabis to lift her mood and make
her feel calm and positive. In many informal conversations with those having multiple
sclerosis, psychoactive effects were always described as undesirable. In clinical trials
with Sativex®, any psychoactive effects are regarded as undesirable events. There is
therefore a paradox. Only resin contained substantial quantities of CBD, a cannabinoid
that was efficacious in its own right and able to reduce the undesirable psychoactive
properties of THC. In-vitro studies and limited clinical trials also suggested that
THC/CBD mixtures were more efficacious than THC alone, when treating cancer pain
(Johnson and Potts, 2005) and other medical conditions (confidential GW
Pharmaceuticals data). Yet illicit users appear to have preferred to use sinsemilla or
herbal cannabis, which evidently lacked CBD.
The lack of preference for resin may have been due to adverse experiences
encountered with such a variable product. It also generally contained much less THC
than sinsemilla, and may have often been too weak to deliver sufficient active
ingredient. For those wishing to smoke cannabis resin, the product would have to be
mixed with tobacco or some other herbal material to support combustion. Preparation
of such mixtures is more difficult than those incorporating herbal cannabis or
sinsemilla. This requires a high level of dexterity which may present difficulties for
some patients. In herbal cannabis, and more so in sinsemilla, the natural plant
structure is still clearly visible and additives would be relatively easy to detect. In resin
43
this is not possible. Large quantities of soil are reported to contaminate the material at
harvest and adulterants added to increase yield or bind together poor quality resin
powder (Clarke, 1998) although forensic analysis has not found clear evidence of the
adulteration (King et al., 2004). However, suspicion remains.
As stated earlier, many MS patients confided that they were medicating with cannabis
that they had cultivated themselves. This finding was also reported in the small-scale
exploratory study performed by Hough et al. (2003). Cannabis seeds of high-THC
cannabis varieties are readily available for purchase. Screening cannabis seed
sources, for the research described in forthcoming chapters, tests were performed to
ascertain the chemotype of twenty four commercially available varieties. Of these,
twenty two produced negligible amounts of CBD. Of the other two varieties, most
seeds produced plants with a high-THC chemotype but three seedlings were of a
heterozygous mixed THC/CBD chemotype. This approximated to 2% of all the
seedlings tested. Patients wishing to grow their own cannabis would therefore find that,
although seeds of a vast number of varieties are commercially available, those that
produce CBD are rare. This suggests that, as with those buying herbal cannabis or
sinsemilla through the illicit market, those growing their own plants are likely to be
raising material almost devoid of CBD.
Any medicinal value obtained from sinsemilla would be mostly attributable to the
cannabinoid THC. However, several researchers have found that preparations made
from the cannabis plant are more efficacious than THC alone (McPartland and Russo,
2001, Costa et al., 2007, Ryan et al., 2007, Williamson, 2001) due to synergies
between THC and other poorly defined components in the plant. Those active
ingredients contributing to this synergy (such as volatile monoterpenes) may be more
abundant in sinsemilla than in resin, thereby increasing its pharmacological activity.
This study did not examine the content of these other agents.
Imported herbal cannabis was typically of low potency, and large quantities of material
would possibly be required to provide a useful medicinal effect. This may be a reason
why some medicinal users in the UK survey reported using as much as 10g per day. A
similar dosage of 7-9 grams a day of marijuana was reported to be used by those with
chronic medicinal conditions in the USA (Conrad, 2004), where THC contents of that
material are typically low (ElSohly, 2000).
Research for this thesis showed that sinsemilla potency increased significantly
between 1996 and 2005 and the efficacy may have altered accordingly. Herbal
cannabis and resin potency appeared to have changed little. However, in recent years
44
the potency of cannabis products in the Netherlands has increased greatly. In 2004
tests showed the mean THC content of resin manufactured and sold in Netherlands
coffee shops to be 39% and CBD content just 1% (Pijlman et al., 2005). This
compares to 4% THC and 4% CBD in resin circulating in England. Cannabis resin may
experience similar potency trends in the UK with marked effects on medicinal efficacy.
For those recreational users who use cannabis for the enjoyment of the psychoactive
effects, especially the young with a disposition to psychoses, this increasingly potent
material introduces an increased threat to safety.
2.6 CONCLUSIONS
As reported reported by Potter et al. (2008), cannabis circulating in England in 2004/5
was a very unpredictable product and the THC content varied greatly. The research
performed here recorded for the first time the variability in the CBD content and the
THC:CBD ratios of illicit cannabis in England. This variability would affect the
pharmacological properties of the material. A decreasing proportion of the material
circulating was in the form of resin, which was the only significant source of the anti-
psychotic cannabinoid CBD. The market was becoming increasingly dominated by
sinsemilla. This was shown to have significantly increased in THC content since 1998.
Increasing concern has been expressed in recent years, regarding the link between
cannabis potency and psychosis – especially in teenage users (Smith, 2005). The
implications of increasing THC content and diminishing CBD content on users of
recreational cannabis has been well documented. For medicinal users, the variability
in THC content also implies an unpredictability in the potential efficacy. However, to a
certain extent the variability in THC content of sinsemilla and herbal cannabis could be
overcome by self-titrating the dose appropriately. In resin however, the ratio of THC,
CBD and CBN varies greatly and this variability cannot be overcome by self-titration. In
sinsemilla and herbal cannabis, CBD is lacking and the rapid decline in availability of
resin indicates that this cannabinoid is becoming less available to medicinal users of
illicit cannabis. CBD is similarly scarce in commercially available cannabis varieties
sold by seed companies, and not easily accessible to those growing their own
cannabis.
A powder found within a ‘herb grinder’ was shown to be dislodged glandular trichomes.
This had a THC content of over 40% w/w and was thus ten times more potent than the
average resin. This illustrated the part that glandular trichomes played in cannabinoid
biosynthesis. It also showed how high cannabinoid purities could be achieved by
separating these from the aerial plant material. With the ultimate aim of exploiting
45
cannabis trichomes, the following chapter reports research performed to gain a greater
understanding of their form and function.
46
Chapter 3. Cannabis trichome form, function, and distribution
Chapter 3 Cannabis trichome form, function, and

distribution
3.1 INTRODUCTION
In Cannabis sativa L, it is widely accepted that the cannabinoids are predominantly, if
not entirely, synthesised and sequestered in small structures called glandular
trichomes (Mahlberg et al., 1984). Most of the monoterpenes and sesquiterpenes found
in Cannabis were also located in these structures (Malingré et al., 1975; Turner et al.,
1980.) Trichomes are arguably therefore the part of the Cannabis plant of greatest
interest to the pharmacognocist. It is reasonable to suspect that the most productive
botanical raw material would contain these glandular trichomes in substantial
quantities, and at their optimum stage of development. Indeed, the reliable production
of optimum feedstocks would be more likely to be achieved if the grower gained a
greater understanding of their form, function and distribution. This would guide the
grower how to judge trichome maturity. Similarly, improved feedstocks might be
possible achieved by selecting plant parts with the optimum array of trichomes. As
seen in the previous chapter, one recreational cannabis product predominantly
consisted of trichome material. Similar enriched trichome preparations (ETPs) could
possibly warrant evaluation as potent sources of secondary metabolites for
pharmaceutical use. When characterising any phytopharmaceutical feedstock, a
systematic and illustrated report of the microscopical details is generally important
(Evans, 2002). This chapter of the thesis constitutes such a report. The following
chapter then examines how to exploit these to full potential.
Cannabis trichomes have been studied in depth for many years, a notable example
being the detailed descriptive and illustrated work of Briosi and Tognini (1894), which is
still regularly quoted. The 1970s was perhaps the period of most intensive study, much
of the research being performed with electron microscopes and meeting the growing
requirement for forensic identification of illicit cannabis products (Fairbairn, 1972;
Ledbetter, 1975; Dayanandan and Kaufman, 1976; Hammond and Mahlberg, 1973,
1977; Turner et al., 1977).
The general term ‘trichome’, when applied to plants, refers to a type of epidermal
appendage. According to one definition, ‘trichomes’ constitute an intermediate group of
appendages between ‘papillae’ and ‘emergences’ (Werker, 2000). ‘Papillae’
(protrusions of the periclinal outer cell wall) and ‘trichomes’ develop from epidermal
cells only (Uphof, 1962), whereas ‘emergences’ develop from both epidermal and
47
subepidermal tissues (Werker, 2000). The distinction between trichomes and

emergences is often far from distinct, and both types are commonly referred to
collectively as trichomes (Dietmar Behnke, 1984). Plant trichomes derive their name
from the Greek word thrix, meaning hair. They exist throughout the Plant Kingdom in
an extremely diverse number of forms, of which over three hundred have been
described (Payne, 1978). Their diversity has attracted much attention since the earliest
microscopists studied them and recorded their detail (Hooke, 1665). Workers have
chosen to catagorise trichomes in many ways according to their morphological and/or
anatomical features. For example, Theobald et al. (1980) placed the various trichome
morphologies under seven general headings. Dickison (1974) adopted three general
headings: - simple, complex and glandular which were each subdivided into various
forms. However, the major distinction was between non-glandular and glandular
trichomes the former being differentiated by their morphology, and the latter by their
secretory materials.
Non-glandular trichomes, in the form of simple plant hairs, occur in the majority of
species within the Tracheobionta (vascular plants), but glandular trichomes are found
in just 20-30% (Dell et al., 1978). Taxonomists commonly regard the Trancheobionta
as being divided into the Spore Forming Division, or ferns (Pteridophyta) and the
Flowering Division (Magnoliophyta). The latter is divided into the gymnosperms
(conifers) and the much larger angiosperms (broad leaved flowering plants). Secretory
glandular trichomes are mostly found within the Magnoliophyta (Dell et al., 1978; Fahn,
1988) but rarer examples are found in existing pteridophyte species, including the ferns
Pityrogramma sp and Polypodium virginianum (Peterson and Vermeer, 1984).
Fossilised remains of the fern Blanzeopteris praedentata reveal that glandular
trichomes existed as long as 300 million years ago, in the Late Carboniferous period
(Krings et al., 2003). The functions of trichomes are either guessed at or totally
unknown and many of the hypotheses have not been experimentally tested (Werker,
2000). These hypotheses include the deterrence of predators and protection against
environmental stresses (Rodriguez et al., 1984; Hallahan et al., 2000; Theis and
Lerdau, 2003; Acamovic and Brooker, 2005). Reviews by Werker (2000) and Wagner
et al. (2004) described seventeen different trichome functions, many of which could be
applicable in Cannabis.
As a result of funding from the tobacco industry, the species most studied for its
multicellular trichomes is Nicotiana tabacum (Glover et al., 2000). The plant family
receiving most trichome studies is the Labiatae, due to the importance of the
terpenoids to the food, cosmetic and pharmaceutical industry. This family incorporates
48
peppermint (Mentha piperita), sage (Salvia officinalis), thyme (Thymus vulgaris),

lavender (Lavendula officinalis) and over four thousand other species. Despite legal
restrictions to plant access, cannabis trichome structures are also well studied
(Fairbairn, 1972; Ledbetter et al., 1975; Dayanandan et al., 1976; Hammond et al.,
1977; Mahlberg et al.,. 1984; Kim and Mahlberg 2003). There has been a great overlap
in the knowledge gained by those studying Cannabis and other species. Cannabis
trichome researchers have commonly described two types of non-glandular trichome,
which have not been associated with terpenoid development. Three types of glandular
trichome have been described on female cannabis, viz bulbous, sessile and capitate
stalked. Males have been found to exhibit a fourth type – the antherial glandular
trichome, which has only been found on anthers (Fairbairn, 1972).
Photosynthesis would be the original source of the carbon utilised in the biosynthesis of
the secondary metabolites in glandular trichomes. In some species however, e.g.
Nicotiana tabacum, photosynthesis would actually take place in chloroplasts within the
trichome (Akers et al., 1978). Dayanandan and Kaufman (1976) described finding
chloroplasts within the stalks of bulbous trichomes in Cannabis (the smallest of the
cannabis trichomes) and reported their absence in the other trichome forms. With no
chloroplasts present, precursors of secondary metabolite biosynthesis would therefore
have to be translocated from elsewhere. Crombie (1977) reported that cannabinoid
biosynthesis had been observed to continue in ‘sport’ tissue lacking chlorophyll,
(although data was limited and statistical analysis not possible). It would appear
therefore that these precursors can be translocated from tissues well away from the
trichome.
As stated in a detailed research paper on cannabis trichomes by Dayanandan and

Kaufman (1976), their study is essential to understand the biogenesis, distribution and
function of the different cannabinoids. Mahlberg et al. (1984) showed that sessile and
capitate stalked trichomes differed in their distribution, as well as their cannabinoid
content and profile, and this was linked to the differing cannabinoid distribution in the
plant. Recent research has greatly increased the knowledge regarding the biogenesis
of the cannabinoids within the glandular trichomes, but further studies are required
(Sirikantaramas et al., 2005; Taura et al., 2007). These studies have added to the
incomplete debate regarding the function of the cannabinoids in the host plant.
3.2 AIM AND OBJECTIVES

In view of the central role of the trichomes in the production of cannabinoid-rich
material the work described here investigated trichome development, structure,
49
function and catabolism in Cannabis sativa L. The knowledge gained would hopefully
be used to make recommendations as to how cannabis should be grown, harvested
and processed to make maximum potential benefit of the secondary metabolites
synthesised in the cannabis trichomes. Maintaining a realistic view that fellow growers
would be unlikely to have ready access to electron microscopes, this work was
unashamedly performed using microscopes costing no more than a few hundred
pounds sterling. The research involved a program of studies with a range of linked
objectives, these being: -
3.2.1 To study and photograph the structure and apparent function of trichomes in
Cannabis sativa L.
3.2.2 To assess how the differing sessile and capitate stalked trichome populations
affected the secondary metabolite content of cannabis tissues.
3.2.3 To assess the effect of capitate stalked trichome density and colour on
cannabinoid content and profile.
3.2.4 To measure the effect of photosynthetic ability, or lack of ability, on the

cannabinoid content of green and yellow tissue in variegated Cannabis sativa.
3.3 MATERIALS
3.3.1 Germplasm
Clone Characteristics Variety Name Supplier
G1-M3 High-THC Guinevere GW Pharmaceuticals

G2-M6 High-THC Galina
G2-M7 High-THC Gina
G5-M16 High-CBD Gill
G60-M55 Variegated Unnamed
M280 High-CBG Unnamed
M186 High-THC Pigmented Sweet Purple Paradise Seeds
P.O. Box 377, 1000 AJ
Amsterdam, Holland
Table 3.1 The names and suppliers of the cultivars used.
50
3.3.2 Microscopy, Tissue Stains, Photography and other Apparatus
Apparatus Source
Olympus OM10 35mm SLR camera Olympus UK Ltd, 2-8 Honduras St,
London EC1Y 0TX
Olympus SP350 8 megapixel camera Brunel Microscopes

Stereo insert 30mm lens tube, Unit 12 Enterprise Centre,
SP-350 Unilink (universal adaptor) Bumpers Industrial Estate
Bumpers Way, Chippenham,
Photonic PL2000 - double arm cold light WILTS. SN14 6QA
source.
MX3 Low Power Light Microscope
High Power Stereo Light Microscope with STE UK Ltd., Staplehurst Rd

Trinocular Head for camera attachment. Sittingbourne, ME10 2NH
Eye Piece Graticule for Specimen Size

Measurement
Hemp Oil (Culinary grade) for mounting of Motherhemp, Springdale Farm,

microscope specimens Rudston, East Yorkshire YO25 4DS
RS 732-139 Hole Punch Kit RS Components Ltd., Birchington Road,

Corby, Northants, NN17 9RS,
Fast Blue Tissue Stain Sigma-Aldrich Company Ltd., Fancy

2,3,5-Tetrazolium Chloride Stain. Road, Poole, Dorset. BH12 4QH
Glycerol
Table 3.2 Photographic, microscopy and other miscellaneous items and commercial sources.
3.4 METHODS
3.4.1 Photomicrograph Studies
3.4.1.1 Choice of Microscopes
Two levels of light microscopy were used in this study. For more general observations
of plant tissue a low power microscope was employed. Using a Brunel MX3 binocular
microscope, fitted with WF10X eyepieces and 1X or 3X objectives, samples of plant
51
tissue up to 15 mm in diameter were observed. For more detailed observations, of

small numbers or individual trichomes, a higher level of magnification was required.
For this, a high power stereo light microscope fitted with x10 eyepieces and x4, x10
and x40 objectives was utilized. This gave fields of view of up to 4.5mm in diameter.
3.4.1.2 Staining
Two stains were utilised to aid visibility of trichome internal cell strictures and to gain
further clarification of the sites of secondary metabolite biosynthesis and storage. The
first stain fast blue, which is commonly used when analysing cannabinoids by thin layer
chromatography, has also been used during histochemical studies on cannabis
trichomes (André and Vercruysse, 1976). A 0.3% w/v aqueous solution stains
cannabinoids orange or pink, but other phenolic compounds are also stained.
The second stain used was the ‘vital stain’ 2,3,5-Triphenyl tetrazolium chloride, also
known as TTC or tetrazolium red. This is an almost colourless water-soluble stain. In
the presence of viable tissue this is reduced (probably by dehydrogenase enzyme
activity) to insoluble red triphenyl formazan (Smith, 1951). A 1% w/w solution was
used for rapid staining (< 1 hour) and a weaker 0.1% w/w used when slowly staining
samples overnight.
3.4.1.3 Unmounted Sample Preparation
The majority of the low power microscope observations were made on unmounted
specimens. For these, small pieces of plant tissue were cut from the plant and placed
directly onto the low-power microscope plate. As a result, areas of pubescence
containing over one hundred trichomes could be observed in a single view. The same
minimalist method of sample preparation was sometimes utilised when viewing plant
tissue on the high power microscope. However, to achieve views where large
proportions of the material were simultaneously in focus, flat samples were likely to
produce success. For this, areas of tissue up to a few millimetres in diameter would be
laid flat on a microscope glass slide. Trichome filtrates were typically observed by
smearing these onto a glass slide.
3.4.1.4 Mounted sample preparation
All the mounted samples prepared in this study were designed to be temporary, and
disposed of immediately after use. To mount intact pieces of plant tissue, small pieces
up to 1cm diameter were placed on a glass slide and a few drops of mounting fluid
placed on the specimen. A cover slip was placed on an angle above the specimen,
52
with one edge of the slip touching the slide. The cover slip was then lowered onto the
specimen. As a result, excess fluid and trapped air bubbles would be expelled from the
underside of the slip, and removed with paper tissue.
Water (refractive index 1.33) was sometimes used as a mounting medium. However,
transparency of biological samples is best achieved by selecting a medium with a
refractive index closest to that of the subject (Delly, 1988). Conversely, definition of
colourless objects is increased by choice of a mountant with a refractive index different
from that of the object, the ideal ratio for clarity being 1.06 to 1.00 (Evans, 2002). The
refractive index of glandular trichome contents was not known. However, initially
aiming for transparency and knowing the refractive index of two of the major
constituents of glandular resin head - myrcene and trans-caryophyllene - to be 1.48
and 1.50, hempseed oil and glycerol (refractive index 1.47) were selected.
3.4.1.5 Illumination
The Brunel MX3 low power microscope incorporated two light sources. These could be
directed vertically upwards and/or downwards onto the subject. The material was
sometimes illuminated, by incident light, using a Photonic PL2000 - double arm ‘cold
light source’. By the inclusion of optical fibres, this enabled white light from a halogen
lamp to be directed at the subject without any accompanying radiant heat. Alteration of
the flexible illuminating arms enabled the angle of incidence of light to be adjusted.
Some samples, when placed on the STE high power microscope, were also illuminated
using the cold light source while others were illuminated from below. When viewing
samples mounted beneath a cover slip, the microscope was set up using the Köhler
illumination method (Delly, 1988), which first ensures that the light from the condenser
lens is correctly focused on an empty microscope slide. When viewing unmounted
specimens the condenser height and aperture were adjusted while viewing the subject
until optimum resolution was achieved.
3.4.1.6 Photography
To enable photographs to be taken through the low power microscope, one eyepiece
was replaced with a compatible 30 mm lens tube to which an Olympus OM10 35 mm
single lens reflex camera or Olympus SP350 8 megapixel digital cameras would be
attached. As in ordinary photography, the depth of field is considered to be the distance
from the nearest object plain to the farthest object plain that is in focus. When objects
are a long distance from the camera lens the depth of field is large. However, depth
53
decreases as the image comes closer to the lens. When taking photomicrographs,
depth of field is measured in micrometers (Delly, 1988). To maximize the chance of
finding substantial areas of tissue simultaneously in focus within this narrow depth of
field, multiple samples were laid as flat as possible on glass slides. Aided by surface
tension, samples mounted beneath a cover-slip were more likely to be retained within a
narrow area of focus.
The earlier studies in this research used single lens reflex photography and colour print
film. For high magnification situations, the required exposure times occasionally
exceeded ten seconds. Following brief tests (data not shown) high-speed ASA 400
film was selected. This reduced exposure times while maintaining sufficiently high
resolving power. Later studies were performed using digital photography. In all cases,
photomicrographs were taken on a solid bench and the shutter activated remotely to
reduce manually-induced camera-shake.
3.4.1.7 Isolation and Observation of Detached Glandular Resin Heads
Adapting a method developed for studying trichome resin heads of thyme Thymus spp,
(Yamaura, 1992) individual resin heads from capitate stalked trichomes were
simultaneously removed and fixed on adhesive tape. One layer of sellotape® was
tightly wrapped around the handles of a domestic clothes peg with the adhesive
surface facing outwards. The taught adhesive surface was then allowed to touch the
surface of cannabis bracts where glandular stalked trichomes were present. The
adhesive surface was promptly fixed to a clean microscope slide. Detached glandular
heads were occasionally trapped in the adhesive and readily viewed.
3.4.2 Effect of glandular trichome array on the secondary metabolite

content of plant tissues
For this study, bracts of nearly-mature cannabis inflorescences were selected where
the pubescence of capitate stalked trichomes was only visible on approximately half of
the area (Figure 3.1). In such cases, it was always the proximal tissue that displayed
this pubescence. Twenty bracts of each of three high-THC clones were selected.
These were cut with a scalpel to separate each bract into portions where stalked
trichomes were judged to be present or absent. Using a low power dissection
microscope, the density of stalked and sessile trichomes in the separate tissues was
assessed by counting the number, within a single randomly-selected field of view of
16.6 mm-2, on both the upper and lower surface of each of these bract sections.
54
Samples were then bulked to produce one sample of proximal bract tissue, and one
sample of distal tissue, for each of the three clones. The samples of proximal and distal
bract material of were then separately dried in an oven at 40ºC for twenty four hours.
The cannabinoid content was assessed by GC (Appendix 1).
Female Flowers
Distal region of bract devoid

of capitate stalked trichomes
Proximal region, densely

covered in a pubescence of
capitate stalked trichomes
Figure 3.1. Upper surface of a bract within a cannabis inflorescence showing glandular stalked
trichomes to be present only within the proximal region (Potter, D..J.).
3.4.3 Organoleptic assessment of the effect of trichome colour and

pubescence density on cannabis potency.
This study utilised over two hundred and fifty illicit samples of sinsemilla cannabis
inflorescence seized by police from five constabularies during 2004/2005 (Potter et al.,
2008). These were collected from police storage and relocated to a central
dehumidified store (30% +/- 5% RH) and kept in darkness at ambient temperatures
prior to assessment.
All samples were visually examined using a Brunel MX3 low power microscope.
Trichome density and trichome colour were awarded single overall scores on a
subjective 1-9 scale (Table 3.3). This was based upon those commonly used by the
National Institute of Agricultural Botany for the subjective assessment of plant
characteristics (NIAB, 2007). For trichome density a score of 1 was awarded to the
maximum trichome density and progressively higher scores denoted a thinner
pubescence with a score 9 denoting that no intact glandular stalked trichomes were
visible. For trichome colour a score of 1 denoted a sample within which the vast
majority of glandular trichome heads were completely clear. With maturation these
resin heads can typically become turbid and then brown. The score would be awarded
55
for the colour of the majority within the sample. The samples were subsequently
analysed for cannabinoid content by GC (Appendix 1).
Score Density Colour

1 Maximum Density Clear
2 Extremely Dense Misty White
3 Very Dense Translucent
4 Moderately Dense Opaque White
5 Intermediate Density Slightly Brown
6 Scarce Light Brown
7 Very Scarce Mid Brown
8 Extremely Scarce Dark Brown
9 Absent Very Dark
Table 3.3 The 1-9 scales for overall capitate stalked trichome density and resin head colour.
Within each sample some variation would occur.

biosynthesis in sessile trichomes.
This study utilised a very rare variegated cultivar of Cannabis sativa L., G60 M55. In a
first of three tests, three plants were selected and leaves collected where pure yellow
and dark green could be viewed on either side of the leaf’s midrib. Using dissection
scissors, areas of approximately equal size were sampled from symmetrically opposite
areas on each side of the leaf’s midrib. These varied in size but were typically each no
more than 100 mm2. For each colour tissue, one bulked sample was produced per
plant - each containing twenty portions of leaf. The emphasis on symmetry was
designed to ensure that the green and yellow samples were at identical stages of cell
expansion. Samples were oven-dried at 40ºC for 24 hours and analysed for
cannabinoid content by GC (Appendix 1). In a second test the above procedure was
repeated with just two plants.
A third test was performed where the cannabinoid quantification was based upon leaf
area rather than leaf weight. For this test a disk cutter was used to cut equal sized
disks from totally green or yellow areas. Leaves of even shape were selected where
disks could be cut from symmetrically-located pure yellow and green areas (Figure
3.2). Twenty disks (1 cm2 diameter) of each colour were cut from each of two plants.
56
Samples were dried as in the above test. From each plant, the twenty disks of each
colour were bulked, thereby providing two replicates per colour. These were analysed
by GC.
Figure 3.2. A variegated leaf of clone M60, with 1cm diameter disks cut from symmetrically
opposite sides of the midrib (Potter, D..J.).
In both tests the samples were viewed through a microscope and an assessment of
sessile trichome density was attempted.
3.4.5 Statistical Methods

Basic calculations of SD (standard deviation), SE (standard error) and ANOVA
(analyses of variance) and regression were performed using Microsoft Excel software.
3.5 RESULTS AND DISCUSSION
3.5.1 Photomicrograph studies
All six forms of trichome described by Fairbairn (1972) were examined. These are
described in turn.
3.5.1.1 Simple unicellular trichomes
An example of this type is shown in Figure 3.3a. The trichome is seen to have
developed from a cell within the epidermis. These simple trichomes, also known as
covering trichomes, were the first to appear. These were initially observed on the
surface of cotyledons immediately after germination. This form continued to develop in
abundance on the underside of leaves (and to a much lesser extent on the upper
57
surface) throughout the plant’s life. These trichomes were typically orientated to face
towards the distal part of the leaf, bract or bracteole and in many instances would lie
almost flat on the surface. This pubescence of trichomes would cover the underside of
the leaf with a layer of trapped air, thereby reducing water loss and providing some
insulation against extreme temperatures (Rodriguez et al., 1984).
3.5.1.2 Cystolythic trichomes
This type of trichome is shown in Figure 3.3b. Specimens were first observed on the
upper surface of the initial pair of true leaves on a cannabis seedling. Always pointing
towards the distal part of the leaf, these trichomes gave the upper surface a texture
that was rough to the touch.
At the base of each trichome is a cystolyth. These concretions are common throughout
the plant kingdom and are typically formed from calcium oxalate or calcium carbonate
crystals. Those found in Cannabis are of the latter type (Dayanandan and Kaufman,
1976; Evans, 2002). These tough trichomes would presumably reduce the palatability
of the foliage to leaf-eating predators. Histochemical chemical staining of these
trichomes with fast blue occasionally resulted in pigmentation of these organelles, but
the presence of phenolic substances on these trichomes was attributed to
contamination from leaking capitate stalked trichomes. This supposition was supported
when the vital stain tetrazolium red was used. No reduction of tetrazolium was
observed apart from in the cells immediately surrounding the cystolyth, where
respirative activities accompanied the formation of these concretions.
20µm 50µm
(a) (b)
Figure 3.3. (a) Unicellular non-glandular trichome. The sample is temporarily mounted under
hemp oil and viewed in transmitted light; (b) Cystolythic trichomes observed on the leaf margin
of a young leaf. The sample was temporarily dry-mounted and viewed in transmitted light.
Cystolyths (concretions of calcium carbonate) are visible at the base of each trichome
(Potter, D..J.).
58
3.5.1.3 Capitate sessile trichomes (more commonly simply called sessile trichomes)
20µm 20µm
(a) (b)
Figure 3.4. (a) a capitate sessile trichome observed on the edge of one of the first pair of true
leaves of a cannabis seedling. The specimen was temporarily dry-mounted and viewed using
both transmitted and incident light; (b) a sessile trichome on a leaf surface stained with Fast
Blue. The still-wet sample was temporarily dry-mounted and viewed using incident light
(Potter, D..J.).
Apart from on the cotyledons and the supporting hypocotyl, sessile trichomes were
observed on all other aerial surfaces throughout the plant’s lifespan. The example
shown in Figure 3.4a was a very rare find, being situated on the margin of one of the
first pair of monofilous leaves of a cannabis seedling. This location enabled the
specimen to be observed in profile, without the need of microtome sectioning.
Those previously studying cannabis trichomes have referred to this type as sessile
(Latin sessilis – sitting) or capitate sessile (Latin caput – head). By definition this type
does not have a stalk. The trichome is in fact connected to the mesophyl cells via a
stalk cell, but this is hidden beneath the trichome resin head. The stalk cell was seen to
contain chloroplasts, enabling some photosynthetic activity. This form of trichome is
found in many other plant families and, due to its flattened shape and short stalk, is
often referred to as a ‘peltate’ trichome (Latin pelta – a short-handled hand-held round
shield). According to some plant physiologists, the development of this structure from
sub-epidermal cells defines this structure as an ‘emergence’ rather than a ‘trichome’ –
the latter always developing from an epidermal cell. However, as stated earlier, most
authors regard both structures as ‘trichomes’ (Werker, 2000).
The glandular head (or resin head) incorporates a disc of secretory cells at the base.
These appear to totally lack chlorophyll. Above the secretory cells, and below the
trichome’s outer membrane, is a chamber within which the secretory cells sequester a
resinous mixture that includes cannabinoids and essential oils (Mahlberg et al., 1984).
59
Mature sessile trichomes are reported to typically have eight secretory cells within the
disk. When the trichomes were viewed from the side, as in Figure 3.4a, it was not
possible to confirm this. When illuminated and viewed from overhead the disk of
secretory cells was also difficult to observe. This problem was exacerbated by the fact
that the glandular head itself acted as a powerful convex lens. As a result it was difficult
to gain an undistorted view of structures with the glandular head. As the sessile
trichomes matured, the interior of the structure would typically turn translucent or
opaque white, further hampering observations of internal structure. Trichomes would
sometimes turn brown in very mature or aged samples. A clearer confirmation of
secretory cell numbers was achieved by immersing the cannabis leaf in fast blue stain
for ten minutes and then rinsing in water immediately prior to examination. Trichomes
varied in the speed at which the stain was absorbed and assimilated. Some samples
exhibited very little staining while others became totally red, allowing no visible
differentiation of tissues. In ideal situations only the membranes of the secretory cells
were stained, allowing a clearer view of the trichome’s internal structure (Figure 3.4b).
Their function is not known but across the Plant Kingdom their role is guessed to be the
protection of the plant tissue against predators.
3.5.1.4 Antherial Sessile Trichomes
Sessile trichomes were also observed on cannabis anthers. The observations

supported the theory of Fairbairn (1972) that these ‘antherial sessile trichomes’ were a
distinct form of trichome. These antherial trichomes were different from all other
sessile trichomes by virtue of their larger size (Figure 3.5). With a diameter of
approximately 70-80µm these unique sessile trichomes were significantly larger than
those located elsewhere. Sessile trichomes were also generally present on the calyx
surrounding these anthers, but these were of the more normal smaller form. Very
similar antherial trichomes are observed in the furrows of the anthers of the male hop
Humulus lupulus L. (Neve, 1991a).
60
1 mm 100 µm
(a) (b)
Figure 3.5. (a) a row of antherial sessile trichomes showing their normal distribution in the
furrow of a cannabis anther. These anthers were captured in incident light through a low-power
microscope. (b) closer view of antherial trichomes (Potter, D..J.).
3.5.1.5 Capitate Stalked Trichomes
These trichomes were generally abundant on the calyx, bracteoles, bracts and
accompanying petioles of female plants. Although sufficiently rare on male plants to be
thought totally absent by some researchers (Hammond and Mahlberg, 1977), their
existence has been confirmed by others (Dayanandan and Kaufman, 1976). During
studies for this study, capitate stalked trichomes were only found on males of a few
varieties and these were restricted to the filaments bearing the anthers. Capitate
stalked trichomes were the most complex. As shown in Figure 3.6, they developed a
resin head, similar to that of the sessile type, but in mature specimens this was
surmounted on top of a multicellular stalk. As experienced by previous workers
(Mahlberg et al., 1984) it was not always easy to distinguish between sessile and
immature glandular stalked trichomes, where the stalk was yet to form. Observations
confirmed the findings of others (Ledbetter et al., 1975) that resin heads on capitate
stalked trichomes are typically 70-100 µm in diameter, compared to sessile types which
were more typically 40-50 µm. The stalk consisted of two distinct cell types.
Hypodermal cells, at the core of the stalk, formed an active channel through which
nutrients could be transported to the glandular head from the phloem. This was
surrounded by a single layer of outer epidermal cells, which was a continuous
extension of that covering the bract or bracteole surface.
61
100µm
Figure 3.6. A capitate stalked trichome (centre) between two cystolythic trichomes. The
specimen is temporarily dry-mounted and illuminated from below. The secretory cells are out-
of-focus due to the optical distortion within the glandular head (Potter, D..J.).
Nineteenth-century studies, using the light microscope, identified a disk of secretory

cells at the base of the resin head (Briosi and Tognini, 1894). More up-to-date electron
microscope studies have been performed which examined the morphology of this
feature (Mahlberg et al., 1984; Mahlberg and Kim, 1991; Kim and Mahlberg, 1991,
2003). The secretory cell disk was readily found during light microscope studies for
this thesis, albeit initially in the form of a poorly defined mass of translucent material at
the base of the glandular head. Viewing techniques had to be adjusted to identify the
individual secretory cells within the disk.
As with the sessile type, the resin head on capitate stalked trichomes would swell
during the early stages of development, as the secretory cells became active. The
contents of the resin head were clear during the earlier stages of development, but
would become opaque-white in older specimens. In most cases this would not occur
until trichomes were at least four weeks old, but the rate of loss of transparency was a
genotypically-dependent characteristic. In CBG-dominant clones (eg M280, Figure
3.13) glandular heads would become a dense opaque-white when just a few days old.
This was an unusual example of where the microscope could be used to help identify a
cannabis plant’s chemotype. Further ageing of most genotypes would sometimes result
in the resin heads turning brown. This colouration was often seen to commence within
the disk of secretory cells, and was possibly due to necrosis of these now inactive
tissues. This browning would continue after plants had been harvested and dried.
Viewing cellular structures within the trichome and surrounding tissues was frequently
hindered by their refractive properties. This was especially the case when viewing
62
unmounted specimens. The interior of the glandular resin head was particularly difficult.
Although the tissue was crystal-clear during the early stages of development, as
observed with sessile trichomes the glandular head acted as a powerful convex lens.
When illuminated from below the transmitted light was refracted by the glandular head
contents. This could lead to a distorted virtual image being formed, which appeared to
be located outside of the structure, as in Figure 3.7.
Secretory cavity
Secretory cells
100µm
Resin head
base
Figure 3.7. Two dry-mounted capitate stalked trichomes viewed in transmitted light. Most of
the features are out-of-focus. In the right-hand trichome, a crisp view of cells within the
secretory cell disk appears as an in-focus image. However these appear to be located outside of
the trichome structure, due to the refractive properties of the resin head (Potter, D..J.).
Mounting the specimen in an appropriate fluid reduced the aberration, as this

decreased the difference in refractive index between the specimen’s contents and that
of the surrounding medium, the refractive index (RI) of air being 1.0. Initial tests found
water to be a poor mounting medium. Air bubbles frequently became trapped between
the hydrophobic waxy plant surfaces and the water. Possibly due to its refractive index
(1.33), image clarity was relatively poor when samples were mounted in water.
Glycerol (1.47), hempseed oil (1.47) and conventional immersion oil (1.51) have similar
refractive indices to those of the main terpenes within the resin head (myrcene 1.48,
trans-caryophylene, 1.50) and, possibly because of this, these were found to be more
suitable mounting media. A 70% v/v aqueous solution of glycerol was less viscous and
sometimes easier to use. This also offered a slightly lower RI of 1.44.
With appropriate lighting, the disk of secretory cells was readily observed at the base of
the glandular head, as shown in Figures 3.8a. When dry mounted, the secretory cells
were poorly defined and appeared dark and translucent. When mounted in glycerol,
more details of the secretory cell disk could be observed. Viewed in profile, the disk
63
appeared as an undulating semi-opaque conglomeration of cells (Figure 3.8b). An

illustration of a trichome viewed from the same angle, as drawn by Briosi and Tognini
(1894), is included for comparison (Figure 3.8c). In all three Figures 3.8a-c, the stalk is
seen to include an inner channel of hypodermal cells surrounded by an outer layer of
epidermal cells. At the uppermost extremity of the hypodermal channel, Figures 3.8b
and c show a more opaque ‘basal cell’. The Briosi and Tognini illustration also shows
this basal cell to be surmounted by two of the four stipe cells, the distal surface of
which acts as a base for the secretory cells. The stipe cells cannot be seen clearly in
this photomicrograph (Figure 3.8b).
100µm 100µm
(a) (b) (c)
Figure 3.8. (a) A temporarily dry-mounted capitate stalked trichome viewed in transmitted light.
An irregular arrangement of poorly-defined secretory cells is visible at the base of the glandular
head (Potter, D..J.); (b) A capitate stalked trichome, temporarily mounted in glycerol and
viewed in transmitted light (Potter, D..J.), and (c) an illustration of a capitate stalked trichome
on Cannabis sativa by Briosi and Tognini (1894).
While observing capitate stalked trichomes mounted in oil, the secretory cells were
perceived to be completely clear structures. These were located above a layer of more
opaque tissue (Figures 3.9a and b). Transmission electron microscope (TEM) studies
have shown the distal part of the secretory cells to contain extensive areas of hyaline
tissue (Kim and Mahlberg, 2003), which may correspond with the clear secretory cell
tissue viewed here. Related TEM studies also identified increased levels of thickening
between the secretory and stipe cells (Mahlberg and Kim, 1991), and similar thickening
64
may explain the presence of the apparently-opaque tissue seen at the base of the resin
head in these studies.
Secretory space
100µm 100µm
Secretory cells
Thickened
glandular head
basal tissue
Stalk with outer
layer of epidermal
cells and inner
channel of
hypodermal cells
(a) (b)
Figure 3.9. (a and b). Similar sized capitate stalked trichomes temporarily mounted under hemp
oil. The samples are viewed in transmitted light. Possibly because of a similarity in the
refractive index of the oil and the secretory cell contents, these cells appear clear. The outer
membrane at the base of the glandular head appears dark and opaque (Potter, D..J.).
TEM studies of the resin head of capitate stalked trichome resin heads have also
indicated that the secretion produced by the secretory cells enters the secretory cavity
in the form of so-called vesicles (Mahlberg et al., 1984). These vesicles are separated
by a network of fibrils, which have a wall thickness approximately half that of the
secretory cell membrane. Studies of untreated specimens for this thesis, using the STE
light microscope only rarely gained a clear view of the vesicles. Indeed it was initially
postulated that the clusters of clear organelles within the secretory head in Figure 3.9a
and b were not secretory cells, but secretory vesicles. These would have been
produced from secretory cells, which were initially perceived to be the dark tissues at
the base of the secretory head. Although this interpretation of the observation is still
possible, the size of the clear structures appears too large for them to be vesicles,
which electron microscope studies have shown to be a few microns in diameter
(Mahlberg et al., 1984). While viewing these secretory cells from directly above the
trichome, using 70% v/v glycerol as a mounting medium and with light being
transmitted from below, more credible identification of secretory vesicles was
65
sometimes possible (Figure 3.10). Illuminated in this manner, the secretory cells and
the secretory vesicles appeared transparent. To maximise the sharpness if the image
within the optically-distorting secretory head, the microscope was initially focused on
the subject and the condenser lens then adjusted to gain maximum clarity.
Appropriately focused, the reported tally of sixteen radially-arranged secretory cells
could sometimes be counted. As shown in Figure 3.10, when viewed in profile, the
disk of secretory cells appeared more opaque.
Secretory cell disk

viewed in profile
Directly overhead
view of secretory
cells, as seen
through the
secretory head
Secretory vesicles
Figure 3.10. A contrasting pair of resin heads on capitate stalked glandular trichomes, naturally
orientated to allow sideways-on (left) and overhead views (right). The specimen is temporarily
mounted in a 70% v/v aqueous solution of glycerol and illuminated from below (Potter, D..J.)..
Mahlberg et al. (1984) stated that secretory vesicles form above the secretory cells and
migrate through the secretory head, depositing some of their contents on the secretory
cavity’s outer membrane, thereby maintaining its strength as it inflates. In their studies
the outer cuticle of the secretory head was seen to increase eight-fold in thickness as
the trichome developed. These workers used THC monoclonal antibodies to detect the
presence of THCA within the secretory head. THCA was only detected when attached
to the vesicle wall and none was discovered within the vesicle interior. If THCA was
indeed restricted to the vesicle surface, it would not be possible to achieve the THCA
concentrations of over 40% w/w found in some cannabis resin samples (Potter et al.,
2008; Hardwick and King, 2008).
The metabolic activity within the secretory cells disk has been well studied by many
researchers and is still the cause of some debate. The biosynthesis of monoterpenes
66
and sesquiterpenes, within the membranes of secretory cells, has been identified as
occurring in plastids and endoplasmic reticulum respectively (Croteau et al., 1984).
Mahlberg et al. (1984) suggested that these plastids were also involved in cannabinoid
biosynthesis. Recent research shows that CBGA, the biosynthetic precursor of THCA
and other cannabinoids, and THCA synthase enzyme are found within the secretory
cavity (Sirikantharamas et al., 2005). This suggests that the secretory cavity is not just
the site for the accumulation of cannabinoids, but also the site of THCA biosynthesis.
The previous theory that THCA is biosynthesised in plastids within the secretory cells is
further weakened by the observation that THCA and CBGA are cytotoxic
(Sirikantharamas et al., 2005). By sequestering the cannabinoids in the secretory cavity
the secretory cells would be protected from damage. The enzymes for cannabinoid
synthesis within the secretory cavity would require water and molecular oxygen to
function. As the hydrophobic terpenes accumulate within the vesicles, the oxygen and
water would most likely be transported via the vesicle fibrillar wall. However, alternative
studies suggested that CBGA indeed may be synthesised in the secretory cells, but
further investigations are needed (Taura et al., 2007).
The secretory cells were confirmed as being metabolically highly-active sites by the
use of the vital stain tetrazolium red (Figure 3.11a). The cells were the first to take up
the stain and develop a red colour. A 1% solution stained most trichomes within one
hour, especially in the case in less-mature trichomes where the glandular head
contents were still clear. When the glandular head’s outer membrane was removed by
abrasion, to expose the disk of secretory cells, the secretory tissue stained red within a
few minutes (Figure 3.11b). The glandular cells within intact secretory resin heads of
more mature trichomes were often unstained, suggesting that metabolic activity had
ceased. However, detailed study of these cells in more-mature trichomes was
obscured by the age-related increasing opacity of the glandular head (Figure 3.11c). As
this figure also shows, prolonged exposure to tetrazolium red resulted in staining of the
trichome stalk, through which metabolites would be transported. Within the stalk’s
hypodermis, the phloem transport of sugars to the secretory cells would be via sieve
tubes. Although phloem transport requires metabolic energy phloem transport is
passive, but energy is needed to maintain the living condition of these sieve tubes
(Clifford, 2004). Respiration would also continue in the surrounding epidermal cells,
resulting in some reduction of the stain.
67
100µm 100µm 100µm
(a) (b) (c)

Figure 3.11. (a) Secretory cells stained red, within the glandular head, after thirty minutes in 1%
tetrazolium red; (b) A capitate stalked trichome with glandular head removed by slight abrasion.
After thirty minutes in 1% tetrazolium the disk of secretory cells is stained bright red; (c) A
mature capitate stalked trichome between two non-glandular cystolythic trichomes, viewed after
twelve hours in 0.1% tetrazolium solution (Potter, D..J.).
As the trichomes aged, it was common for the resin head to become detached from the
stalk. This would usually arise as a result of the entire head parting with the stalk (as
commenced in Figure 3.12a). In this photomicrograph the epidermal cells have pulled
away from the resin head, the hypodermal cells maintaining contact via a narrow
channel of stipe cells, which are still connected to the disk of secretory cells within the
resin head. In Figure 3.12b the resin head has completely detached and been
removed. The stipe cells are just seen protruding from the top of the stalk. Less
frequently, the resin head would detach as a result of a fissure apparently forming
above the secretory cell disk (Figure 3.13).
68
100µm
(a) (b)
Figure 3.12. (a) A capitate stalked trichome temporarily dry-mounted and viewed in transmitted
and incident light. The glandular head has become partly detached from the stalk to expose the
stipe cells, which connect the disk of secretory cells to the hypodermal cells within the stalk;
(b) The stalk of a capitate stalked trichome after detachment of the resin head. The specimen
was dry-mounted and illuminated with incident light. The stipe cells can just be seen protruding
from the top of the stalk (Potter, D..J.).
100µm
Figure 3.13. Separation of the glandular head during (left) and after (right) the appearance of a
fissure above the secretory cells. The example shown was observed on (M280). The specimen
was viewed in incident light (Potter, D..J.).
Figures 3.14a and 3.14b show an alternative view of a capitate stalked trichome before
and after detachment of the resin head. Many of the cells in this clone are naturally
pigmented red, some of which are on view within the secretory head (Figure 3.14a).
During secretory head detachment, the base of the secretory cell disk has remained
69
attached to the stalk, but the secretory cells have been detached along with remainder
of the secretory head (Figure 3.14b). The secretory head tissue remaining attached to
the stalk would appear to be formed of thickened resin-head cuticle. In an earlier
publication it was stated that the red structures within the secretory head were the
secretory cells (Potter, 2004). With the improved knowledge of trichome morphology
since gained, it would now appear that these flavonoid-pigmented structures were not
secretory cells but the vessels carrying carbohydrate and water to the secretory cells.
These appear embedded in and surrounded by a dense opaque material. A substantial
proportion of the weight of the resin head clearly has to be attributed to this rigid basal
material. It is quite likely that much of this material is made of cutin, suberin or both.
The lateral walls of glandular trichomes in many species become thickened with this
material, forming a so-called impermeable casparian strip. This prevents apoplastic
flow of secretory products down the trichome stalk (Werker, 2000).
100µm 50µm
(a) (b)
Figure 3.14. (a) An intact glandular stalked trichome of a naturally pigmented clone M186, with
coloured cells visible within the resin head. The sample was temporarily mounted in hemp oil;
(b) A direct overhead-view of the stalk of a capitate stalked trichome on clone M186 after resin
head detachment. The resin head has become detached leaving thr base of the secretory head
attached to the stalk. The sample was temporarily mounted in oil (Potter, D..J.).
Steady views of the outer surface of the base of more-mature secretory heads were
made possible by trapping the structures on the surface of adhesive tape. (Attempting
to remove resin heads from very immature capitate stalked trichomes usually ruptured
the thinner outer membrane, causing spillage of the contents.) The scars of
attachment to the four stipe cells were readily observed (Figure 3.15). This is
surrounded by the apparently-encrusted material of the secretory head outer wall. It
would appear from these studies that there are three methods of secretory head
70
detachment. In the vast majority of cases it would result from a split occurring in the
basal or stipe cells. Less commonly however, some of the secretory head could remain
attached to the stalk, and this may or may not include the secretory cells. Whether or
not the secretory cells are included in the collected material, this would not affect the
balance of the main secondary metabolites, as research suggests that the secretory
cells contain no cannabinoids (Mahlberg et al., 2004; Sirikantaramas, 2005; Taura et
al., 2007).
Scar of attachment
to stipe cells (four in
number).
Ring of heavily
thickened cell wall
tissue beneath
secretory cells.
Figure 3.15. A detached resin head (approximately 100 µm diameter) from a capitate stalked
trichome, viewed from below to gain a clear view of the scar where the stipe cells were
originally attached. The head had been trapped on the surface of clear adhesive tape
(Potter, D..J.).
On the female plant’s calyx, bracteoles, bracts and associated petioles capitate stalked
trichomes would commonly form a dense pubescence (Figure 3.16a), which would act
as a physical barrier to small phytophagous insects. By trapping a layer of air close to
the surface, it would also provide some protection against desiccating cold winds
(Mahlberg et al., 1984). By reflecting infra-red light a dense trichome pubescence has
cooling properties and, being equally effective across the complete light spectrum, it
also reflects ultra-violet (Roberecht and Caldwell, 1980). Phenolic resins like the
cannabinoids have also been shown to offer UV protection (Rhodes, 1977). This is
especially welcome in floral structures housing gametophytic tissues, which are
susceptible to damage by UV-B radiation (Caldwell et al., 1983). Although glandular
trichomes never exhibited a complete cover on bracts and bracteole tissue, UV
protection could never be complete, but any degree of protection would improve
survival chances.
71
Struggling insects were frequently found trapped to the resin heads of capitate stalked
trichomes, thereby inhibited from further feeding and reproduction (Figure 3.16b). This
defensive role of trichomes is observed in many other plant species, e.g. Alfalfa weevil
Empoasca fabae on lucerne Medicago sativa (Shade et al., 1975) and other insect
species on Pelargonium spp. (Harman et al.,. 1991).The most common victim observed
struggling on Cannabis was the cotton melon aphid Aphis gossypii. When attacked by
predators, Aphis gossypii emits an alarm pheromone to warn other others of danger
(Byers, 2005). It is possible that a trichome-ensnared aphid responds similarly. One of
the most common pests of cannabis – the tobacco thrip Thrips tobaci - would also
become trapped. It too is capable of emitting an alarm pheromone (Anathakrishnan
1993). If this theory is correct, the loss of a few trichomes to insects could discourage
a more extensive attack. Restricted allocation of capitate stalked trichomes to floral
tissue is widespread throughout the Plant Kingdom, where plants optimise investment
in defence by allocating secondary metabolites to tissues in direct proportion to their
value (Herms and Matson, 1992). It was notable that sessile trichomes played no part
in insect entrapment, suggesting that these had a different function. The cannabinoids
CBGA and THCA have been shown to cause apoptosis in insect cells, and it has been
suggested that this is an important defensive role for cannabinoids in capitate stalked
and sessile trichomes (Sirikantaramas et al., 2005).
(a) (b)
Figure 3.16 (a) A dense pubescence of glandular stalked trichomes on a bract within a cannabis
female inflorescence. The specimen was illuminated from behind and photographed with a
tripod-mounted camera incorporating a macro lens. The orange/brown structures are senesced
stigmas; (b) two young cotton-melon aphids Aphis gossypii. All six legs on each specimen are
irreversibly adhered to the resin heads of capitate stalked trichomes (Potter, D..J.).
72
3.5.1.6 Bulbous Trichomes
With a diameter of approximately 10-20 µm, these were the smallest of the glandular
trichomes (Figure 3.17a). First seen on the stem and the lower leaves, these were
widespread across the entire surface of the aerial part of the plant. Connected to the
epidermis by two cells (the top one much larger than the lower) these would produce a
simple spherical glandular head (Figure 3.17b) or a rarer complex multi-
compartmented glandular head (Figure 3.17c). Their function is not known. It was
notable that the CBG chemovar, shown in Figure 3.13, produced opaque white capitate
stalked and sessile trichome resin heads. This opacity is attributed to the CBGA. The
bulbous trichomes on this chemovar were clear or brown, as in Figure 3.16b. This
suggests that, in this genotype at least, the bulbous trichomes do not contain significant
concentrations of the cannabinoid.
100 µm 10 µm 10 µm
(a) (b) (c)

Figure 3.17 (a) A small bulbous trichome (left) alongside a fully developed glandular stalked
trichome. The contrast in resin head diameter (10 µm v 100 µm) is clear; (b) a simple bulbous
trichome and (c) a complex bulbous trichome. These are 10-15 µm in diameter. These samples
were temporarily dry-mounted and viewed in mixed transmitted and incident light
(Potter, D..J.).
3.5.1.7 Effect of age and storage on glandular trichome colour
Glandular trichome resin heads were observed to turn brown as the plant aged, as
previously reported by Turner et al. (1977) and Mahlberg et al. (1984). Bulbous
trichomes were the first to discolour, as seen in Figures 3.16b and c. In the glasshouse
this browning would rarely be seen on capitate stalked trichomes that were less than
five weeks old. The colouration continued during storage at ambient temperatures.
The clear colouration of freshly harvested floral material is contrasted with the brown
colour of a three-year old sample in Figures 3.18a and 3.18b. The width of the
73
glandular stalk in the dried sample has narrowed during drying due to the collapse of
the epidermal cells. This has also resulted on the stalks losing their turgid upright
appearance.
(a) (b)
Figure 3.18. (a) Clear glandular stalked trichomes on freshly harvested young cannabis floral
tissue; (b) Brown trichomes on three-year old stored cannabis (Potter, D..J.).
3.5.2. Effect of glandular trichome array on the secondary metabolite

content of plant tissues
The densities of capitate stalked and sessile trichomes counted on the proximal and
distal sections of bracts are shown in Figure 3.19.
3.5
3
Trichomes per sq mm
2.5
1.5
0.5
0
Proximal Stalked Proximal Sessile Distal Stalked Distal Sessile
Trichome Type and Distribution
Figure 3.19 The mean density of capitate stalked and sessile trichomes (± 1SE) (upper and
lower surfaces combined), on each of three high-THC cultivars M3, M6 and M7. On each clone
twenty randomly selected fields were counted on both the upper and lower surface in the
proximal and distal areas.
74
This confirmed that the bract dissection method had been successful in separating the
samples into proximal sections, where capitate stalked trichomes were abundant, and
distal sections where they were absent. Despite the polarisation of capitate stalked
trichome populations, the density of sessile trichomes was found to be similar in
proximal and distal areas. It was also noted that glandular stalked trichomes were in
the greatest numbers on the upper surface of the bract in this variety. Studies with
other varieties (not reported here) showed trichome density to be greatest on the
underside of the bract.
In all three clones analyses of variance showed that distal tissue had a significantly
higher proportion (p < 0.001) of CBC within the cannabinoid profile. Figure 3.20
showed that the capitate stalked trichomes in the proximal area outnumbered sessile
trichomes by a factor of approximately five. Individual capitate stalked trichome resin
heads have a volume approximately eight times greater than that of a sessile trichome.
Combining these factors, the potential difference in volume of the capitate stalked and
sessile populations is therefore a factor of forty. The findings of previous workers
suggest that the difference in quantity of cannabinoid stored could be even greater, as
capitate stalked trichome contents were also found to have a higher cannabinoid
concentration (Turner et al., 1977). This suggests that, on tissues that have abundant
capitate stalked trichomes, the sessile trichomes population would have minimal
influence on the overall cannabinoid potency and profile of that tissue.
25
CBC as % THC+CBC
20
15 Distal
10 Proximal
0
M3 M6 M7 Mean
*** *** *** _

Clone
Figure 3.20. The proportion of CBC expressed as % of THC+CBC in the cannabinoid profile of
proximal and distal tissue of bracts from each of three high-THC clones. (Error bars on clones
represent ± 1SD, and on the mean ± SE). In all three clones the difference was highly significant
(ANOVA, *** denotes p < 0.001).
75
Producers of illicit sinsemilla cannabis commonly recognise that distal bract tissue has
a much lower cannabinoid content. It is common practice to remove the distal part of
the less resinous bracts. The material that remains is consequently reduced in weight,
but its overall potency is increased. Illicit producers commonly refer to this process as
‘manicuring’ (Potter, 2004). The practice has also been followed by producers of
pharmaceutical grade medicinal cannabis in the Netherlands (Institute of Medical
Marijuana, 2008). In pharmacognosy, the removal of less desirable material from a
pharmaceutical feedstock is sometimes called ‘garbling’ (Tyler et al., 1988). This study
suggests that, where bracts only exhibit glandular stalked trichomes on the proximal
tissue, removal of the distal tissue will indeed increase the potency of the material that
remains. The cannabinoid profile of the remaining material will be minimally affected,
because of the relatively small influence of the associated sessile trichome population.
3.5.3 Effect of capitate stalked trichome density and colour on

cannabinoid content and profile.
The mean THC content was calculated for all sinsemilla samples attracting the same
score for trichome density. The vast majority of samples attracted scores on the first
five points of this 1-9 scale (76% being awarded a score of 2 or 3) and meaningful
standard error calculations were only possible on values in this range (Figure 3.21).
18
16
14
12
%THC
10
8
6
4
2
0
1 2 3 4 5 6 7 8 9
14 64 126 32 6 3 2 0 2
1-9 Score
n
Figure 3.21 The correlation between capitate stalked trichome density (visually assessed 1-9
scale, 1 = high, 9 = low) and the overall THC content of the sample. Values shown are the
mean % w/w THC content (± SE only where n > 5) of all samples for each density score.
Regression line is shown in red. The regression model is:- % THC = 18.051 – 1.629 * Density
Score. (p < 0.001, R2 = 0.17).
76
The results show that there is clearly a significant tendency for plant samples with
greater capitate stalked trichome densities to have a higher cannabinoid content. In
plants with a less dense pubescence, high cannabinoid contents are not simply
maintained by more secondary metabolite being sequestered in each trichome. A high
density of glandular trichomes is therefore a very useful visual parameter upon which a
judgement can be made of the cannabinoid content of dry plant material. However, the
low R² value shows that this is only an approximate guide when assessing samples of
variable genotype and provenance.
The relationship between trichome colour and cannabinoid content is shown in Figure
3.22. The data were more evenly spread than those from the corresponding trichome
density study. The slope of the regression line shows that there is a weak but
significant tendency for darker coloured trichomes to be associated with low-potency
cannabis samples. Turner et al. (1977), comparing the THC content of colourless and
brown capitate stalked trichomes of a single clone, reported that brown trichomes
contained less THC per trichome. However, there were no accompanying statistical
analyses to clarify the significance of the reported finding.
20
18
16
14
12
% THC
10
8
6
4
2
0
1 2 3 4 5 6 7 8 9
8 26 35 52 51 49 20 6 1
1-9 Score
n
Figure 3.22. The correlation between capitate stalked trichome colour (visually assessed using a
1-9 scale, 1 = clear, 9 = darkest brown) and the overall THC content of the sample. Values
shown are the mean % w/w THC content (±SE) of all samples for each colour score.
Regression line is shown in red. The model for this is: - Percent THC = 15.990 – 0.587 *
Trichome Colour Score (p < 0.001, R2 = 0.053).
77
The pigment causing this brown colouration is not known. Commencement of browning
was commonly observed in the region around the secretory cells and could be due to
oxygen entering the resin head via the scar tissue (Figure 3.13) when it becomes
detached from the epidermal cells on the trichome stalk, as shown in Figure 3.12a.
Alternatively it could be caused by the biproducts of catabolism of the secretory cells
contents.
The mean average CBN content of the samples awarded each score is shown in
Figure 3.23. For those samples where CBN contents were very low and below the
minimum detectable level of 0.10% a nominal value of 0.05% was used.
3
2.5
y = 0.0957x
CBN % w/w
2
R2 = 0.5616
1.5
1
0.5
0
1 2 3 4 5 6 7 8 9
8 26 35 52 51 49 20 6 1
Trichome Colour 1-9 Score
n
Figure 3.23. The mean CBN content (± SD) of populations of sinsemilla samples awarded each
of the 1-9 ratings for trichome colour 1 = clear, 9 = darkest brown. All samples were seized by
police in 2004/2005.
There was a trend for samples with darker trichomes to have a higher CBN content but
the data was extremely variable, as shown by the large standard deviations. The
degree of THC catabolism was assessed by expressing the CBN as a proportion of the
CBN+THC total for each sample (i.e. CBN/(THC+CBN). The correlation between
trichome colour and CBN/(THC+CBN) is shown as a scatter-graph in Figure 3.24. This
shows that very little CBN existed in clear or white trichomes. A high level of CBN
formation was restricted to plants with brown glandular trichomes, but dark coloured
trichomes could still be devoid of this catabolite. Six values were identified as clear
78
outliers, according to the ISO recommended Grubb’s Test (Miller and Miller, 2005) and
excluded.
30
25
CBN as % THC+CBN
20
15
10
0
1 2 3 4 5 6 7 8 9
Trichome Colour
Figure 3.24. The variability in degree of THC catabolism to CBN as related to their colour. Of
249 original samples 6 were rejected as outliers and not included here. Trichome colour was
assessed visually and scored on a 1-9 scale where 1 represents totally clear and successively
higher scores denote an increasing opacity and darkening in colour.
The comparison between trichome colour and CBN content showed that the lower
average potency in darker colour trichomes could be almost entirely due the THC
having catabolised in these samples. There was a greater likelihood of finding CBN in
brown coloured trichomes but its presence was not guaranteed. This shows that if a
pharmaceutical company demands cannabis material containing minimal levels is
CBN, a brown colouration does not immediately imply rejection of the sample.
It appears that the conditions required for the oxidative catabolism of THC to CBN are
not the same as those causing the brown colouration. Experienced observation of
many samples has shown that this pigmentation is not reversible and the colour
stabilises. Indeed, this colour was seen to have been maintained in trichomes
photographed while contributing to studies on an archaeological cannabis sample
discovered in an ancient Chinese tomb. Carbon dating estimates the age of the
sample in Figure 3.25 as 2700 years old. It is accepted that additional catabolic
reactions would have occurred over time and possibly had an influence on the colour,
79
but catabolism appeared to have stabilised and cannabinoids were still present in these
trichomes (Russo et al., 2008).
100µm
Figure 3.25. Aged brown sessile glandular trichomes on 2700 year old cannabis. The sample is
illuminated with incident light and photographed digitally. The pubescence of unicellular non-
glandular trichomes is also clearly visible (Potter, D..J.).

biosynthesis in sessile trichomes on variegated leaf tissue.
The potency of the leaf tissue samples in these tests is shown in Table 3.4. On a dry
weight basis, yellow leaf material was clearly more potent than green. This would have
been influenced by the fact that green tissue weighed 33% more per unit area,
probably due to the production and storage of starch in these tissues following
photosynthesis. By assessing potency in terms of cannabinoid per unit area, the effect
of starch accumulation would be removed. Assessed this way, the yellow material was
still significantly more potent (p < 0.05), although the difference was much smaller.
Microscope observations suggested that this difference was not due to differences in
densities of sessile trichomes on yellow and green tissue, but an accurate count was
not possible because of the density of non-glandular trichomes obscuring the view.
The earlier work by Crombie (1977) had shown that cannabinoid biosynthesis still
continued in ‘albino’ tissue. This study went further and suggested that cannabinoid
biosynthesis was totally unrelated to the immediate tissue’s ability to photosynthesise.
However, the rate of biosynthesis in these tissues would be affected by the overall
photosynthetic-ability of the plant.
80
The biosynthesis of terpenoids has a higher energy requirement than most other
primary and secondary metabolites, because of the extensive level of chemical
reduction within these compounds (Gershenzon, 1994). The energy for this
biosynthesis would be derived from light energy, captured within the chloroplast.
During periods of light exposure, some chloroplasts would be directly illuminated and
others would be in varying degrees of shade. This study showed that plant tissues
devoid of chlorophyll, and thereby unable to photosynthesise, could support the
biosynthesis of cannabinoids within their own glandular trichomes. The carbon source
required for cannabinoid biosynthesis would have been produced elsewhere within the
plant and then translocated to these trichomes. This demonstrates that in a normal
growing environment, where some parts of the plant are in full sun and others in
varying degrees of shade, all aerial parts the plant will be able to synthesise
cannabinoids. The resultant increase in trichome content uniformity is fortunate for the
grower of cannabis for the pharmaceutical industry.
ANOVA
Green Tissue Yellow Tissue p value
Potency by Weight Test 1. THC % w/w
dry (n = 3 bulked samples) 1.51 (± 0.26) 2.30 (± 0.26) 0.008
Potency by Weight Test 2. THC % w/w
dry (n = 2 bulked samples) 1.18 (± 0.08) 1.92 (± 0.08) 0.005
Potency by Area Test 2. THC g m-2
(n = 2 bulked samples) 0.50 (± 0.03) 0.61 (± 0.01) 0.038
Table 3.4. The potency (THC content) of yellow and green leaf tissue of the variegated cultivar
G60-M55 assessed in each of two tests (± SD). The potency in the second test is also shown as
a weight of THC per unit area.
.
Mahlberg et al. (1983) compared the cannabinoid content of capitate stalked trichomes
on the upper and lower surfaces of cannabis bracts and found those on the upper
surface to be more potent. This was attributed to the upper surface receiving more
light. This study with variegated cannabis weakens that argument, showing that the
ability of leaf tissue to photosynthesise has minimal effect on its ability of the local
glandular trichomes to synthesise cannabinoids.
81
3.6 CONCLUSIONS
While performing a detailed study of cannabis trichomes, Dayanandan and Kaufman
(1976) stated that the study of these organelles was essential to understand the
biogenesis, distribution and function of the different cannabinoids. The research
reported in this chapter clearly showed how the distribution of THC and CBC differed.
The subcellular details of cannabinoid secretion were also shown in greater detail than
perhaps previously captured with light microscopes.
Within the pubescence on Cannabis sativa L there are six forms of trichome, of which
only five are found on the female plant. With appropriate mounting and staining, it was
shown that what was believed to be a previously unavailable undistorted views of the
internal structures of these trichomes could be differentiated to a high degree with the
use of low cost microscopes. The use of these mounting media could facilitate future
further studies of these structures.
It was shown that the cannabinoid profile of bract tissue appears to be affected by the
ratio of so called ‘capitate stalked’ and ‘sessile’ trichomes present. This suggests that
the two types of trichome had differing cannabinoid profiles. By separating tissues with
differing ratios of capitate stalked and sessile trichomes, it was shown that the grower
had the capability to produce samples with significantly different cannabinoid profiles.
The research also showed that insects were regularly ensnared by glandular stalked
trichomes but never by capitate sessile trichomes. This was a clear indication that
these two trichome forms differed in function. It was hypothesised that this was
possibly attributable to a difference in secondary metabolite content. The observed
contrasting sizes and densities of glandular stalked and sessile trichomes suggested
that in floral material, exhibiting both trichome types, the capitate stalked trichome
population had a potentially greater influence on cannabinoid profile than the sessile
trichome population.
Vital staining techniques provided supportive evidence that that the secretory cells
within the trichome head were the area of greatest metabolic activity. Novel studies
with variegated cannabis also showed that secretory cells within leaf tissue lacking
chlorophyll suffered no reduction in cannabinoid synthesis ability. In fact a small but
significantly higher concentration of cannabinoid (weight per unit area) was found in
tissue lacking chlorophyll. Although the overall ability of the plant to biosynthesise
cannabinoids was likely related to the total amount of energy available to the plant, this
study implied that where part of a plant was in shade, this did not affect its ability to
82
biosynthesise cannabinoids ability. Thus partial shading did not increase the variability
of plant tissue cannabinoid contents.
As glandular trichomes age it was observed that there is a colour change, the initially
clear resin head first turning white and then brown. On feedstock exhibiting brown
trichomes there is an increased likelihood, but not a guarantee, that there will be a
reduced THC content and a raised content of the THC catabolite CBN. A high density
of capitate stalked trichomes on the female floral tissue indicates an increased
probability, but not a guarantee of high cannabinoid content.
As the capitate stalked form matures, the resin head is increasing likely to detach itself
from its stalk. When this occurs, the secretory cells are usually retained within the
resin head, but they are sometimes left attached to the stalk. This mode of
detachment would be expected to affect the proportion of waxes, proteins and other
unidentified ingredients within any enriched trichome preparations made.
Having assessed the form and distribution of the trichomes, Chapter Four looks in
detail at the chemical content of the capitate types. It also assesses how this alters as
trichomes age. The knowledge gained would help the grower identify the optimum time
for harvest and evaluate the possibly of producing feedstocks by removing these
trichomes from the plant.
83
Chapter 4. The Function and Exploitation of Secondary Metabolites from Glandular Trichomes of Cannabis sativa L.
Chapter 4. The Function and Exploitation of Secondary

Metabolites from Glandular Trichomes of Cannabis sativa L.
4.1 INTRODUCTION
Across the Plant Kingdom, glandular trichomes are responsible for the biosynthesis
and/or sequestration of a vast range of secondary metabolites. Of these the terpenoids,
which includes the cannabinoids and cannabis-related terpenes, are the most common
and structurally diverse compound group, and these are spread throughout many plant
families (Kelsey et al., 1984). Other carbon, hydrogen and oxygen based categories of
secondary metabolites are also found, but are restricted to the glandular trichomes of
smaller ranges of species. Examples include the methylated flavonoids of the creosote
bush Larrea tridentata (Thompson et al., 1979) and the quinones found in the
semitropical plant family Hydrophyllaceae (Kelsey，1984).
The total number of secondary metabolites found in plants has been estimated at
approximately thirty thousand. Of these, about 40% contain nitrogen, of which all but a
thousand belong to the alkaloid group (Acamovic and Brooker, 2005). Of the secondary
metabolites found in glandular trichomes, the proportion that contains nitrogen is much
less than 40% (Wagner et al., 2004). Examples include histamine, which is synthesised
within the stinging glandular trichomes of nettles Urtica spp. Another example of great
importance to world health is the alkaloid nicotine, which is abundant in trichomes of
some cultivated varieties of tobacco Nicotiana tabacum. Although sequestered in
trichomes, this alkaloid is actually translocated there from the roots where it is
biosynthesised. Trichomes of the velvet bean Mucuna pruriens also contain alkaloids
and additionally produce serotonin (5-hydroxytryptamine) (Ghosals Singh et al., 1971).
The previous chapter described the different forms of non-glandular and glandular
trichomes in Cannabis sativa L and compared their differing functions. It also reported
studies which supported previous researchers’ findings (Turner et al., 1977, Mahlberg
et al., 1984) that capitate stalked and sessile trichomes differed in their cannabinoid
profile. It was shown that by segregating plant tissues, according to their trichome
pubescence characteristics, it was possible to produce feed-stocks with significantly
different cannabinoid profiles. This chapter looks in greater detail at the secondary
metabolite content of these trichomes. It was hypothesised that, if the sessile and
capitate stalked trichomes could be removed from the plant and then separated, it
might be possible to provide feedstocks with more closely controlled cannabinoid
84
profiles. Separating trichomes that were at different stages of maturation may also
have benefits.
The removal of glandular trichomes from living Cannabis plants is common practise,
being a major part of cannabis resin manufacture. Many centuries-old cultural methods
are practised, but few are of relevance to the modern pharmaceutical industry. One
method of resin manufacture includes manual rubbing of the inflorescences, resulting
in the trichomes adhering to the hands (Cerniak, 1985). British physician W.B.
O’Shaughnessy, who introduced cannabis to Western medicine from India, reported
that workers dressed in leather would walk through the crop and then scrape off the
resin that adhered to their clothing (Samuelsson, 1999). Both techniques rupture the
trichomes, accelerating the loss of the more volatile components and the oxidation of
others. The materials also become heavily contaminated with other plant fragments.
More common methods of trichome removal for resin production are based upon the
sieving of dried cannabis plants. This practise is utilised in Morocco, which supplies the
vast majority of cannabis resin entering the UK (UNODC, 2006). Plants are dried and
sometimes left for many weeks before processing, by which time the trichomes will
have lost much of their more volatile terpene content. Techniques used to make so-
called ‘modern hashish’ (Clarke and Watson, 2007) perhaps offer greater opportunities
for collecting undamaged glandular trichomes, with more of their secondary
metabolites intact. In one method, cannabis material is simply agitated in cold water
and the dislodged capitate stalked trichomes collected using sieves. Jansen and Terris
(2002) reported that, with a Dutch Government subsidy, this technique had been
adapted to make hashish for pharmaceutical research purposes. Separation of
capitate stalked and sessile trichomes was not reported. For the remainder of this
thesis, the terms hashish and resin are avoided when referring to products made for
scientific as opposed to recreational purposes. The term enriched trichome preparation
(ETP) is preferred.
As well as the cannabinoids, most of the monoterpenes and sesquiterpenes found in

Cannabis are also located in the glandular trichomes (Malingré, et al., 1975; Turner et
al., 1980). These are generally present in much smaller quantities than the
cannabinoids, and this hinders their detailed study. Studies of the terpene content of
cannabis trichomes are few and have involved the steam distillation of intact cannabis
tissue as the source material (Mediavilla and Steinemann, 1997). The bulk collection of
intact cannabis trichomes would clearly facilitate more detailed terpenoid analysis by
providing a richer source of material for analysis. To enable this to happen, research
was needed to assess the feasibility of removing and then segregating large quantities
85
of intact glandular trichomes from cannabis. A number of techniques have been

employed to remove intact glandular trichomes from other species. These have been
reviewed by Wagner et al. (2004). Techniques included removal of individual trichomes
with forceps, shaking in aqueous solution with an abrasive (e.g. sand, fine glass beads
or powdered dry ice) and the gentle brushing of fresh or frozen tissue. In some cases
different trichome forms have subsequently been separated by sieving and/or
centrifugation, e.g. separation of bulbous and peltate trichomes of sweet basil Ocimum
basilicum (Gang et al., 2001). All these collection methods produce extremely small
trichome samples. In contrast, removal of glandular trichomes on an industrial scale is
currently performed in the hop (Humulus lupulus) industry. By agitating plant material
in extreme cold temperatures, the sessile glandular trichomes (known commercially as
lupulin) are dislodged and separated from the plant (Rigby, 2000).
The previous chapter confirmed the findings (Ledbetter and Krikorian, 1975) that the
four types of glandular trichome found in Cannabis, namely bulbous, capitate sessile,
antherial sessile and capitate stalked, had diameters of approximately 10 µm, 50 µm,
80 µm and 100 µm respectively. Based on their comparative spherical dimensions a
bulbous trichome would have a potential secondary metabolite content approximately
one hundred and twenty five times less than that of a capitate sessile trichome.
Previous workers have expressed difficulties in working with bulbous trichomes
because of their small size, and evidence for the existence of cannabinoids in these
trichomes is lacking. Regular observations of bulbous trichome populations, while
studying other trichome forms for this thesis, suggested that the population densities of
bulbous trichomes were similar to those of sessile trichomes. This supported the
observation of previous researchers (Turner et al.,1978; Mahlberg et al., 1984). This,
combined with the minute size, suggests that bulbous trichome contribute little to the
overall secondary metabolite content of cannabis. Bulbous trichomes were not studied
further for this thesis.

A series of studies were performed with the aim of separating, purifying and analysing
substantial quantities (>1g) of intact glandular trichomes from cannabis. It was
envisaged that trichome filtrates derived from cannabis tissues of differing chemotype,
stage of maturity and location on the plant might exhibit different secondary metabolite
profiles. As progress was made the following sequence of objectives was developed: -
4.2.1. To collect and separate intact capitate stalked and sessile glandular trichomes
from fresh floral material.
86
4.2.2. To isolate intact sessile glandular trichomes from vegetative material.
4.2.3. To collect sessile trichomes from foliage of a high-CBC chemotype as a means

of isolating the minor cannabinoid CBC
4.2.4. To examine the ontogenetic changes in terpenoid content of glandular trichomes

during plant maturation.
4.3 MATERIALS
4.3.1 Germplasm
Clone Chemotype Variety Name Supplier
G1-M3 High-THC Guinevere GW Pharmaceuticals Ltd,

G2-M6 High-THC Galina Porton Down Science
Park, Salisbury, Wiltshire.
G5-M13 High-CBD Grace
G5-M16 High-CBD Gill
M240 High-CBC Unnamed
Table 4.1 Name, source and chemotype of clones used.
4.3.2 Apparatus
Kenwood HM 220 Food Mixer Dixons Ltd. Internet purchase.
25 µm, 43 µm, 73 µm and EveryoneDoesIT Unit D1, Phoenix Industrial Estate,

220 µm ‘Bubblebag’ Sieves Rosslyn Crescent, Harrow, HA1 2SP
Ebb and Flood Irrigation Tank Bridge Greenhouses Ltd., Chalk Lane,
2.0 x 0.7 x 0.4 m Sidlesham, Chichester, PO20 7NQ
Table 4.2 Miscellaneous items and commercial sources.
4.4 METHODS
4.4.1 Separation of sessile and capitate stalked trichomes glandular heads

from mature fresh floral material.
In the first of a series of investigations, this test adapted a method of removing
trichomes from cannabis material described by Jansen and Terris (2002). The high-
THC clone G1 M3 was used for this study. Approximately 500 g of fresh floral material
was removed from a batch of fully mature plants. These had been in flower for eight
87
weeks and were at the normal stage for harvesting. Using scissors, the material was
cut into pieces, each being no greater than approximately five grams. Six randomly
selected pieces were retained for analysis. The material was then plunged immediately
into twenty five litres of slurry containing tap water and crushed ice. The mixture was
gently mixed for ten minutes to allow the plant material to cool and then agitated for ten
minutes using a Kenwood HM220 domestic food-mixer operating at maximum speed.
Manual examination confirmed that the glandular trichomes lost their sticky texture at
low temperatures. A large proportion of the trichome glandular resin heads were
dislodged from the plant by this process. Being denser than water they sank and
readily separated from the pulp when poured through a fine sieve (220 µm approximate
mesh). The resin heads passed through the mesh and the ‘spent pulp’ was retained.
The resin heads are then efficiently separated from the bulk of the water by pouring the
suspension through a 73 µm sieve and then a finer 25 µm sieve. In theory, the resin
heads from glandular stalked trichomes (reported typical diameter 75-100 µm) should
have been trapped on the 73 µm sieve, sessile trichomes (typically 50 µm) falling
through and being caught on the 25 µm mesh.
The resin heads collected from each sieve were removed and a few milligrams taken
for microscope study. The remainder was frozen while awaiting chemical analysis.
4.4.2 Bulk-Production of Pure Sessile Trichome Preparations.

Five kilograms of leaves were removed from plants of the high-THC clone G1 M3. The
plants had been grown for three weeks in continuous lighting (75 Wm-2 PAR) at 25ºC,
and were therefore still in the vegetative phase and devoid of capitate stalked
trichomes. The collected material was thoroughly mixed and approximately twenty
grams frozen (-20ºC) pending analysis by GC (Appendix 2). The 5 kg of foliage was
thoroughly mixed in a shallow tank with 200 litres of iced water for thirty minutes using
a domestic food mixer at maximum speed. The bulk of the treated foliage was then
manually removed and the liquid remains drained through 220 µm and 25 µm sieves.
The material collected on the second sieve was examined for purity, using a
microscope, and a sub-sample (approximately 1 g) taken for cannabinoid assay. The
remaining material was frozen prior to steam distillation and qualitative chemical
analysis by GC at Botanix Ltd. A batch of plants (≈ 300) of the same clone was also
grown to full maturity, then harvested and dried for routine drug production. Six random
subsamples (approximately 5 g) of the stripped leaf and foliar material were thoroughly
mixed prior to analysis by GC (Appendix 1).
88
4.4.3 Production of a cannabichromene-rich sessile trichome preparation.

The production of a CBC-rich sessile trichome preparation was attempted, using the
above trichome collection method. The plant material utilized was stripped foliage of
the CBC-rich clone M240. These plants had been grown vegetatively for four weeks in
the same regime.
4.4.4 Ontogenetic changes in Secondary Metabolite Content of Glandular

Trichome Contents.
Two cultivars were selected for this study; high-THC clone G2 M6, and high-CBD clone
G5 M13. Plants were grown according to the routine methods for the propagation of
the medicinal cannabis crop. Ten randomly selected plants of each cultivar were
removed from the crop after three weeks of vegetative growth. Further samples were
made on each of five harvest dates - six, seven, eight, nine and ten weeks after the
induction of flowering. The foliage and floral material from each batch was stripped
from the stem. This was thoroughly mixed and six subsamples (approximately 10 g
each) were taken. All materials were frozen at -20ºC pending analysis. Stems were
discarded. The six subsamples were subsequently analysed for cannabinoid content by
GW Pharmaceutical Ltd using GC (Appendix 1).
The main samples were distilled by Botanix Limited according to their Standard
Operating Procedure for the Determination of Volatile Hop Oil (Institute of Brewing
Method). Each was ladled into a one-litre round bottom flask with a B55 neck. A few
anti-bumping granules were added before connecting each one to a ‘British
Pharmacopoeia Still’ using a B55/34 adaptor. The flasks were then heated using a
heating mantle and the contents distilled for three hours. Employing this method,
Howard (1970) showed that the extraction of essential oils from hop lupulin (glandular
trichomes) was reliably complete in three hours and indeed no further distillate was
seen to be collected after this time. During this period the flow of condensate was
controlled to cause minimum disturbance to the oil in the trap. At the end of three
hours the volume of oil was recorded and decanted. These oil fractions were analysed
for qualitative terpene content by Botanix Ltd. The quantitative content of a small range
of terpenes within this mixture was determined by GC at GW Pharmaceuticals Ltd
(Appendix 2).
89

Heads from mature fresh floral material
Table 4.3 shows the cannabinoid content of single trichome preparations collected from
the fine and coarser sieves and a bulked sample of plant material before and after
trichome collection. Three sub-samples of each material were taken for analysis. The
standard deviations therefore only indicate the uniformity of the preparations and the
precision of the analytical method.
% of each cannabinoid in the cannabinoid

total ± SD % w/w total
Sample Details cannabinoid
CBC CBG THC ± SD
Sessile trichome rich

1.44 ± 0.02 1.19 ± 0.04 97.37 ± 0.05 37.00 ± 7.26
filtrate - 25 µm sieve
Capitate stalked trichome

0.93 ± 0.02 0.86 ± 0.03 98.21 ± 0.02 58.70 ± 4.29
rich filtrate - 73 µm sieve
Starter Material 1.13 ± 0.14 1.30 ± 0.05 97.58 ± 0.16 11.50 ± 0.34
Spent Pulp 1.07 ± 0.18 1.21 ± 0.03 97.72 ± 0.19 7.30 ± 1.53
Table 4.3. A comparison of the cannabinoid profile and content of fresh cannabis floral
material (clone G1 M3) and sieved trichome filtrates made there from. One sample of each
fraction was produced. Analyses show mean analytical results of three subsamples (±SD). Also
shown is the cannabinoid content of the floral material before trichome extraction and the spent
pulp after extraction (n=3).
Microscope analysis confirmed that most of the material caught on the 73 µm sieve
consisted of the larger resin heads from capitate stalked trichomes. The material
caught on the 25 µm sieve appeared to consist predominantly of the smaller sessile
trichome resin heads. However, the resin heads from some immature capitate stalked
trichomes may have also passed through the 73 µm sieve and been included here.
The material collected on the finer 25 µm sieve had a much lower cannabinoid content
(37% w/w) than that removed from the coarser 73 µm sieve (59% w/w). This supports
previous observations in a previous study that showed the cannabinoid concentration
in individually collected sessile trichomes to be lower than that of the capitate stalked
form (Turner et al., 1978, 1979).
90
The sessile trichome rich filtrate on the finer sieve had a notably higher proportion of
the cannabinoid CBC. The result supports the finding of the previous chapter,
suggesting that the two trichome forms appear to have different cannabinoid profiles.
However, in that earlier study, the comparison was made between capitate stalked and
sessile trichome populations from individual bracts, and the two trichome forms were of
the same age. In this later study with sieved trichome samples, the average age of the
sessile trichomes would have been greater than that of the capitate stalked trichomes.
This is because the sessile form would have been produced throughout the lifetime of
the plant, whereas the bulk of the capitate stalked trichomes would have been formed
within the last few weeks of growth. It has been shown that, in most chemotypes, the
proportion of CBC within the cannabinoid profile of the whole plant drops sharply during
the life-time of the plant (Shoyama et al., 1975; de Meijer et al., 2009). The drop is
greatest during the first weeks of growth when no capitate stalked trichomes are
present, and the decrease in CBC proportion thus cannot be attributed to a changing
balance of sessile and capitate stalked trichomes. The higher proportion of CBC in the
filtrate on the finer sieve would have thus likely been due to two factors: -
i) It contained more sessile trichomes, which have been shown to have a higher
proportion of CBC to capitate stalked trichomes of the same age, and
ii) The trichome on this sieve contained a higher proportion developed within the first
weeks of growth when the proportion of CBC within the whole plant was much higher.
Table 4.3 also shows that there was a higher proportion of CBG in the cannabinoid
profile of the material collected on the finer sieve. It is not possible to state if this is due
to direct differences in the cannabinoid profile of sessile and capitate stalked trichomes,
or due (at least partly) to the younger average age of the capitate stalked trichome-rich
material. (Younger material would reasonably be expected to contain a higher
proportion of trichomes within which less of the GPP and olivetolic acid had been fully
converted into THC via the intermediate in question – CBG.) In the previous chapter,
where the cannabinoid profiles of similar aged trichome populations on proximal and
distal bract tissue were compared, CBG was not present in detectable quantities in all
samples. A comparison with that test is therefore not possible. Even though the
reasons were not totally clear, from this test it was important to note that separating
trichomes with sieves appeared to offer a means of producing phytopharmaceutical
feedstocks with a favorably altered cannabinoid profiles.
91
4.5.2 Isolation of intact sessile glandular trichomes from vegetative

material
This test clearly demonstrated that intact sessile trichome resin heads could be readily
collected from non-flowering cannabis vegetation. This, and the previous test,
overcame concerns that the technique might only be efficient on capitate stalked
trichomes which, by virtue of their greater size and upright shape, would experience
greater shearing forces from the agitated iced-water. A sub-sample of this preparation
is shown in Figure 4.1.
Figure 4.1. A smeared sample of sessile glandular trichome resin heads prepared from
vegetative foliage of high-THC clone G1 M3. The freshly captures specimens were collected
from the surface of a 25 µm sieve and are free-floating in water. A few pieces of leaf fragment
are also present as a minor contaminant.
It can be seen that there was some minor contamination of this material, with
fragments of green leaf tissue. The proportion of resin heads within this filtrate was
higher than that generally produced when collecting resin heads from capitate stalked
trichomes. The latter would be routinely contaminated with capitate glandular trichome
stalks (which are not present on foliage) and cystolythic trichomes (which more readily
pass through the coarser sieve). The cannabinoid content of materials collected from
both the 25 µm and 45 µm sieves were analysed. In Table 4.4, this is compared to the
cannabinoid content of the combined foliar and floral material collected from mature
plants of the same cultivar. Although THC was still the dominant cannabinoid in the
sieve contents, it can be seen that the proportion of CBC within the cannabinoid profile
was greatly increased. The results show the content of trichome preparations collected
from 45 µm and 25 µm sieves. For comparison, the cannabinoid content is shown in
92
aerial material stripped from the stems of 3 week and 11 week old plants of the same
cultivar. Three replicates of each sample were analysed.
Material analysed
CBC as % of
(n = three analytical samples from CBC % w/w THC % w/w
THC+CBC
a single batch) Mean ± sd Mean ± sd
Mean ± sd
25 µm sieve contents
4.25 ± 0.74 22.70 ± 4.36 14.98 ± 0.19
From 3 week old vegetative plants
45 µm sieve contents
6.37 ± 0.31 35.31 ± 2.44 14.92 ± 0.23
From 3 week old vegetative plants
Foliage –
0.08 ± 0.03 0.67 ± 0.16 10.43 ± 0.89
3 week old vegetative plants
Foliage and floral material –
0.13 ± 0.01 8.61 ± 0.45 1.48 ± 0.47
11 week old plants
Table 4.4. The relative proportions of CBC and THC in sessile trichome rich preparations made
by sieving dislodged trichomes from vegetative foliage of high-THC clone G1 M3. Three
subsamples of each were analysed by GC.
The CBC potency of the preparation collected on the 45 µm sieve (6.37 % w/w) was
approximately eighty times greater than that of the plant material from which it was
derived (0.08 % w/w). However, the preparation collected on the 25 µm sieve was less
potent (4.25 % w/w). The reason for this was not known. The smaller trichomes would
have a greater surface area to volume ratio and it is possible that these contained
proportionally more outer membrane and vesicle-associated fibrillar material. Despite
the high proportions of THC remaining in these preparations, the results indicated that
the trichome separation technique does enable samples to be produced which have a
favourably altered cannabinoid profile. It is notable that the proportion of CBC within
the cannabinoid profile of the separated trichomes (approximately 15% w/w) was
greater than that of the foliage (approximately 10% w/w). The proportion fell further as
the plant matured, repeating the observation reported by de Meijer et al., (2008).
The reason for the observed difference in proportion of CBC in the trichomes and the
foliage from which they were collected is not clear. Two possibilities are suggested: -
1) Although the foliage was collected from three week old plants, many of the
trichomes residing there would have formed earlier during the plants development.
93
The cannabinoid profile of these trichomes would reflect the proportionally higher CBC
expected in vegetation that was less than three weeks old.
2) In the three week old foliage there would be many new sessile trichomes forming.
Following the well reported ontogenetic pattern, these would have a lower proportion of
CBC than their predecessors. It is possible that these were not sufficiently resilient to
be dislodged intact during agitation in iced water. The filtrate would therefore be
skewed towards more mature trichomes.
4.5.3 The collection of sessile trichomes from foliage of a high CBC

chemotype as a means of isolating the minor cannabinoid CBC
The preparation of sessile trichomes collected from vegetation of clone M240 was dried
overnight. The material proved an extremely rich source of CBC – 44% w/w. The
sample also exhibited a high level of purity. As found in the previous test, the fall in the
proportion of CBC in the cannabinoid profile of the enriched trichome preparation
appeared to lag behind that of the plant material from which it was collected. The
foliage, in which CBC accounted for 61% of the total cannabinoids, had produced a
trichome filtrate within which CBC accounted for 94% of the cannabinoid total (Table
4.5). Patent protection of this procedure, as a means of producing purer sources of
CBC, was applied for (GB 0806553.4.). This material was subsequently further
processed to produce a CBC sample exceeding 99% purity levels, which was
submitted for in-vitro pharmacological evaluation.
Cannabinoid % w/w mean ± SD

Sample % Purity
CBC CBCV CBG CBD
analysed CBC ± SD
Trichomes 44.37 ± 4.95 0.57 ± 0.07 0.33 ± 0.04 0.60 ± 0.06 94.2 ± 0.3
Foliage 1.36 ± 0.04 - 0.62 ± 0.02 0.25 ± 0.01 61.0 ± 0.9
Table 4.5. The proportions of principal cannabinoids (±SD) found in a trichome-rich filtrate
clone M240, containing only sessile trichomes. One bulk sample was prepared and four sub-
samples analysed. Also shown is the purity of the CBC (expressed as a %w/w of total
cannabinoids detected) within the foliage from which these trichomes were collected.
4.5.4 Ontogenetic changes in glandular trichome secondary metabolite

content
The changing terpene profile of glandular trichomes collected from plants at a range of
developmental stages is shown in Tables 4.6 (THC chemovar) and 4.7 (CBD
chemovar).
94
In the high-THC plants harvested 6 to 10 weeks after being placed in short daylength,
the monoterpenes accounted for between 60%-67% of the monoterpene/sesquiterpene
total (when assessed by peak area). There was no clear trend against time. The
trichomes from the young foliage were notable for having a much lower proportion of
monoterpenes (26%) than those from the flowering material. However, it can be seen
that not all monoterpenes showed the same contrasting pattern. While the proportion of
α-pinene, β-pinene and limonene is similar in both, the proportion of myrcene in the
profile of foliar sample is much lower. The reason for this is not known. However, it is
important to note, that in some cases, a range of terpenes can be synthesised with the
involvement of a single enzyme (Croteau and Johnson, 1984). As a consequence, the
ratio of the terpenes produced always maintains a fixed ratio. α-pinene, β-pinene and
limonene, may be linked in this way while myrcene appears not to be linked to other
terpenes.
95
Raw material used Foliar Foliar and Floral Mixture

Trichome type Sessile Predominantly Stalked
Wks after planting 3 9 10 11 12 13
Wks in 12 hr days 0 6 7 8 9 10
Monoterpenes % Peak Area

α-pinene 0.9 0.7 0.8 0.8 0.9 1.0
β-pinene 1.7 1.5 1.7 1.9 2.1 2.2
Myrcene 14.8 47.7 51.8 46.8 44.8 44.7
Limonene 8.5 8.4 9.3 9.1 9.2 9.4
Linalol - 2.5 3.0 3.0 3.1 3.2
Sesquiterpenes % Peak Area
trans caryophyllene 21.1 18.1 14.8 17.4 17.5 17.5
trans α
- 1.9 1.7 1.9 2.0 1.9
bergamotene
(z)-β farnesene 9.7 3.9 3.3 3.5 3.7 3.5
α caryophyllene 16.0 7.4 6.4 7.2 7.6 7.3
(e)-β farnesene 8.6 1.7 1.6 1.8 2.0 2.0
gurjunene 4.9 2.0 1.8 1.9 2.1 2.0
Δ guaiene 2.2 - - - - -
(e)-nerolidol 3.3 4.1 3.9 4.6 5.0 5.2
Unidentified 3.5 3.1 3.2 3.8 4.3 4.6
Unidentified 1.8 3.1 2.8 3.0 2.9 2.7
α-bisabolol 2.9 - - - - -
Monoterpenes 25.9 60.9 66.6 61.6 60.1 60.6
Sesquiterpenes 74.1 39.1 33.4 38.4 39.9 39.4
trichomes extracted from high-THC clone G2 M6 at various stages in the plant’s development.
The results are the relative peak area after analysis of the essential oil by GC. Missing values
occur where individual terpene contents were below the detectable limits.
96
Raw material
Foliar Foliar and Floral Mixture
used
Trichome type Sessile Predominantly Glandular Stalked
Wks after
3 9 10 11 12 13
planting
Weeks in flower 0 6 7 8 9 10
Monoterpenes % Peak Area
α-pinene - 9.1 8.0 6.8 8.0 6.7
β-pinene - 3.7 3.1 3.3 3.6 3.1
Myrcene 16.0 41.8 39.7 50.3 49.1 51.1
Limonene - 4.6 4.6 6.0 5.9 6.3
β-ocimene 3.0 9.0 8.4 11.0 9.9 10.3
Sesquiterpenes % Peak Area
Trans
49.8 22.5 25.6 15.1 15.9 14.4
caryophyllene
α caryophyllene 3.2 5.9 6.8 4.2 4.4 4.2
(e)-β farnesene 12.4 2.4 2.7 2.3 2.2 2.6
Unidentified 0.6 - - - - -
Unidentified 6.2 - - - - -
(e)-nerolidol 6.3 1.0 1.1 1.1 1.1 1.2
Caryophyllene
0.7 - - - - -
Ox
α bisobolol 1.7 - - - - -
Monoterpenes 19.1 68.2 63.8 77.4 76.4 77.6
Sesquiterpenes 80.9 31.8 36.2 22.6 23.6 22.4
trichomes extracted from high-CBD clone G5 M13 at various stages in the plants development.
The results are the relative peak area after analysis of the essential oil by GC. Missing values
occur where individual terpene contents were below the detectable limits.
Unfortunately, in the CBD-chemovar, the α-pinene, β-pinene and limonene

concentrations were below the detectable minimum in the foliar fraction and a similar
comparison was not possible. In this chemovar monoterpenes accounted for 64-68% of
the terpene total in the first two harvests. However, this rose to 76-78% and remained
stable for the last three weeks when plant development ceased. As with the high-THC
clone, the proportion of monoterpenes was much lower in the young vegetative foliage
(19%). It appears likely that the slightly higher proportion of sesquiterpenes in the
97
younger flowering plants was due to these having a higher proportion of sesquiterpene-
rich leaves. The proportion of foliage would have fallen as floral development
continued up to the eighth week after induction of flowering. The higher proportion of
sesquiterpenes in less developed plants observed here mirrors a similar observation
reported when a hemp crop was grown outdoors for the commercial production of
essential oils (Meier and Mediavilla, 1998). A detailed evaluation of terpene production
in trichomes in field-grown plants is described later in Chapter 6.
Insufficient entriched trichome preparation was available to enable analysis of the

cannabinoid content. However it was possible to assess the cannabinoid profile of the
original plant material. The results are shown in Table 4.8 and 4.9.
Weeks after planting 3 9 10 11 12 13

Weeks in flower - 6 7 8 9 10
Cannabinoid %w/w
THC 0.40 7.92 8.94 8.92 8.92 8.33
CBC 0.06 0.14 0.13 0.12 0.12 0.11
CBG - 0.11 0.11 0.12 0.13 0.11
Table 4.8. The cannabinoid profile of the original plant material (high-THC clone G2 M6) from
which the trichome rich preparations were made. Six subsamples were combined and milled to
produce one sample for analysis by GC. A missing value denotes that the cannabinoid level was
below the detectable threshold.
Weeks after planting 3 9 10 11 12 13

Weeks in flower _ 6 7 8 9 10
Cannabinoid %w/w
CBD 0.20 2.93 3.14 3.10 3.73 3.16
CBC _ 0.16 0.16 0.14 0.16 0.13
THC _ 0.20 0.17 0.15 0.16 0.14
CBG _ 0.03 0.03 0.02 0.03 0.02
Table 4.9. The cannabinoid profile of the original plant material (high-CBD clone G5 M13)
from which the trichome rich preparations were made. Six subsamples were combined and
milled to produce one sample for analysis by GC. A missing value denotes that cannabinoid
level was below the detectable threshold.
Plant development was seen to have ceased by their eighth week in short daylength,
and plants harvested after nine and ten weeks showed very high levels of foliar
98
senescence. The results of the cannabinoid assay suggest that both cultivars had
reached maximum potency at the end of the seventh week in short daylength. In both
chemotypes, delaying the harvest beyond the eigth week of flowering appears to have
no clear effect on the terpene or cannabinoid profiles. The foliage from three week old
vegetative plants showed very low levels of cannabinoid and terpene, as expected.
The remarkably unchanging cannabinoid and terpene profiles, from the eighth week of
the flowering period onward, suggest that the secondary metabolites are stable while
sequestered in the trichome. This facilitates the task of the grower of pharmaceutical-
grade cannabis, as delaying harvest does not appear to affect the specified ratio of
metabolites. This is in contrast to observation in other species where the terpene profile
changes due to enzymic conversion of one terpene to another. In peppermint Mentha
piperita L. for example, menthone is converted to menthol in the latter stages of
flowering. (Gershenzon, 2000) Research also showed that catabolism of
monoterpenes occurs in intact glandular trichomes of mint (Croteau and Martinkus,
1979).
The much higher secondary metabolite content of floral tissue compared to the foliage
is typical of most plant species. The loss of floral tissues is likely to have greater
impact on a plant’s ability to pass on its genes to the next generation. The Optimal
Defence Theory suggests that, away from the influence of man, plants will have
evolved to generally allocate secondary metabolites to tissues in direct proportion to
their value. (Herms and Matson, 1992). The very different balance of monoterpenes in
the sessile trichomes on the foliage and the predominantly capitate stalked trichomes
on floral tissues is supporting evidence that these trichomes have different functions.
Both types contain bitter sesquiterpenes which can act as anti-feedant repellents. The
increased monoterpene content of capitate stalked trichomes would be expected to
lower the viscosity of the contents, thereby making it more able for them to ensnare
insects, as reported in the previous chapter. The monoterpenes are more volatile, and
being hydrophobic they are highly persistent in the atmosphere. Insect olfactory
systems are devoid of the mucous membranes found in mammals, and they are
especially sensitive to such lypophylic chemicals. Monoterpenes are thereby detected
by insects at considerable distances from the plant. In many cases these
monoterpenes are repellent to insects, (e.g. α-pinene and ants) those insects
apparently misidentifying the monoterpene as an alarm pheromone (Kelsey et al.,
1984). It is notable that the secondary metabolite profiles of both chemotypes (Tables
4.6 and 4.7) were dominated by the most reduced forms of terpenes – the ‘true
terpenes’ - containing just carbon and hydrogen. These were the most costly type to
99
biosynthesise (Gershenzon, 1994). The slightly more oxidised monoterpene alcohols

(e.g. linalool,) are more water soluble (Merck Index, 1996) and less persistent in an
atmosphere containing water vapour. It is possible that it is a function of the
cannabinoids, by virtue of their anti-oxidant properties, is to prevent oxidation of the
monoterpenes to more soluble monoterpene alcohols.
Some of the high monoterpene content of glandular trichomes is possibly due to

human influence over many centuries. The highest monoterpene contents are most
likely found in plants producing high quantities of resin. These plants would have been
selectively planted by growers striving for maximum resin yields. Many would also find
the odour attractive and of greater demand for those using fragrant resins as religious
sacraments. Focused plant breeding of cannabis for pharmaceutical use continues to
make dramatic changes to the glandular trichome contents.
4.6 CONCLUSIONS
The previous chapter showed that on mature flowering plants, only a very small
proportion of the total cannabinoid content is sequestered in sessile trichomes, the
majority being synthesized in the capitate stalked form. On a mature flowering plant,
the cannabinoid content of the sessile trichomes has an almost negligible influence on
the total cannabinoid content. Comparisons of proximal and distal bract sections, with
differing ratios of sessile and capitate stalked trichomes, suggested however that the
former had a higher CBC content within the cannabinoid profile. By separating and
then analysing sessile and capitate stalked trichomes, the research reported in this
chapter confirmed this to be the case.
Although yields were comparatively small, enriched trichome preparations could be

produced from foliage, within which only the sessile form was present. Agitating the
foliage in cold water (<1°C) appeared to dislodge a higher proportion of the more
mature sessile trichomes, within which there was a higher proportion of CBC than that
of the overall sessile trichome population. Using the foliage of a chemovar, within which
CBC accounted for approximately 61% w/w of the cannabinoid total, an enriched
trichome preparation could be produced, within which CBC constituted 94% w/w of the
cannabinoid total. The CBC content of the dried preparation was 44% w/w. This was
possibly the first time that such a rich source of CBC had been extracted from cannabis
material by such simple means.
In marked contrast to some other species, this chapter showed that in Cannabis the
terpenoid content of trichomes appears to be stable at the end of the flowering period,
at least in the glasshouse growing environment. The cannabinoid and terpene profile
100
of the phytopharmaceutical feedstock would be minimally affected by a few days

alteration to the intended harvest date. This indicates that some flexibility in harvest
timing is justified.
Capitate stalked trichomes appear to have a higher monoterpene:sesquiterpene ratio

than the sessile type. Although only the sessile trichome form is present on foliage,
floral material predominantly exhibits the capitate stalked form. As a consequence of
this marked difference if trichome distribution, floral material has a higher
monoterpene:sesquiterpene ratio than foliage. Any alteration to growing conditions or
harvest timing, which affected the ratio of foliar and floral material, would be likely to
affect the balance of mono and sesquiterpenes in the feedstock produced.
The previous chapter showed that only the capitate stalked trichome form was involved
in insect entrapment. The higher proportion in monoterpenes in this form implies that
the secretory head contents would consequently be of lower viscosity than those of the
sessile form. If punctured by insect contact, the contents would more readily spread
over the insect surface. As the monoterpenes from the punctured trichome volatalised,
this would leave an increasingly viscous adhesive coating on the insect. Microscopic
observations indeed showed the capitate trichome contents to have extremely
adhesive qualities.
The following chapter evaluates glasshouse growing methods with the aim of enabling
the reliable and uniform propagation of high yielding phytopharmaceutical feedstocks,
with a high density of capitate stalked trichomes. Uniform foliage:flower ratios would
appear to be vital if both materials are included in the feedstock.
101
Chapter 5. Indoor Propagation of Medicinal Cannabis
Chapter 5 Indoor Propagation of Medicinal Cannabis
5.1 INTRODUCTION
To enable year-round propagation of high yielding good quality plant material for
GW Pharmaceuticals, extensive research was needed to identify suitable horticultural
methods. The techniques initially adopted were recommended by consultants from
HortaPharm BV. Additional advice was sought from the books on cannabis horticulture
then available from at least a dozen authors (e.g. Clarke, 1981; Frank, 1997; Rosenthal,
1998). Although the first of these books sourced the work of both scientific and the
clandestine growers, most publications appeared to rely totally on the latter and
contained minimal input and peer review from named qualified and respected
scientists. Indeed the growing of cannabis was often described as more of an art than a
science. Within three weeks of planting the first seed, insect damage was observed on
the crop and the battle against pest and disease had begun. Growth media had to be
developed that supported good root development, and fertilizer regimes had to be
devised to give the correct balance of nutrients at the appropriate time in the plants’
development. As the results of these studies are also of interest to illicit cannabis
growers, this aspect of the research cannot be included in this thesis.
Along with the three most commonly used legal drugs in the west – alcohol, nicotine
and caffeine – cannabis-derived THC is unusual, if not unique, amongst illicit drugs in
that part of the enjoyment for many users is the taste and odour of the plant material
from which it is sourced. In addition to raising yield and potency, many of the cannabis
growing techniques for illicit cannabis were designed to improve the taste of the
harvested material. This would be of little or no relevance when growing cannabis as a
phytopharmaceutical, and this typifies how the aims of the illicit and pharmaceutical
growers differed. The research discussed in this thesis also differs in that it investigated
the effect of growing conditions on the biosynthesis of cannabinoids other than just
THC. To enable the tight specification expected of a phytopharmaceutical feedstock to
be met, the research reported here placed greater emphasis on crop uniformity than
crop yield. There was little doubt that environmental conditions influenced the quantity
of cannabinoids in various plant parts at different growth stages, as demonstrated by
Fairbairn (1976), Lydon and Teramura (1987) and others. An improved knowledge of
how alterations to the growing method affected the secondary metabolite profile would
reduce the likelihood of unacceptable feedstock being produced. It would also increase
the ability to perform a reactive diagnosis, if a batch of feedstock was produced that
102
was ‘out of specification’. Reviewing the findings of several researchers, Raman and
Joshi (1998) highlighted the magnitude of this challenge. They reported that even if
genetic and environmental factors could be tightly controlled, high interplant variability
in cannabinoid content could be expected between plants of the same chemotype
when grown under identical conditions.
Initial advice was that plants should be propagated from cuttings rather than seed.
Most researchers agreed that genetics exert the greatest control of a plant’s
cannabinoid profile (Beutler and Der Manderosian, 1978; Fournier et al., 1987;
deMeijer, 1994). The propagation of cannabis from cuttings guaranteed the genetic
uniformity of all plants produced from the same source (Samuelsson, 1999).
Propagation from seed however was perceived to be less labour intensive, and with
appropriate research perhaps the foreseen variation could be dispelled or overcome.
Research reported here compared the variability of cloned and seed-sown plants.
To produce a medicine containing a desired even balance of cannabinoids, separate

THC and CBD chemotypes were routinely propagated for Sativex® production. By
blending Botanical Drug Substances from these two, it could be guaranteed that the
correct ratio of these cannabinoids would be delivered. It has been concluded
elsewhere that the inheritance of CBD and THC chemotype was controlled by a
monogenic, co-dominant mechanism (de Meijer et al., 2003). A single locus referred to
as B was postulated, with two alleles, BD and BT, encoding CBD- and THC synthase
respectively. According to this model, the CBD chemotype had a BD/BD genotype at the
B locus, and the THC predominant plant had a BT/BT genotype. Heterozygous BD/BT
genotypes were available that produced both cannabinoids, thereby possibly avoiding
the need to blend materials. Research reported here was performed to assess how
growing methods and harvest timing affected the ratio of these two cannabinoids in
BD/BT genotypes.
Prior to research for this thesis, large seasonal variations were seen in plant yield. In
winter crop yields fell by two-thirds and the fall in cannabinoid yield was even more
pronounced. In addition to the economic implications, this variability in potency may
have also unacceptably affected the secondary metabolite profile of the Botanical Drug
Substance (purified extract) made from that feedstock. While glasshouse temperature
remained constant throughout the year, summer light levels far exceeded those in
winter, and this was the suspected main cause of the yield variation. Limited growth
room tests, performed in winter, showed that when summer irradiance levels were
recreated, both plant yield and cannabinoid assay values returned to those typical of a
summer crop. Extensive written guidance was available regarding the use of lighting
103
systems to support cannabis growth. Most was aimed at those growing under totally
artificial lighting conditions and few described the use of electrical lighting to
supplement natural daylight. Growing cannabis this way in a glasshouse had been and
continues to be less common, at least in the UK (ACMD, 2008), partly due to it being
too difficult to keep the activity secret and secure. The effect of lighting on cannabis
growth was discussed in little detail in peer-reviewed scientific literature.
It was well recognised that the floral material of the female plant was the most potent
source of cannabinoids (Fairbairn, 1976). To confirm this, various parts of some
unpollinated THC chemovar plants were analysed. The THC content of dried seeds,
roots, stems, leaves and inflorescences was found to be 0.0%, 0.0%, 0.3%, 0.8% and
15.2% w/w respectively. When the outer bracts and stalks were removed the THC
content of the remaining inflorescence material was found to be over 18% (Potter
2004). In a separate test pollination was found to reduce THC content by more than
half, with THC yield per unit area being decreased by over 75%. It was therefore
essential to grow plants though to the flowering stage to obtain the highest yielding
material.
Cannabis is generally a ‘short day plant’ that by definition only commences to flower
late in summer, once the day length starts to fall (Clark, 1981). When the so called
critical daylength is reached floral development is stimulated. More correctly, the plant
is responding to the increasing length of the night, during which time a light-sensitive
phytochrome protein slowly dimerises to a different form. Within seconds of light
exposure the protein reverts back to the original structure. It is only when the night time
is sufficiently long that a required balance of the dimers is reached to signal
commencement of flowering (Halliday and Fankauser, 2003). To reliably induce
flowering in most varieties of cannabis, the night-length must be greater than the critical
value. The critical daylength for an individual variety is greatly affected by its
geographical origin and would generally be greatest in those plants derived well away
from the equator (de Meijer and Keizer, 1994). Exceptions to this response occur in
plants adapted to grow in equatorial regions, where there is minimal variation in day
length. Flowering in tropical cannabis plants is more closely related to plant age. In
contrast, rapid flowering ecotypes are found at latitudes of 60º (Callaway, 2002) or
more. These have typically adapted to survive in the very short growing season, and
commence flowering early in the season irrespective of the daylength. Despite these
differences in critical daylength, it has been common practise amongst illicit cannabis
indoor-growers to induce plants to flower by placing them in a twelve hour daylength.
The use of such a short day length to induce flowering was perceived as somewhat
104
ironic, as in the northern hemisphere an autumn day length of twelve hours occurs
naturally at the equinox during the last days of September, when cannabis flowering
would be finishing. A slightly longer day length would still initiate flowering and would
result in proportionally more light energy being delivered to the plant in one day.
Research reported here investigates how plant development, yield and cannabinoid
profile are affected by differing daylengths of eleven, twelve and thirteen hours.
Recreational cannabis users would often harvest their plants when about 95% of the
visible stigmas had senesced, but this would vary according to the variety and the
grower’s personal preference (Clarke, 1993). This was partly related to taste of the
material as well as its potency. It was not known how closely inflorescence
development stage was related to the secondary metabolite content. It was predicted
that the length of the propagation period would not greatly affect the cannabinoid
profile, especially of those cannabinoids that are the final products of separate
biosynthetic pathways. Rowan and Fairbairn (1977) and others had shown that during
the flowering phase CBD chemotypes would produce a small proportion of THC and
THC chemotypes would produce a small proportion of CBD. In THC chemotypes,
although substantial proportions of CBC were often present in the early stages of
growth, THC dominated the profile of THC chemotypes throughout the flowering stage.
However, how the state of maturity of the plant affected the relative proportions the
precursor CBG and the final cannabinoids was less well studied, and was investigated
in this study. Based on visual assessment, clones of the THC chemotype typically
appeared ready for harvest eight weeks after being placed in a twelve hour day length.
The objective of this investigation was to check that this flowering time was the
optimum to achieve efficient production of THC. The investigation would also
investigate how the duration of the flowering period affected the relative proportions of
THC and its precursor CBG.
The methods of production of licensed medicines had to comply with Rules and
Guidance for Pharmaceutical Manufacturers and Distributors 2007 as directed by the
Medicines and Healthcare Products Regulatory Agency (MHRA, 2007) to ensure that
they were of sufficient quality. It was perceived that GMP did not apply to the
propagation phase of the production process, but did apply once the plant material was
processed. Although the cannabis plants were to be grown in a glasshouse, the
research described in this chapter was performed in the knowledge that the
propagation phase would have to comply with the EMEA Good Agricultural Practice
(GAP) Guidelines (EMEA, 2006). The use of the word agriculture in this quality control
105
system reflects that most, if not all of the existing phytopharmaceuticals in western
medicines, were derived from outdoor grown materials.

A series of tests was performed, all of which had the aim of identifying how glasshouse
growing methods could be altered to maximise plant quality and uniformity. The linked
objectives were as follows:
5.2.1 To gain a greater understanding of the irradiance levels on plant yield and
cannabinoid content.
5.2.2 To confirm or dispel the belief that plants propagated from clones were
significantly more uniform in cannabinoid profile than plants grown from seed.
5.2.3 To ascertain how the length of flowering period affected the cannabinoid yield.
5.2.4 To ascertain how the length of flowering period affected the cannabinoid
profile of THC chemovars and heterozygous plants with a mixed THC/CBD
profile.
5.2.5 To compare the effect of 13, 12 and 11 hour daylengths on plant development,
cannabinoid yield, potency and profile.
106
5.3 MATERIALS
5.3.1 Plant Propagation and Drying Materials
Materials Source
Complete light fittings incorporating 600 watt Osram
Nav T High Pressure Sodium Lamps, compatible
ballasts and Silver Wing® reflectors.
Hortisystems UK Ltd
Complete light fittings incorporating 600 watt Hortilux-
Pulborough, UK
Scroeder High Pressure Sodium Lamps, compatible
ballasts and Deep reflectors
1000 watt Mercury Vapour MBFU lamps
Seradix® Rooting Powder (MAPP No 1331) 0.3% w/w Certis UK 1b Mills Way,
4-Indol-3yl butyric acid Amesbury, UK
Jiffy Products International
Jiffy 7 Rooting Plugs
AS, Kristiansand, Norway
Thripex-Plus sachets Amblyseius cucumeris for
biocontrol of Thrip Thrips tabaci
Spidex Phytoseiulus persimilis for biocontrol of red Koppert B.V. Rodenrijs,

spider mite Tetranychus urticae NL
En-strip Encarsia formosa for biocontrol of white-fly
Trialeurodes vaporariorum
Aphipar bottles Aphidius colemani for biocontrol of
cotton-melon aphid Aphis gossypii
Ebac BD150 commercial building drier Ebac Ltd., Durham, UK
Table 5.1. Plant Propagation Materials
107
5.3.2 Germplasm Details
GW Variety/ Supplier
Accession Clone
Number Name
G1-M1 Gwen
G2-M7 Gina Seed from HortaPharm BV Schimkelhavemkade 1075 VS
G5- M11 Gayle Amsterdam, NL, used to produce clones by
G5- M13 Grace GW Pharmaceuticals Ltd, Porton Down Science Park,
Salisbury, Wiltshire
G5- M16 Gill
M237 - Clone bred by GW Pharmaceuticals (As above)
Various suppliers as follows:
Sensi Seeds B.V., Rotterdam, NL

G41 to
G52, G63, Natural Mystic, Rochdale, UK
Various
G91 G130 Nirvana, Zaandam, NL
and G159 Dutch Passion B.V. Amsterdam, NL
Serious Seeds, Amsterdam, NL
Seedsman Ltd, London, UK
Table 5.2. Germplasm Details
5.3.3 Light Measurement and Weighing Equipment
Skye 500 Hand-held Light Sensor Skye Instruments, Powys, UK

Kipp and Zonnen solarimeter Priva UK Ltd, Gloucestershire, UK
E&G Websales Ltd, Delfryn, Lixwm,
Salter M323 Top Pan Balance
Holywell, UK
Table 5.3. Light Measurement and Weighing Equipment
5.3.4 Growth Medium

The growth medium used for these studies was developed at GW Pharmaceuticals Ltd,
Porton Down Science Park, Salisbury, Wiltshire and its precise formula has to remain
proprietary information (Wilkinson 2006). The medium (or compost) primarily consisted
of peat and perlite and included sufficient nutrient to maintain healthy plant growth
through to harvest with no additional feeding.
108
5.4 METHODS
Before describing the more detailed experimental methods that apply specifically to
individual studies, sections 5.4.1.1 to 5.4.1.11 describe the routine propagation and
botanical raw material production processes which were routinely used and applied to
most of the studies described in this chapter. Sections 5.4.2.1 to 5.4.2.11 describe
more specific methods applying to individual tests within this chapter.
5.4.1 Routine Propagation and Plant Production Methods
5.4.1.1 Seed sowing and transplantation of seedlings
Seeds were sown in trays of a range of proprietary seed compost. There is no light
requirement for germination and seeds were sown approximately 1 cm deep and at
least 1cm apart. The compost was sufficiently deep (> 3 cm) to allow some natural
downward root development prior to replanting. The compost was lightly compressed
after sowing, and well watered. The temperature was maintained at 20-25°C and
lighting adjusted to maintain a 24 hour daylength. Horticultural fleece was placed over
the seed tray or pot until seedling emergence to reduce surface drying of the compost.
In ideal conditions seedlings were ready for transplanting approximately ten days after
sowing, by which time the hypocotyls were typically around 10 cm tall and the first pair
of true leaves well formed. Seedlings were teased from the seed compost and
transplanted into individual pots of the appropriate growth medium. A sufficiently deep
hole is prepared in this medium to allow the seedling to be lowered undamaged so that
the first pair of true leaves sat 1-2 cm above the surface of the medium. The
surrounding growth medium was firmed by hand, and the growth medium sufficiently
watered to maintain moisture around the seedling.
5.4.1.2 Production of Cuttings (Clones)
Branches of vegetative material were removed from plants producing ample numbers
of axial buds. (Figures 5.1 a-f). The branch was then cut into sections, each carrying
one axial bud and retaining approximately 5cm of stem below the bud. The stem below
the bud was dipped in rooting powder containing 0.3% w/w 4-Indol-3yl butyric acid and
then promptly into a very moist peat plug. These plugs were placed on a bench of
regularly wetted gravel under a polythene cover to maintain very high humidity.
Irradiance levels under the polythene was maintained at approximately 10 – 15 W m-2
for 24 hours per day. Successful cuttings were normally producing sufficient roots
within fourteen days to enable repotting.
109
5.4.1.3 Nurturing Vegetative Growth of Seedlings and Cuttings.
For the first three weeks vegetative growth was encouraged by maintaining the
cannabis plants in a twenty-four hour daylength. This supported maximum growth rate
by enabling continuous photosynthesis. Being ‘short-day plants’, flowering was
naturally prevented in this regime.
Unless otherwise stated, all high-THC plants are potted in five litre pots of GW
Pharmaceuticals’ standardised peat-based growth medium. High CBD plants of variety
G5 are potted in three litre pots of the same medium. These were closely placed pot-to-
pot on the glasshouse or growth-room bench for three weeks before being induced to
flower. A proportion of plants were retained in long daylength however to produce
ample vegetative material for the production of cuttings.
5.4.1.4 Induction and Maintenance of Flowering
Plants were relocated, or the lighting regime adjusted, so that plants were now in a
short daylength. Unless otherwise specified, this was twelve hours per day. The
response to daylength change was for plants to commence flowering within seven to
fourteen days. The short daylength was maintained for the entire flowering period,
which typically lasts eight weeks. During this period, unless otherwise specified, crops
were spaced at a density of 10 m-2 (THC chemotype) or 17 m-2 (CBD chemotype).
5.4.1.5 Biological Pest Control
The major pests encountered were red spider mites (Tetranychus urticae), the onion
thrip (Thrips tabaci), white fly (Trialeurodes vaporariorum) and cotton-melon aphid
(Aphis gossypii). These were controlled by regular introduction of the predatory mite
and insect species Phytoseilius persimilis, Amblyseius cucumeris, Encarsia formosa
and Aphidius colemani (Malais and Ravensberg, 1992).
5.4.1.6 Harvesting
Plants were harvested by cutting the stems just below the lowest side-branch and
placed on a clean surface if being temporarily stored prior to drying. As respiration
could rapidly generate heat within piled fresh material, drying of harvested plants was
commenced as soon as possible to avoid heat-induced catabolism.
110
a) b)
c) d)
e) f)
Figure 5.1. Production of cannabis cuttings. A vegetative cannabis branch (a) is cut into
sections (b). Each cutting has been cut leaving approximately five centimetre of stem below a
single axial bud and up to one centimetre above (c). The base of the cutting is dipped in rooting
powder (d) and then placed in moist peat plugs (e). After two weeks roots are protruding from
the peat plugs and the cutting is ready to be planted.
5.4.1.7 Crop Drying
Plants were hung to dry in a closed environment and industrial dehumidifiers used to
lower air moisture (Figure 5.2). Horticultural fans were used to maintain air circulation.
Freshly harvested material had an aqueous content of approximately 80% (w/w). The
air surrounding the chamber was conditioned as appropriate to ensure that waste heat
from the dehumidifiers did not raise temperatures within the chamber above 35oC. The
111
crop was considered dry enough for storage or processing when below 15% (w/w). At
his point, the crop was clearly crisp to touch, the inflorescence tissue closest to the
stem feeling dry and the floral material would readily pull away from the stem without
excessive force.
Figure 5.2. A newly harvested crop hung to dry on wires.
5.4.1.8 Stripping
When sufficiently dry, all floral and leaf material was stripped from the stem by pulling
the plant though the tightly clasped thumb and forefinger of a gloved hand. This floral
and foliar mixture is referred to as Botanical Raw Material (BRM).
5.4.1.9 Garbling
Tyler et al., 1998 defined garbling as ‘’the final step in the preparation of a crude drug
and involving the removal of extraneous matter such as other parts of the plant, dirt
and added adulterants. This is done to some extent during collection and should be
carried out after the drug is dried and before it is baled and packaged’’.
When grown in the glasshouse, minimal adulteration or contamination with dirt arose.
The presence of low-potency stem material was undesirable and all material assessed
as more than 2 mm in diameter would be manually removed.
5.4.1.10 Storage
As with other dried herbs, e.g. hops (Henderson, 1973) the aqueous content of stored
dried herbs is greatly affected by the surrounding humidity. To minimise the possible
effects of varying moisture content on microflora activity, secondary metabolite
volatilisation and catabolism the herbal material was stored in a controlled environment
112
at 35 ± 5% RH. Low air moisture was achieved using industrial dehumidifiers. THC
degrades more rapidly in the presence of light (Fairbairn, 1976). To discourage this,
the material was stored in the dark prior to analysis or use.
5.4.1.11 Environmental Control System
Glasshouse temperatures and lighting conditions were controlled by a Priva Universal

Computer (907) system. Warmed or chilled air was supplied via a rapidly responding
air handling system to ensure that a target glasshouse temperature was achieved +/-
1.5°C. Supplementary lighting was programmed to be turned on at the desired times
of the day, and only when natural daylight irradiance levels were below a selected
minimum level. Similarly, glasshouse roof shades and or blinds were automatically
opened or closed to control the ingress of natural light and to prevent light pollution
from glasshouse lights after sunset.
5.4.2 Specific Methods

5.4.2.1 Uniformity of Plants Grown from Cuttings or Seed
THC variety G1 was selected for this test. Thirty plants raised from cuttings were
compared to a similar number raised from seeds. The clones were all derived from one
plant and code named clone M1. For the seed sown crop, seedlings were raised as
described in paragraph 5.4.1.1. For the plants raised from cuttings, the propagation
timings and conditions were as described in paragraph 5.4.1.2. Ten days after sowing,
eighty individual seedlings were transplanted into pots using identical materials and
methods employed for the propagation of plants from cuttings. The rooted cuttings
were also transplanted on the same day.
Three weeks after transplanting, the plants were moved to a twelve hour light / twelve
hour dark regime to induce flowering. All male plants and excess female plants were
removed to leave thirty seed and thirty clone derived plants. These were maintained in
neighbouring blocks at a density of ten plants per square metre. Eight weeks after the
move to twelve hour days, the plants were harvested and dried. The total weight of
foliar and floral material produced by each plant after removal of the stem was
recorded. This mixture was then milled and the cannabinoid content studied using gas
chromatography.
5.4.2.2 Effect of Duration of Flowering Period on Yield
113
Two clones used for the production of clinical trials raw material were selected viz. M1
and M7. Twelve commercially available varieties were also selected. All were
commonly used for illicit recreational purposes and were described as being derived
from tropical or subtropical areas. Plants truly derived from these contrasting locations
would be expected to differ in the natural duration of their flowering period.
Seedlings of each of the twelve commercial varieties were raised in the glasshouse in a
twenty-four hour daylength. When sufficiently developed, cuttings were taken from
each plant and encouraged to root using the standard method (section 5.4.2). The
original plants, from which the cuttings were taken, were moved to a twelve-hour
daylength regime to encourage flowering. All plants consequently identified as male
were disposed of along with all cuttings taken from them prior to gender identification.
Up to three female clone lines of each variety were retained for further evaluation, each
being regarded as the most potent or prolifically flowering examples of their variety.
Just one clone line was chosen for evaluation from four of the varieties. Two or three
clone lines were selected from the remainder. In the latter case, these varieties had
shown a high level of phenotypic variability.
The propagation regime was based on that used for regular production of THC for
clinical trial (Section 5.4.2 – 5.4.8). However, a temporary installation of supplementary
glasshouse lights was in place for this test and this gave a slightly lower minimum
irradiance level of 40 W m-2 PAR. Fifteen plants of each clone were propagated and
five of each harvested after six, eight and ten weeks in flower. Harvested plants were
dried in a dehumidified environment (final humidity < 35%) for seven days and the floral
and foliar material stripped from the stem. The floral and foliar material of the five
replicate plants of each clone line were thoroughly mixed. Five small samples of
approximately 1 g were taken at random from the mixture, blended and analysed by
High Performance Liquid Chromatography (Appendix 1).
5.4.2.3 Effect of Daylength on Cannabinoid Profile (Part 1) Comparison of 12 and 13

hour daylength
All plants were propagated in the glasshouse using the standard materials and
methods. Eighty plants of each selected clone line were grown in five-litre pots and
maintained in constant twenty-four hour daylength for the first three weeks after potting.
Thereafter half were transferred to a glasshouse area with a twelve-hour daylength and
the remainder relocated to a similar glasshouse area with a thirteen-hour daylength
regime. In both areas the glasshouse target irradiance level was 75 Wm-2. Within each
regime, each clone line was divided into two batches of twenty plants. Single plant
114
batches of each clone line were placed alongside each other at a pot-density of 10 m-2.
Plants were watered by hand throughout the test. Eight weeks after the move to short
daylength, one batch of each clone was harvested and hung to dry. The remaining
batches were harvested and dried fourteen days later.
5.4.2.4 Effect of Daylength on Cannabinoid Profile (Part 2) Comparison of 11 and 12

hour daylength
The above method was duplicated. However, this test was located in a growth room
with no natural lighting (Figure 5.3), and twenty plants of each clone (rather than five,
as used in the previous test) were included in each replicate. High-pressure sodium
lamps gave a uniform irradiance level of 70 W m-2. Temperature was maintained at 25
± 1°C. ‘Ebb-and-flood’ benches provided uniform levels of water to all plants. Vertical
blinds between areas enabled plants to be divided into areas with equal environmental
conditions but altered daylength.
Figure 5.3. An experiment to compare plant development and cannabinoid content when
flowered in 11 and 12 hour daylengths. Plants are maintained on ebb-and-flood benches and
lighting provided by high pressure sodium lamps. Duplicate batches of plants are maintained
either side of the curtain with plants on the right receiving the longer daylength regime.
5.4.2.5 Plant height assessment
The height of each single plant (from the surface of the pot to the top of the tallest
inflorescence) was measured by hand immediately before harvest.
5.4.2.6 Stigma senescence assessment
115
The stage of inflorescence development on each clone was recorded weekly, up to the
final harvest date, by making a visual estimate of the relative proportions of still-viable
white stigmas and senesced non-viable brown stigmas (Figure 5.4).
Newly formed white stigmas are

viable and receptive to pollen.
Aged stigmas have senesced to

a brown colour and are no
longer viable.
Figure 5.4. A close-up view of part of an unpolllinated cannabis inflorescence, showing viable
and older non-viable stigmas.
5.4.2.7 Plant Weight Assessment
After being dried for seven days, the foliar and floral material from each individual plant
was stripped from the stem and the latter discarded. The weight of combined foliar and
floral material was weighed on a top pan balance.
5.4.2.8 Cannabinoid Content and Profile
The cannabinoid content and profile of the samples from the twelve versus thirteen
hour assessment were analysed by HPLC. This facility was not available for the eleven
versus twelve hour study. Cannabinoid yields and profiles were assessed using GC
(Appendix 1).
116
5.4.2.9 Effect of Irradiance Level on Plant and Cannabinoid Yield
The mercury vapour (MBFU) lamps originally fitted in the glasshouse were only able to
convert approximately 10% of the consumed electrical energy into photosynthetically
active radiation. This compared to 30% with more modern High Pressure Sodium
(HPS) and metal halide (MH) lamps. By replacing each 1000 watt MBF fitting with two
600 watt high pressure sodium (HPS) or metal halide (MH) units the electricity
consumption within the glasshouse would have been raised by 20% to the maximum
capacity of the existing power supply. However, the increase in irradiance achieved by
new fittings would have theoretically been three-fold, taking levels from 16 to 53 W m-2.
To measure the effect of introducing such improvements to irradiance levels, identical

batches of plants were grown in the glasshouse and in a walk-in growth chamber. The
latter was equipped with a combination of HPS and MH lamps to deliver an irradiance
level of 70 W m-2 at the plant canopy. This equates to the light level typically existing in
the glasshouse, as equipped at the time, during mid-afternoon on a bright summer day.
The experiment commenced in September and finished in December so as to recreate

a harvest date when the previous year’s findings suggested that glasshouse yields
would be at or near their minimum. Records from the Priva Glasshouse Control
computer system showed that on rare very clear days, noon irradiance levels inside the
glasshouse peaked at approximately 50 W m-2. However a maximum of approximately
25 W m-2 was more common, with morning and evening levels falling to a minimum of
17 W m-2. Irradiance levels were measured using a hand-help Skye SKE 500 light
metre at ten locations on the surface of the crop canopy.
One hundred rooted cuttings of each of the THC chemovars G1 and the CBD
chemovar G5 were placed in pots of growth medium following the standard procedures
for botanical raw material production. Fifty of each chemovar were propagated in the
growth room and the remainder were propagated in the glasshouse along-side the crop
being routinely grown for botanical raw material production. In both the glasshouse
and growth room the temperature was maintained at 25.0 ± 1.5°C throughout the trial
period. Plants were placed on a bench at maximum density (pot-to-pot) for the first
three weeks and maintained in a twenty-four-hour-per-day lighting regime. The THC
chemovar plants were then spaced out to ten plants per square metre or seventeen per
sq metre (G5 CBD chemovar) for the remaining eight weeks before harvest. During
these eight weeks the lighting regime was maintained at 12 hours light/12 hours dark
per day. Plants in both areas were watered by hand. No additional plant nutrient was
given to the plants in either growing regime.
117
Plants were harvested after the usual eight-week flowering period and dried and
weighed. Mixed samples of stripped foliar and floral dry material of each chemovar
from both regimes were analysed for cannabinoid content using HPLC.
In a subsequent larger scale test 250 m-2 of the glasshouse was converted to raise
supplementary lighting levels from 16 to 53 W m-2. Temperature and daylengths were
kept identical in both areas. The monthly average crop yields in both areas were
compared over the following year.
5.4..2.10 Effect of the length of flowering period on the cannabinoid profile of

heterozygous plants of the mixed THC/CBD chemotype.
In addition to clones specifically bred to produce a mixed THC/CBD profile, one was
discovered by chance (code name G159) while screening plants grown from
commercially available seed. This was marketed for outdoor growing in the UK.
Although most seedlings from this variety were found to be of the THC chemotype, five
plants were of the BTBD genotype, as described by de Meijer et al. (2003). These
produced plants containing THC and CBD in ratios varying between 1.5:1 and 1:1.5.
These were adopted for detailed study in this thesis as they could be included in indoor
and outdoor tests. Four rooted cuttings of each were grown in five litre pots of the
standard growth media. Plants were kept in a walk-in growth room in continuous
lighting for the first three weeks. A twelve hour light/ twelve hour dark regime was then
implemented. Temperature was maintained at 25.0 ± 1.5ºC. Using high pressure
sodium lamps, irradiance levels were kept at 75 W m-2 at foliage height. After four
weeks in short day length a program of weekly sampling commenced, one entire
inflorescence being removed from the side of each plant. These were analysed for
their cannabinoid content by gas chromatography (Appendix 1).
5.4.2.11 Statistical Analysis
Analyses of variance (ANOVA), paired t-tests and F-tests were used as appropriate,
utilising Microsoft Excel 2003 related software.
118
5.5.1 Comparison of the Yield and Uniformity of Plants Grown from

Cuttings or Seeds
The yield of raw material obtained from plants grown from seed (494 g m-2) and
cuttings (515 g m-2) was very similar. An analysis of variance (n = 30) showed this
small difference not to be significant (p > 0.05). However, the mean THC content of the
cloned plants (14.6% THC w/w) was significantly higher (ANOVA, p < 0.01) than those
grown from seed (11.1% THC). As a consequence, the cloned plants produced
significantly more THC per unit area - 75.4 g m-2 (p < 0.01, ANOVA) than those from
-2
seed - 54.9 g m . Clone M1 was one of five originally selected for further testing from
approximately one hundred female plants of the variety G1. The selection was on the
basis of its favourable ratings for vigour, yield, glandular trichome density, THC content
and purity. The THC content of these one hundred plants had exhibited a normal
distribution. It is not clear from this investigation whether the increased THC yield from
the cloned material was entirely as a result of this selection process.
Only THC, CBG and CBC were found in detectable quantities in all samples. The mean
CBG content seed sown plants was very similar to that collected from cloned plants,
and no significant difference was found. Although the mean CBC content of cloned
plants was 15% higher than that of seed derived plants, this increase was not
significant (ANOVA, p > 0.05). F-tests showed the CBC potency of seed derived plants
to be significantly more variable (p < 0.01). The difference in variability of CBG content
of seed sown plants was much less pronounced. Seed derived plants were not
significantly more variable in CBG potency than those raised from clones (p > 0.05).
When the ratios of cannabinoids in the seed-sown and cloned plants were compared it
was found that the cannabinoid profile of seeds-sown plants was more variable than
that those raised from cuttings. F-tests showed that the CBC:THC ratio and CBG:THC
ratios to be significantly more variable in plants grown from seed (p < 0.01).
To meet acceptable quality standards, phytomedicines have to undergo

standardization process, this being -
the establishment of reproducible pharmaceutical quality by comparing a product with

established reference substances and by defining minimum amounts of one or several
compounds or groups of compounds….. (or in some cases) a maximum and minimum
amount (Heinrich et al., 2004).
119
The efficacy of cannabis can be attributed to more than one cannabinoid. Two of
these, THC+CBD, have been shown to act together synergistically. (Williamson, 2001;
Musty, 2004). The interaction of the other pharmacologically active cannabinoids is
less well understood. Where active ingredients within a medicine act synergistically,
alterations to the ratio of the synergists can have a greater effect than in a medicine
where two active ingredients act additively. This investigation shows that the ratio of
synergists is more strictly controlled in cloned plants. When pharmacologically active
ingredients are extracted from botanical raw material, it could be argued that the overall
secondary metabolite profile of the entire batch is important and that the profile of
individual plants within that batch is of minimal significance. More work would be
required to test this hypothesis.
5.5.2 Effect of Irradiance Level on Plant and Cannabinoid Yield

During the first year of regular propagation of cannabis chemovars the average
monthly yield was recorded. A large seasonal yield variation was observed (Figure
5.5) with winter crops yielding less than those grown in summer.
Routine monthly analysis of cannabinoid content by GC showed that the potency of

winter crops was as little as half that of summer crops. As a consequence of the
combined drop in crop-yield and potency the cannabinoid yield of winter crops was
found to be roughly a quarter of that achieved in summer (Figure 5.6).
500
450
400
350
Yield g/sq m
300
C BD
250
TH C
200
150
100
50
0
Jan Feb Mar Ap r Ma y Jun Jul Aug Sep O ct N ov De c
Ha rvest Month
Figure 5.5. Average monthly BRM yield (two to four crops per month) (± SD) of THC and
CBD chemovars during the first full year of propagation. (No THC chemovar was harvested in
April and no CBD in February and November.)
120
70
60
50
Yield g/sq m
40 THC
30 CBD
20
10
0
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec
Harvest Month
Figure 5.6. The seasonal variation in cannabinoid yield of THC and CBD chemovars during the
first full year of propagation. Values shown were estimated by combining the average monthly
Botanical Raw Material yield (two to four crops per month) and the average monthly THC or
CBD content (w/w).
Table 5.4. shows the yield and potency of identical batches of crops propagated in the
glasshouse and in the growth-chamber. The glasshouse light level varied according to
time of day and outside lighting conditions, with mercury vapour lamps providing up to
17 W m-2. A theoretical maximum irradiance level of 50 W m-2 was achievable in the
glasshouse at noon on a bright winter day with supplementary lighting operating. An
actual maximum of approximately 25 W m-2 was more typical. Growth room conditions
were much brighter at 70 W m-2. These crops were harvested during November and
December when glasshouse grown yields were close to their minimum.
Daytime Cannabinoid Cannabinoid

Chemovar Location BRM Yield
Irradiance Potency Yield
W m-2 PAR g m-2 % w/w g m-2
THC Glasshouse 17- 50 188 11 23
THC Chamber 70 397 16 62
CBD Glasshouse 17 - 50 157 6 9
CBD Chamber 70 251 11 28
Table 5.4. A comparison of Botanical Raw Material yield and potency of a THC and a CBD
chemovar when grown in two irradiance levels during winter.
Following this compelling observation that increased irradiance resulted in such a large
increase in yield of plant material and cannabinoid, part of the glasshouse
121
supplementary lighting system was upgraded. Existing mercury vapour lamps were
replaced with high pressure sodium lamps giving a greatly improved light output of a
level similar to that achieved in the growth room test. During the following winter
months, ten batches of each of the THC and CBD chemovar crops (> 50 plants per
batch) were grown under both the mercury vapour lamps delivering 17 W m-2 PAR and
the brighter high pressure sodium supplementary lighting system delivering 55 W m-2
PAR. Under the mercury vapour lamps the THC and CBD chemovar yields were 234 (±
25.6 SD) g m-2 and 162 (± 46.6 SD) g m-2 respectively. Under high pressure sodium
lamps this increased to 478 (± 71.2 SD) g m-2 and 410 (± 82.6 SD) g m-2. In a paired t-
test the improvement in yield of both chemovars under sodium lamps was highly
significant (p < 0.01).
The crop yields were monitored over twelve months, following the complete conversion
of the glasshouse to the new lighting regime. The average monthly yields achieved
before and after the lighting improvements are shown in Figure 5.7.
700
600
Yield g/sq m
500
400 Mercury
300 Sodium
200
100
0
P
T
R
EC
B
N
R
V
N
G
Y
C
JU
FE
SE
JU
AP
O
A
JA
AU
D
M
N
M
Harvest Month
Figure 5.7. The average yield of the THC chemovar before and after the replacement of mercury
vapour lamps (17 W m-2) with high pressure sodium lamps (55 W m-2) of improved
supplementary lighting (± SD). The mean is typically for four crops per month. No crop was
harvested in April of the first year.
This showed that the monthly average yields under the brighter HPS lamps were
significantly much greater than those previously obtained under the mercury vapour
fittings (t-test, p < 0.001). This was even the case in the late summer months when
abundant natural light was available. Even on the brightest days, when outside light
122
levels reached 85 klux or more, the supplementary lighting was activated when outside
light conditions fell below 60 klux - this always being the case in the early and late part
of the day. The amount of total light energy to which plants were exposed daily was
thus greatly increased.
The new lighting regime also significantly improved the uniformity of monthly average
yields over the year (F-test, p = 0.013). However, despite this there was still some
seasonal variation in yield. The average THC-chemotype summer yield (harvested
May – October 573 g m-2) was significantly greater (ANOVA, p < 0.001) than that of
crops harvested over the rest of the year – 516 g m-2. In theory, further increases in
supplementary lighting would have increased winter yields. Alternatively uniformity
could be improved by reducing the irradiance levels in summer, by less use of
supplementary lighting or increased use of glasshouse shading.
The yields achieved in the first full year of crop growth in a solid building with no natural
lighting were similarly monitored. Yields showed a downward trend, commensurate
with the manufacturer’s predicted age-related fall in irradiance from the lamps. A
comparison of the variability in monthly yields of crops grown entirely indoors under
lamps initially delivering 75 W m-2 , with those grown in the glasshouse under HPS
supplementary lamps, revealed no significant difference (F-test, p = 0.05) between the
two growing environments.
Having demonstrated such a clear correlation between irradiance levels and cannabis
growth, the initial seasonal yield fluctuation prior to lighting improvements was
reconsidered. With hindsight it was striking that crop yields were seen to peak in those
crops harvested in August and September, a few weeks earlier than if grown outside at
the same latitude. The peak in glasshouse average daily irradiance levels would exist
around the summer solstice, five to nine weeks earlier, in late June. This suggests that
the irradiance conditions at the very beginning of flowering have the greatest potential
impact on the final yield. This supposition is supported when the clearly-similar
seasonal pattern of glasshouse yields (prior to lighting improvements) is plotted
alongside the seasonal variation of irradiance conditions existing at the
commencement of flowering of each of these glasshouse crops (Figure 5.8). It would
appear that Cannabis sativa has evolved to make maximum use of the light energy
available, during the longest and brightest days of mid-summer, to develop as dense a
foliar canopy as possible. In the following generative (flowering) phase the additional
foliage formed would undoubtedly intercept more light energy. This would increase the
photosynthetic ability of the plant, some of which would be diverted into secondary
metabolite biosynthesis. The later senescence of the additional foliage would enable
123
proportionally more metabolites to be translocated via the phloem to the inflorescences

for secondary metabolite biosynthesis.
100
80
% of Maximum
60 Irradiance
Yield
40
20
0
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec
Month Plants Commenced Flowering
Figure 5.8. Pattern of irradiance level in the glasshouse between 7 am and 7 pm (prior to the
improvements in supplementary lighting) and the pattern of average monthly yields of THC
chemovar raw material ± SD (n = 4).
This data was alternatively viewed by plotting average monthly yield as a function of
the glasshouse light level at the beginning of flowering (Figure 5.9). On both axes the
data was expressed as a percentage of the maximum observed. A regression
2
calculation confirmed the highly significant close-to-linear correlation (R = 0.96, p <
0.001) between monthly irradiance level at the commencement of flowering and the
subsequent final yield.
This clear linear correlation complemented the findings of growth room studies by
Lydon et al. (1987) who showed that cannabis assimilated carbon dioxide linearly up to
a photon flux density of approximately 500 µmol m-2 s-1, 400 - 700 nm. This is
-2
equivalent to solar radiation levels of approximately 100 W m and is well above the
maximum daily-average irradiance encountered in this study. It is assumed that the
linear assimilation of carbon dioxide observed by Lydon et al. (1987) corresponded to a
linear increase in photosynthesis. A subsequent increased accumulation of both
primary and secondary metabolites would be expected, although the relative
proportions of these would possibly differ. The increased potency of plants grown in
brighter conditions observed in this study supports the Carbon Nutrient Balance
hypothesis proposed by Bryant et al. (1983). This predicts that the increased net
124
photosynthesis results in an increased Carbon/Nutrient ratio within the plant. This

favours the development of carbon-based secondary metabolites.
100
80
Yield % of Max
60
40
20
0
0 20 40 60 80 100
Irradiance % of Max
Figure 5.9. The average monthly yield as a function of the glasshouse light level at the
beginning of flowering. On both axes the data was expressed as a percentage of the maximum
observed (r2 = 0.92, p < 0.001).
The requirement for such high levels of supplementary lighting is understandable. At

latitude 50ºC, ignoring effects of cloud cover, average solar radiation levels at sea level
are approximately half those encountered at 30ºC (Albuisson et al., 2006). Although
Cannabis sativa grown for fibre or seed is often planted at this latitude, the THC-
chemotype grown outdoors for its cannabinoid content is more commonly found at
latitudes of 30º or less (Small and Beckstead, 1973a,b). In addition to unfavourable
latitude, light transmission through a typical glasshouse roof is further reduced by at
least 30% (Heuvelink et al., 1995). Supplementary lighting has been increasingly used
in commercial glasshouses in the UK for food-crop production. Rarely do these deliver
light levels more than 12 W m-2, but recent installations of 25 W m-2 have been reported
(Vale, 2008). The installation of a lighting system delivering 55 W m-2 over an area of
5000 m-2, as used here, was highly unusual and possibly unique in the UK.
However this lighting level is less than that typically utilised by illicit growers. Evidence
from indoor UK cannabis-growing scenes of crime (private communication) show much
brighter lighting conditions. These appear to comply with the illumination levels
recommended in cannabis growing guides (Green, 2003), which commonly suggest the
use of one 600W high pressure sodium lamp per square metre of flowering crop.
125
Hough et al. (2003) quotes an experienced illicit UK cannabis grower as saying that “A
decent grower can quite easily get one gram of dried flower head per watt of lighting
used.” This refers to electrical energy consumed per unit area. It implies that the use of
one 600 Watt lamp per square metre could result in the production of 600 g m-2 of dry
mature cannabis flower head. Such illicit yields are regularly claimed (Rosenthal,
2001) and appear credible. Six hundred watt HPS lamps typically convert about 30%
of the electricity consumed into photosynthetically active radiation (Langton and Fuller,
2001). The irradiance in such a situation would therefore be 180 W m-2 PAR, i.e.
approximately twice that encountered by the pharmaceutical crop studied for this
thesis. This highlights the typical illicit growers’ desire for yield over uniformity. This
may be partly to meet consumer demand. However, it is extremely common for illicit
growers to be using ‘extracted’ (stolen) electrical energy, and as such energy
consumption and cost is often not a consideration.
5.5.3 Effect of Duration of Flowering Period on BRM and Cannabinoid

Yield
Between the sixth and eigth week of flowering the 33% extension in flowering duration
resulted in a mean cannabinoid yield increase of over 50% (Figure 5.10). In all clones
this extra two weeks flowering was clearly advantageous. A further 25% increase in
flowering period from eight to ten weeks resulted in a mean increase in THC yield of
31%, but for approximately half of the clones, including Sativex-related G1M1 and
G2M7, the observed THC yield increase was less than 25% and the economic benefit
of the extra two weeks flowering was doubtful. Paired t-tests showed that the mean
increases in yield, from six to eight weeks and from eight to ten weeks, for the twenty
five clones were highly significant (6-8 weeks p < 0.001, 8-10 weeks p = 0.032).
126
100
90
80
70
THC g/sq m
60 6 weeks
50 8 weeks
40 10 weeks
30
20
10
0
G50-5
G52-9
G52-5
G47-5
G45-3
G49-9
G49-8
G51-1
G1 M1
G44-13
G50-3
G43-12
G50-4
G44-18
G44-16
G46-4
G48-12
G46-14
G45-11
G42-1
G41-14
G47-8
G52-6
G48-13
Mean
G2 M7
Clone
.
Figure 5.10. The yield of THC achieved by each of the clones (n=5) after six, eight and ten
weeks in flower. For clarity, the clone lines have been sorted in order of descending THC
yields after ten weeks in flower. (Paired t-test 6-8 weeks p < 0.001 and 8-10 weeks p < 0.01).
The clones that clearly benefited from a longer ten week flowering period were
generally those reported to have an equatorial and subequatorial areas provenance. In
their natural habitat these would experience a longer growing season. However, as
with all cannabis varieties that are used for illicit purposes, these variety names are not
registered in the International Code of Nomenclature for Cultivated Plants, 1995
(Snoeijer, 2002). Thus true provenance cannot be authorised and caution is required
when referring to the properties of such varieties.
5.5.3.1 Effect of Duration of Floweringt Period on Ratio of THC and CBG in THC
Chemovars
Between the sixth and ten week of flowering, the mean average content of both THC
and CBG increased in the twenty five clones. The mean proportion of CBG in the
cannabinoid profile fell from 3.66% to 3.24% between the sixth and eighth week of
flowering and fell further to 2.66% in the tenth week. Both decreases were highly
significant (p = 0.0208 and p = 0.0003 respectively in two-tailed t-tests). This was
probably due to the fact that as plant development slowed, CBG was being converted
into THC faster by THC synthase than it was being renewed by
127
geranylpyrophosphate:olivetolate geranyltransferase. The relative proportions of CBG

and THC in a pharmaceutical product must meet a tight specification to ensure its
quality, safety and efficacy. This investigation shows that the ratio of these two
cannabinoids is not genetically fixed and that harvest timing can have a significant
effect.
Plant genetics were shown to have a greater impact on THC:CBG ratios than harvest
timings. For example, the THC:CBG ratio of clone G44-13 was approximately 1:200 at
all harvest timings whereas G48-12, G48-13 and G41-14 showed a ratio greater than
1:20. When the THC:CBG ratios from all three harvest date were combined (Figure
5.11) highly significant differences were observed between clones.
100
THC as % of THC+CBG
99
98
97
96
95
94
93
92
91
90
G1 M1
G2 M7
G44-13
G45-11
G45-3
G44-18
G50-4
G44-16
G50-5
G42-1
G51-1
G47-8
G46-14
G47-5
G43-12
G50-3
G49-8
G46-4
G52-9
G52-5
G52-6
G49-9
G48-12
G48-13
G41-14
* * *** ** ** ** *** ** * ** ***
Clone
Figure 5.11. A comparison of the mean relative proportions (±SD) of THC and CBG in twenty
five clones at three harvest dates. Analyses of variance (one-way) compared the proportion of
THC in each clone to that in the Sativex-dependent clone G1 (shown in red). (* p < 0.05, ** p
< 0.01, *** p < 0.001).
In some cases (G44-13 v G44-16 and G50-3 v G50-4 or G50-5), there were highly
significant differences in the THC:CBG ratios of clones derived from single varieties
(ANOVA, p < 0.01). This supports the data reported in Section 5.5.1 which indicates
that uniform drug feedstock is more likely to be produced by plants grown from clones
rather than seed. The especially high degree of variability in varieties G44 and G50
(provenance unknown) was possibly due to the parent-crosses having a higher degree
of heterozygosity.
128
5.5.3.2 Effect of the length of flowering period on profile of heterozygous chemotypes

with mixed THC/CBD profiles
Four weeks after being placed in twelve hour days, inflorescence development was
rapid and capitate stalked trichomes were becoming abundant. This suggested that the
plants were actively biosynthesising cannabinoids. Sampling commenced and was
continued for the following six weeks. By the tenth week in short day length, no more
viable stigmas were observed, indicating that inflorescence development and capitate
stalked trichome formation had ceased. Table 5.5 shows that there were minimal
increases in cannabinoid concentration in the last two weeks of the sampling period,
but prior to that, cannabinoid contents had escalated rapidly.
In clone G159/1 the CBD/THC ratio remained constant over the entire flowering phase.
However, in G159/9 and G159/16 there were significant proportional increases in CBD
content (regression p < 0.05). Conversely, in G159/11 and G159/12 the proportion of
CBD in the cannabinoid profile showed significant decreases with time (regression p <
0.05). The stable cannabinoid profile exhibited here by clone G159/1 suggests that
single clones of the mixed CBD/THC chemotype can be used as phytopharmaceutical
feedstocks. However, to produce such clones the plant breeder has the combined
challenge of identifying genetics that not only exhibit a desired cannabinoid profile but
also deliver this reliably and uniformly during the growing process. In the absence of
such clones, desirable cannabinoid mixtures can be only be achieved by blending
materials with differing profiles.
129
Mean Clone
Weeks
Cannabinoid
in 12 h G159-1 G159-9 G159-11 G159-12 G159-16
Content %
days
w/w (± SE) CBD as % of CBD+THC (± SD)
1.55 61.78 58.79 61.41 42.60 59.49
4
(± 0.08) (± 0.14) (± 0.70) (±0.37) (± 0.35) (± 0.49)
3.12 61.43 58.55 62.21 43.14 60.02
5
(± 0.69) (± 0.41) (± 0.29) (± 1.14) (± 0.63) (± 0.80)
5.69 62.03 59.65 61.20 42.95 61.19
6
(± 1.27) (± 0.85) (± 0.15) (± 0.22) (± 0.49) (± 0.54)
9.22 61.84 62.05 59.23 43.06 61.49
7
(± 1.22) (± 0.38) (± 0.38) (± 0.30) (± 0.44) (± 0.11)
11.15 61.53 62.79 58.91 41.86 62.21
8
(± 1.08) (± 0.09) (± 0.21) (± 0.57) (± 0.90) (± 0.58)
11.65 61.32 63.00 59.11 42.06 62.17
9
(± 0.92) (± 0.37) (± 0.95) (± 0.58) (± 0.05) (± 0.23)
11.95 61.99 64.16 59.60 41.08 62.29
10
(± 1.24) (± 0.53) (± 0.73) (± 1.22) (± 1.17) (± 0.20)
No Significant Significant Significant Significant

significant Increase Decrease Decrease Increase
Regression change
p > 0.05 p < 0.01 p < 0.05 p < 0.05 p < 0.05
r = 0.03 r = 0.96 r = 0.81 r = 0.79 r = 0.95
Table 5.5. The relative proportions of CBD and THC during plant development in five clones
derived from variety G159. Results are shown as the proportion of CBD expressed as % of
CBD+THC (± SD). The regression calculations test the significance of the changing proportion
of CBD and THC in each clone, between the 4th and 10th week in 12 h daylength.
5.5.4 Effect of Daylength on Plant Development and Cannabinoid Profile
5.5.4.1 Comparison of Twelve and Thirteen Hour Daylength Regimes
Despite their varying provenance, all the varieties included in this investigation
commenced to flower within ten days of being placed in a thirteen-hour day length.
This suggested that all had a ‘critical daylength’ of over thirteen hours. As shown in
Table 5.6 a daylength of thirteen hours occurs naturally at around August 18th, at a
latitude of 26ºN, and progressively later at more northerly locations. Had a Columbian
variety been included, perhaps originating for Santa Marta (latitude 11° north), it is
130
possible that it would not have flowered - as even at the summer solstice the day
length is just 12 hour 45 minutes. The critical day length for plants from this latitude
would be therefore expected to be substantially less than thirteen hours. It is possible
that the plant would have eventually have flowered in response to plant age rather than
daylength, as is common for tropical varieties.
All of the locations cited in Table 5.6 have been important in the breeding history of the
varieties currently developed for pharmaceutical use. Skunk #1 and subsequently
some of the other varieties within this test were originally bred using a mixture of
Columbian, Mexican and Afghan genetics (de Meijer, 1999; Clarke, 2001). Despite the
Columbian inclusion, all varieties appear to have inherited from their more northerly
ancestors an ability to flower readily in a thirteen-hour daylength. The CBD-rich variety
G5 was bred from landrace plants derived from the Black Sea coast in Northern
Turkey, where thirteen hours day length occurs in early September, a few days earlier
than in Southern England.
Latitude: - 11º 26º 36º 41º 51º

Climate: - Tropical Semi-tropical Temperate
Location: - Santa Marta, Monterrey, Mazar al sharif, Samsun, London,
Columbia Mexico San Fransisco and Turkey UK
Ketama*
Jul-15 12.43 13.35 14.25 14.50 16.10
Aug-01 12.35 13.19 14.05 14.20 15.24
Aug-15 12.28 13.03 13.37 13.49 14.37
Sep-01 12.19 12.38 12.59 13.06 13.34
Sep-15 12.11 12.18 12.27 12.28 12.40
Oct-01 12.01 11.54 11.48 11.45 11.37
Oct-15 11.53 11.34 11.19 11.10 10.45
Table 5.6. The falling late-summer daylength (hours.minutes) in a range northern hemisphere
cannabis growing areas. (* Contrasting locations at similar latitude in Afghanistan, USA and
Morocco.)
Although floral development of all plants commenced equally promptly in a twelve or

thirteen-hour daylength, the pattern of floral development differed greatly in the two
131
regimes. The visual assessments the degree of stigma senescence (Table 5.7)
showed that the rate of flower formation slowed more rapidly in the twelve-hour regime.
Assessed eight and ten weeks after being placed in short daylength, the mean
proportion of senescence stigmas was significantly higher in the twelve hour daylength
(paired t-test, p < 0.01, at both assessment dates). The difference was most dramatic
in clone line M84, which after ten weeks in a twelve-hour daylength, had completely
ceased to develop new flowers whereas the same clone in thirteen hours daylength
continued to flower prolifically.
Eight weeks in Short Days Tenth Week in Short Days
12 hr Day 13 hr Day 12 hr Day 13 hr Day

length length length length
Clone
M1 60 33 50 45
M6 60 38 80 45
M57 80 35 80 55
M59 70 50 35 10
M60 95 35 100 98
M61 99 38 100 75
M79 70 15 60 50
M82 20 4 30 15
M84 50 3 100 10
M87 30 4 25 5
Mean 63.4 ** 25.5 ** 66 ** 40.8 **
Table 5.7. A comparison of the proportion of senesced stigmas observed on ten clone when
induced to flower in daylengths of twelve or thirteen hours. Assessments were made eight and
ten weeks after the plants were placed in short daylength. ** Significant difference (p < 0.01,
paired t-test).
A twelve-hour day length occurs at all northern hemisphere latitudes in the last week of
September (Table 5.6), at around which time a cannabis crop grown for seed would
typically be harvested (Bocsa and Karus, 1998). Any female flowers formed after early
September would be unlikely to have sufficient light and warmth to produce viable seed
before the plant died. Moreover, pollen formation typically ceases before the end of
132
August after which time male plants typically die thereby not competing with the female
plants as they set seed. The formation of further female flowers would be a futile waste
of plant resources.
After eight weeks in short daylength there was no significant difference in the mean
height of the ten clones in twelve or thirteen hour days (paired t-test). After ten weeks
significant differences were observed in the height of five clones (ANOVA), and this
was most pronounced in vigorous F1 hybrids M82 and M87 (Figure 5.11). As a
consequence there was a significant difference in the mean height for all ten clones in
the two regimes (paired t-test p < 0.05). M82 and M87 were much taller than other
clones. Excessive height is a disadvantage in a glasshouse grown crop and the longer
daylength exacerbated this problem. Tall plants are difficult to support and to handle,
and are less easily examined for pest and disease.
300
250
Mean Height (cm)
200
12 hours
150
13 hours
100
50
0
Mean
M1
M6
M57
M59
M60
M61
M79
M82
M84
M87
* * *** *** *** *

Clone
Figure 5.12. Effect of Daylength on Plant Height ± SD (n = 20) ten weeks after induction of
flowering (* p < 0.05, *** p < 0.001, ANOVA for individual clones and paired t-test for the
overall mean).
Eight weeks after flower initiation, minimal differences were observed between the
botanical raw material yields of plants grown in twelve and thirteen hour daylengths.
The two F1 hybrid clones M82 and M87, that had shown large differences in height,
showed marked increases in yields at the longer daylength when harvested ten weeks
after flower initiation (Figure 5.13). However, overall the longer daylength did not
significantly increase the mean yield of the ten clones (paired t-test, p > 0.05).
133
1400
1200
1000
BRM g/sq m
800 12 hours
600 13 hours
400
200
0
M1
M6
M57
M59
M60
M61
M79
M82
M84
M87
Mean
* * ***
Figure 5.13. Effect of Daylength on Yield of Botanical Raw Material ± SD (n = 5 plants) ten
weeks after induction of flowering (* p < 0.05, *** p < 0.001, ANOVA).
The cannabinoid content of the higher yielding clones M82 and M87 was greatly
reduced in the longer daylength and this resulted in a marked decrease in mean
cannabinoid yield of the ten clones at both harvest dates (Table 5.8). Conversely the
clone lines M1 and M6, showed a possible yield benefit. When assessing the overall
effect of daylength on the mean cannabinoid yield of the ten clones, the paired t-tests
showed no significant advantage for either regime.
134
Clone 8 weeks in flower 10 weeks in flower

12 hr days 13 hr days 12 hr days 13 hr days
M1 57 79 70 73
M6 44 54 65 83
M57 64 64 51 87
M59 52 34 54 30
M60 41 45 36 39
M61 29 30 32 31
M79 50 53 43 49
M82 84 33 122 63
M84 34 35 55 53
M87 84 38 100 73
Mean 54 47 62 58
Table 5.8. Effect of Daylength on cannabinoid yield (g m-2), eight and ten weeks after induction
of flowering.
Daylength had highly significant effects on cannabinoid profile. Harvested eight or ten
weeks after being placed in a twelve or thirteen hour daylength, all ten clones had a
proportionally higher CBG content if grown in the thirteen hour regime (Table 5.9). The
mean difference was significantly higher after eight weeks (p < 0.01) and ten weeks (p
< 0.05, paired two-tailed t-tests).This was likely due to the fact that more new florets
and accompanying glandular trichomes were being formed in the thirteen hour
daylength, and these would be synthesising more CBG. Conversely plants in the
thirteen hour regime possessed proportionally less THCV (p < 0.001 after eight weeks
and p < 0.01 after ten weeks). This finding supports other results (not reported here)
that propyl cannabinoid synthesis is more rapid than that of pentyl cannabinoids. As a
consequence, in aging tissue the biosynthesis of THCV reaches completion before that
of THC.
135
CBG as % of CBG + THC THCV as % of THCV + THC

Daylength 12 h 13 h 12 h 13 h 12 h 13 h 12 h 13 h
Time in Short Days 8 weeks 10 weeks 8 weeks 10 weeks
M1 1.00 1.20 0.65 2.00 0.52 0.48 0.55 0.46
M6 0.77 1.40 0.81 1.25 0.65 0.57 0.73 0.53
M54 3.96 4.38 3.30 3.97 0.39 0.34 0.33 0.29
M57 0.69 0.98 0.45 0.82 0.42 0.34 0.51 0.35
M59 1.27 2.63 1.11 0.96 0.47 0.45 0.57 0.42
M60 3.30 4.57 2.51 2.59 0.54 0.52 0.54 0.48
M61 1.53 3.43 0.98 1.44 0.29 0.20 0.32 0.18
M79 1.55 3.78 1.29 1.57 0.43 0.37 0.45 0.37
M82 5.77 7.15 4.41 6.11 0.32 0.16 0.31 0.17
M87 4.10 6.55 3.29 3.40 0.39 0.35 0.32 0.37
Mean 2.39 3.61 1.88 2.41 0.44 0.38 0.46 0.36
p, t-test
0.0011 0.0176 0.0006 0.0015
(2 tailed)
Table 5.9. The effect of day length during flowering on cannabinoid profile. Results shown are
the proportion of CBG and THCV, expressed as a % of the CBG+THC or THCV+THC total, in
ten clones after eight and ten weeks in short daylength.
5.5.4.2 Comparison of Eleven and Twelve Hour Daylength Regimes
Whereas reducing the daylength from thirteen to twelve hours had a dramatic effect on
floral development, a further reduction to eleven hours only marginally accelerated the
cessation of new flower formation, Table 5.10.
136
Clone Eight weeks in Short Days Tenth Week in Short Days

11 hr 12 hr 11 hr 12 hr
M1 30 30 60 55
M6 35 40 60 60
M11 80 70 100 98
M16 50 38 100 98
M57 15 8 65 55
M58 28 30 45 55
M59 30 40 85 70
M60 95 80 99 100
M61 20 25 98 70
M82 8 10 23 20
M84 15 10 50 45
M87 10 12 40 30
Mean 35 33 69 63
Table 5.10. A comparison of the proportion of senesced stigmas on ten clones in twelve and
eleven hour daylength regimes when assessed eight and ten weeks after the induction of
flowering. Just one overall visual assessment was made for each clone.
F1 hybrids M82 and M87 showed very large differences in height when grown in
contrasting twelve and thirteen hour regimes. However, along with all other clones,
these showed no significant difference in height when grown for eight or ten weeks
(Figure 5.14) in eleven and twelve hour daylengths (ANOVA). Similarly, a paired t-test
revealed no significant difference in the mean heights of the ten clones in either
daylength.
The extra hour of day length resulted in significantly greater yields of most clones
(ANOVA) at both assessment dates. For simplicity, just the result of the latter
assessment is shown in Figure 5.15 (* p < 0.05 and *** p < 0.001). A paired t-test
showed the mean weight of the twelve clones to be greater in the longer day length at
both assessment dates.
137
250
H eig h t (cm ) 200

150
11 hour
100
12 hour
50
0
M1
M6
M 11
M 16
M 57
M 58
M 59
M 60
M 61
M 82
M 84
M 87
M ea n
Clone
Figure 5.14. Effect of Daylength on Plant Height ± SD (n=20) ten weeks after induction of
flowering. In a paired t-test there was no significant difference in the mean height of plants
grown in 11 and 12 hour days.
100
90
80
Y ie ld g /s q m
70
60 11 hours
50
40 12 hours
30
20
10
0
M1
M6
M11
M16
M57
M58
M59
M60
M61
M82
M84
M87
Mean
* * *** *** *** *** *** *** ***

Clone
Figure 5.15. Effect of Daylength on Yield of Botanical Raw Material ± SD (n = 20 plants) ten
weeks after induction of flowering. In the Analyses of Variance, the levels of significance were
shown as * p < 0.05 and *** p < 0.001. In a paired t-test, the mean height of plants in the 12
hours days was significantly higher p < 0.001.
When harvested after ten weeks in the flowering regime, all clones showed a lower
cannabinoid yield in the shorter daylength (Figure 5.16). The overall mean 22%
reduction in cannabinoid yield was highly significant, (p < 0.01, two-tailed t-test). This
reduction in yield could not be attributed to a prominent visible change in plant
138
morphology. Plants in the shorter eleven-hour daylength had received 9% less light
energy per day. Conversely, these plants spent 9% longer in darkness and would be
expected to have lost yet more energy as a result of a longer period of respiration.
This would seem a major contributing cause of the yield reduction in the eleven hour
daylength and is in marked contrast to the lack of yield difference observed when
comparing the twelve and thirteen hour daylengths. In the thirteen hour regime the
potential yield benefits of increased energy were not utilised by the plant.
70
Cannabinoid g/sq m
60
50
40 11 Hour
30 12 hour
20
10
0
M1
M11
M16
M57
M58
M59
M6
M60
M61
M82
M84
M87
Mean
**
Clone
Figure 5.16. Effect of Daylength on cannabinoid yield ten weeks after induction of flowering.
(Paired t-test, two tail ** p < 0.01).
The HPLC analysis method that was used to analyse the cannabinoid profile showed
limited sensitivity to CBG and THCV, and these proved to be below the detectable level
in some clones. Nevertheless, sufficient data was available to assess the proportion of
CBG, as a percentage of the CBG+THC+CBD total, in six clones (Table 5.11). In a
paired test, there was no significant difference in the mean proportion of CBG present.
139
CBG as % of Cannabinoid Total

Daylength
11 h 12 h
M57 0.45 0.90
M58 4.23 4.52
M59 1.31 1.33
M60 2.56 2.29
M84 1.73 1.67
M87 3.13 2.82

Mean 2.24 2.26
Paired t-test p = 0.98
Table 5.11 The effect of day length during flowering on cannabinoid profile. Results shown are
the proportion of CBG expressed as a % of the CBG+THC+CBD total, in six clones after eight
weeks in short daylength.
5.5.4.3 Review of the comparisons of plants induced to flower on daylengths 11, 12

and 13 hours.
Amongst the major findings, from the studies comparing the effect of daylength, it was
found that increasing the daylength from twelve to thirteen hours increased energy
consumption by at least 8%, but resulted in no beneficial increase in botanical raw
material yield. There was sometimes also an unwelcome increase on plant height.
Effects of increased day length on cannabinoid yield were variable, with some clones
showing a large decrease in cannabinoid production at the longer day length.
Conversely, reducing the day length to eleven hours saved energy by approximately
8% but resulted in an economically unacceptable and statistically significant 15 – 20%
mean reduction in plant yield (p < 0.01). There was also a significant reduction in the
mean cannabinoid yield (p < 0.01 after 10 wks in short days).
Placing indoor-grown cannabis in a short daylength to induce the plants to flower is

widely practised. The concept of flower initiation in response the arrival of the critical
daylength, and the part that phytochromes play in this are widely reported (Halliday and
Fankhauser, 2003). The effect of short daylength on inflorescence development is less
well studied. In addition to a critical daylength, at which point flowering is induced, this
140
study suggests that there is a previously unreported second critical daylength at which
development of further female flowers is inhibited.
In a glasshouse setting, where crops in various stages of floral development grow side-
by-side, it is not practicable to recreate the steadily shortening day length experienced
by outdoor grown cannabis. A standardised daylength has to be maintained once
flowering has commenced. The flowering daylength adopted can dramatically affect the
appearance of the inflorescence, with plants in thirteen hours day length continuing to
produce significantly more fertile female flowers for much longer than those in twelve or
eleven hour regimes (p < 0.05). An assessment of the proportion of senesced stigmas
in an inflorescence is often regarded as useful when deciding upon harvest date of illicit
crops. However, this investigation concluded that stigma senescence is not useful for
judging the cannabinoid content of crops grown for pharmaceutical purposes.
Eleven or thirteen hour daylength regimes are less economical in producing cannabis
for the production of cannabinoid-based medicines than the twelve hour regime initially
adopted. Future research to explore small adjustments to the adopted twelve-hour
regime may be beneficial. Research with ornamental pot-grown chrysanthemums
(short day plants) has shown that adjustments of just ten minutes to the daylength can
have a dramatic effect on the economic value of the plants produced (Langton and
Fuller, 2001).
Rather than simply altering the lighting regime to save costs, much greater savings
could be made by growing the crop outdoors. Whereas the glasshouse environment
provided uniform temperature, irradiance levels and daylength the outdoor environment
exposes the crop to variable conditions. If these variabilities were only found to affect
secondary metabolite yield this would be tolerable. The production of plants with a
significantly different secondary metabolite profile or poor quality would raise serious
concerns. The next chapter evaluates this possibility.
5.6 CONCLUSIONS
This chapter examined the development, and more especially its secondary metabolite
content, when grown indoors. The overall aim was to determine ways of identifying
growing methods and harvest timings that improved yield and crop uniformity when
growing medicinal cannabis crops indoors in the UK.
Specific tests with the Sativex® dependent variety G1 showed that cannabinoid profile
uniformity was significantly improved by growing plants from clones (cuttings) rather
than seeds.
141
The duration of the flowering period prior to harvest was seen to have a statistically
significant effect on cannabinoid profile when tested on a wide population of genotypes.
However, limited data suggested that the proportion of CBG in the cannabinoid profile
of Sativex-dependent varieties G1 and G2 had stabilized after eight weeks in flower.
The ratio of THC and CBD in heterozygous plants did not remain stable during floral
development, and this instability is one supportive reason for producing THC and CBD
for Sativex® from separate homozygous chemotypes.
When inducing and maintaining floral growth, a twelve hour daylength was shown to be
more energy efficient than eleven and thirteen hour regimes.
The most pronounced findings, from the research for this chapter, resulted from studies
with glasshouse supplementary lighting. Significant improvements were achieved in
the year round uniformity and yield of botanical raw material. Cannabinoid uniformity
and yield were similarly improved. It was shown that the irradiance levels at the
commencement of flowering had the greatest effect on yield. However, these
improvements were only achieved by the consumption of lighting energy levels well
above those typically used in the UK horticultural industry. One way of avoiding this
energy use would be to grow the crop outdoors. Rather than simply altering the lighting
regime to save costs, much greater savings could be made by growing the crop
outdoors. Whereas the glasshouse environment provided uniform temperature,
irradiance levels and daylength the outdoor environment exposes the crop to variable
conditions. If these variabilities were only found to affect secondary metabolite yield
this would be tolerable. The production of plants with a significantly different secondary
metabolite profile or poor quality would raise serious concerns. The next chapter
evaluates this possibility.
142
Chapter 6. The Outdoor Propagation of Phytopharmaceutical Cannabis
Chapter 6．The Outdoor Propagation of
Phytopharmaceutical Cannabis
6.1 INTRODUCTION
The concept of growing phytopharmaceutical cannabis outdoors was initially rejected
due to plant quality and security concerns. However, prior to this thesis, an
experimental outdoor trial was planted. This demonstrated that potentially high yielding
cannabis crops could be grown in southern England (Potter, 2004), the Sativex-
dependent clone G5 M11 producing 500 g m-2 of dry BRM. This was approximately the
same yield as that achieved in a glasshouse crop. The trial did not aim to explore the
maximum yield that could be achieved and with improved knowledge higher yields may
have been obtained. No attempt was made to analyse the effect of outdoor growing on
secondary metabolite profile. In the following two seasons further batches of the clone
G5 M11 were planted and crop establishment was good. However, both crops
encountered fungal spoilage by Botrytis in the last week before planned harvest.
Subsequent attempts to grow clones of the Sativex-dependent THC chemovar G1
failed, as the plants commenced flowering far too late in the season to produce a
worthwhile crop. However, another high-THC genotype did achieve high yields and
produced a potential phytopharmaceutical feedstock containing 11.8 ± 1.5% (w/w)
THC. This was close to the average potency of illicit sinsemilla cannabis circulating in
England in 2005 (Potter et al., 2008), most of which was thought to have been secretly
grown indoors. The first field trial also compared the establishment of plants grown
from rooted cuttings and from seed. Both methods proved satisfactory but
establishment from seeds was markedly less labour intensive.
From an agricultural perspective it therefore appeared that outdoor growing of cannabis

was feasible in England, although there was little data to demonstrate the potential
quality of this material. Concerns remained about the security of outdoor cannabis
crops as the medicinal cannabis crop has a potentially high cash value, and an unusual
cachet that appeals to the curious and mischievous. Hemp-fibre and seed crops of
Cannabis sativa grown in the UK have frequently been damaged by the public, who
have mistaken the crop for recreational cannabis. An all-female CBD chemovar crop
bears a closer similarity to recreational cannabis, both in appearance and odour, and
would be more likely to be tampered with. Any pharmaceutical crop is also possibly
143
vulnerable to the attention of those protesting against the animal testing associated
with the industry. An outdoor crop would require a very discrete and secure location.
In marked contrast to the field environment the glasshouse conditions were well
controlled, and harvest timings could be routinely anticipated and labour timetabled
accordingly. If the timing of harvest of field-grown crops could be predicted with similar
confidence there would be clear benefits. When commencing this research it was not
known how predictable the date of outdoor harvest would be, and how long crops could
be regarded as being at the appropriate stage. As only one crop had been successfully
harvested prior to this thesis, the predictability of harvest date was still little understood.
It was anticipated that a major difficulty of outdoor growing would be the requirement to
dry large quantities of material at the exact point when crops were deemed ready for
harvest. An early attempt to dry bulk quantities of outdoor grown crop resulted in
failure. The crops were dried using the same method employed for the glasshouse
crop, and much of the material was lost due to fungal spoilage. Exceeding the microbial
count threshold, as stated in the European Pharmacopoeia, is the most common cause
of rejection of plant material (Baier and Bonne, 1996). Prior to this thesis a crop was
successfully dried in the oast house, using a diesel fuelled oast furnace, in a similar
fashion to that traditionally used for hops. The lower humidity and rapid drying
conditions prevented fungal proliferation. This type of drying appeared to warrant
further evaluation. However, although the diesel used here is an acceptable fuel source
for the drying of food-grade hops, this would not be acceptable for a medicinal crop.
GAP Guideline 6.6 dictated that medicinal herb crops had to be dried naturally or using
methane, propane or butane (EMEA, 2006). For future drying experiments, alternative
GAP-compliant methods would be necessary.
A major advantage of growing outdoors is that it avoided the large energy cost incurred
in the glasshouse (> 500 kW hr per kilo of dry botanical raw material) and consequently
created a much smaller carbon footprint. At the start of this study it appeared that
outdoor growing was feasible and, if disease and crop drying difficulties could be
overcome, there were clear cost and environmental advantages. However, the effect of
outdoor growing on cannabinoid and terpene profile remained untested.
6.2 AIM and OBJECTIVES

Trials were performed annually over four consecutive years. The overall aim was to
enable the growth and harvest of good quality plants, with the same cannabinoid and
terpene profile as glasshouse crops. This necessitated an improvement in the
understanding of how outdoor growing affected plant development, secondary
144
metabolite production and susceptibility to pests and disease. Specific tests would
then be needed with the objective of overcoming these problems and improving
harvest and drying techniques. These objectives are described in specific details as
follows.
6.2.1 The effect of growing environment on female plant development

Unlike the glasshouse crop, which routinely took eleven weeks to grow from a rooted
cutting to fully mature plant, the outdoor grown plant took up to twice as long. As such,
specific observations were performed to ascertain how the development of outdoor
grown cloned all-female plants compared to that of a glasshouse crop and how
optimum time for harvest could be visually judged.

plant material and enriched trichome preparations made from them
As shown in Chapter 4, enriched trichome preparations could be produced from
glasshouse grown crops which captured most of the cannabinoid content of the raw
plant material. These were less bulky materials from which essential oil samples could
be steam-distilled to allow detailed terpene analysis. Trials were performed to compare
the secondary metabolite profile of freshly harvested plant material and enriched
trichome preparations. If the results indicated that glandular trichomes collected from
fresh material had the same secondary metabolite profile as the intact plant material,
this collection method might prove more efficient than harvesting, drying and then
processing whole plants.
6.2.3 The effects of harvest timing on secondary metabolite yield and

profile
One detailed trial aimed to assess how the cannabinoid and terpene yield and profile
changed during the latter phases of inflorescence maturation. This would help
ascertain the growth stage at which plants were most appropriately harvested to meet
the preset specification.
6.2.4 Comparison of the secondary metabolite profiles of glasshouse and

outdoor grown plants
In 2006 studies were performed to compare the terpenoid content and profile of
glasshouse and field grown plants, during inflorescence maturation. The objective was
to ascertain if and how secondary metabolite content was altered by growing
conditions.
145
6.2.5 Evaluation of outdoor pest and disease issues

The trial program would aim to find out what major pest and disease problems might
arise and assess their potential effect on plant quality and yield.
6.2.6 Evaluation of Crop Drying Methods

A range of drying methods was evaluated with two objectives. The first was to find a
way of speeding the processing of a single seasonal harvest. The second was to
assess how the drying method might be improved, to reduce fungal spoilage.
6.3 MATERIALS
Materials Source
Seed for cannabis varieties G1, G5 HortaPharm BV Schimkelhavemkade 1075
VS Amsterdam Netherlands
Seed for cannabis variety G159 Undisclosed commercial source.
Propane gas heater and accompanying Local hire. Location withheld for security
bottled gas. reasons.
Diplex portable handheld thermometer/ Diplex Ltd., PO Box 172, Watford
hygrometer (calibrated in-house) England, WD17 1BX
Carbolyte drying oven Carbolite - Parsons Lane, Hope, Hope
Valley. S33 6RB UK
Table 6.1 Propagation Materials and Equipment used in the field trials program to evaluate the
outdoor propagation of Cannabis sativa L.
The sieves, microscopes and analytical equipment were the same as those described
in Chapters Two to Four.
6.4 GENERAL AGRONOMIC METHODS
6.4.1 Seedbed Location, Preparation and Crop Establishment
Being mindful of security concerns, and to satisfy Home Office license conditions, the
crop was grown in a discrete high-walled garden. Seedbeds were prepared manually
using a fork and rake and metaldehyde-based molluscide pellets applied to the soil to
provide slug control.
Cuttings were raised as described in Chapter Five. These were placed outside the
glasshouse for seven days to acclimatise to outdoor conditions. They were then
planted in rows one metre apart at a density of four to six cuttings per square metre, as
146
found adequate in first year’s trial. Predictably, variable weather and soil moisture
conditions were encountered in the first weeks of growth and plants were watered by
hand when insufficient soil moisture was naturally available.
6.4.2 Field Trial Design

In the first three years single 100 m-2 plots of clone M16 were established, and specific
tests performed on this single clone. The 2006 trial compared the performance of the
G5 CBD chemovar at five harvest timings. Seven plots of each regime were laid in a
randomised block. A small area outside of the main trial was dedicated to some
heterozygous cultivars producing approximately equal quantities of THC and CBD. All
were derived from the single variety G159.
6.4.3 Soil Nutrition

All trials were given a light dressing of approximately 40kg/ha of fertilizer, containing
34.5% w/w nitrogen, before or soon after planting. However, no account was taken of
residual fertilizer already existing in the soil. In all years lush growth was observed and
no deficiency symptoms were observed.
6.4.4 Pest and Disease Monitoring and Management

In each season the trials were examined at least once per week throughout the
season. Pests and disease were identified and their severity noted. At no stage were
any chemical treatments applied. Rabbits caused extensive damage to a crop planted
in 2001 and subsequent trials were surrounded with protective fencing. Weeds were a
constant problem and were manually removed on a regular basis each year.
6.4.5 Harvest
In all cases plants were cut at the base below the lowest side-branch upon which any
floral material was observed. Plants were placed on clean sheeting prior to relocation
to an oast house for drying.
6.4.6 Crop Drying and Stripping

Crops were hung to dry on suspended wires in an oast house drying chamber.
Propane gas burners provided hot air and fans ensured that this moved freely within
the drying environment. Excess heat was allowed to escape through controlled
ventilation ducts in the ceiling above the drying chamber. Thermostats within the drying
chamber maintained drying conditions at the selected temperature, and this was
monitored (in compliance with GAP Guideline 4.6) using the oast house environmental
management recording system. Humidity was read with a hand-held meter.
147
In each of the three years 2003-2005, separate batches were dried in temperatures of
30º, 40º and 50ºC (± 2ºC). At the end of the drying period the relative humidity in these
respective regimes was recorded as being approximately 30%, 20% and 10%
respectively. Ten samples (approximately 1 g each) were taken at three-hourly
intervals during the drying process. Their moisture content of these was determined by
weighing the moisture loss when dried for 24 hours at 105ºC. Once judged by touch to
be sufficiently dry, the leaf and floral material in the oast was stripped from the stems
by hand. This material was relocated to a dehumidified store at 30-35% RH to
equilibrate. It was expected that in this environment the samples would settle to a
uniform predicted moisture content of 15 ± 2%, based on experience with glasshouse-
grown crops.
6.4.7 Assessment of Crop Development

The 2005 field trial was visited weekly and the height of thirty randomly selected plants
was measured. A typical plant would produce a dominant main inflorescence, and
many shorter side branches. Plant height was taken as the height of the top of this
central inflorescence as measured from the soil surface. In the following year’s trial,
subjective weekly assessments were made of the state of development of the
inflorescence, from the appearance of the first stigmas through to complete stigma
senescence. An assessment was also made of the overall proportion of senesced
stigmas in each of the seven replicates. To compare the pattern of development of the
field and glasshouse-grown plants, the same observations were made on the
inflorescences of three consecutive crops of the same clone.
6.5 Secondary Metabolite Purification and Analytical Methods
6.5.1 Production and Collection of Enriched Trichome Preparations

Ten litres of crushed ice were placed in a fifty litre bucket and twenty litres of tap water
added. Seven plants were collected at random from the field trial. Approximately ten
litres of foliar and floral tissue were stripped from the plants. These were promptly
plunged into the iced water and the temperature allowed to stabilize at below 1ºC. The
mixture was agitated with a food mixer at maximum speed for ten minutes. The
resultant slurry was then poured through 220 µm and 25 µm ‘Bubblebag’ sieves. The
material settling on the 25µm sieve was placed in a 100ml Duran bottle and moved to a
refrigerator at 4ºC. The procedure was repeated until all the material from the seven
plants had been processed. A subsample was examined, using a Brunel MX3 binocular
microscope, to check the efficiency of the process. Over 90% of the solid material
present was judged to be glandular trichome resin heads, the remaining proportion
148
consisting of capitate trichome stalks, cystolythic trichomes and fragments of mesophyl

tissue. The ETP was then relocated to a deepfreeze at -20ºC prior to analysis.
6.5.2 Steam distillation of trichome rich preparations

Samples were distilled at Botanix Limited, Hop Pocket Lane, Paddock Wood, Kent, UK
according to the company’s Standard Operating Procedure for the Determination of
Volatile Hop Oil (Institute of Brewing Method). In compliance with this method samples
were sufficiently thawed within their sealed Duran 200ml bottles to enable the contents
to be decanted into individual one-litre round bottom flasks with a B55 neck. A few anti-
bumping granules were added to the flasks before connecting each one to a British
Pharmacopoeia still, using a B55/34 adaptor. The flasks were then heated using a
heating mantle, and the contents distilled for three hours. Using this method Howard
(1970) showed that the extraction of essential oils from hop lupulin (glandular
trichomes) was reliably complete in this time. During this period the flow of condensate
was controlled to cause minimum disturbance to the oil in the trap. At the end of three
hours the volume of oil was recorded and then decanted for analysis by GC
(Appendix 2).
6.5.3 Steam distillation fresh foliar and floral material

Two kilograms of each sample of fresh material and five litres of water were placed in
ten litre round bottomed flasks and a few anti-bumping added. The flasks were then
connected to the still using a B55/34 adaptor. The mixture was boiled for two hours,
after which time no further volatiles were seen to be condensing within the stills.
During this period the flow of condensate was controlled to cause minimum disturbance
to the oil in the trap. At the end of two hours the volume of oil was recorded and
decanted. These oil fractions were analysed for qualitative terpene content by GC at
Botanix Ltd. The quantitative content of a small range of terpenes within this mixture
was determined by GC at GW Pharmaceuticals Ltd (Appendix 2).
6.6 Statistical methods

Standard deviation, analyses of variance, and regression calculations were performed
using Microsoft Excel 2003 software.
6.7 RESULTS and DISCUSSION
6.7.1 Observations on Crop Establishment and Plant Development

Crops were successfully harvested over five seasons. A summary of planting dates,
flowering dates and yields are shown in Table 6.2. Planting dates ranged from 12th
149
May to 25th June and yields varied between 451 and 728 gm-2. No clear correlation
was observed between planting date and final yield. However, yields data needed to
be made with caution as the trials differed in their location, planting density, soil fertility
and harvest dates and as expected weather conditions varied between seasons
(Appendix 4). To put the observed yields into perspective, clinical trials with Sativex®
indicated that when patients were able to self-titrate, for the symptoms of MS and
neuropathic pain, the mean daily consumption of CBD was 25 – 40 mg day-1 (9 - 14 g
year -1) in combination with a similar dose of THC (Wade et al., 2004, Barnes, 2006).
This suggests that one square metre of the CBD-chemovar could have provided
sufficient feedstock for one patient for one year (not allowing for losses during
extraction and formulation). If approved as a prescription-only medicine, the entire UK
demand of CBD for this medicine could theoretically have been met by a crop area of
just a few hectares.
Year 2000* 2003 2004 2005 2006

Cultivar(s) M11 M16 M16 M16 M16
Planting Date 12th May 11th Jun 25th Jun 23rd Jun 23rdJun
First Flowers No record 1st Sep 22nd Aug 25th Aug 22nd Aug
17th Sep to
Harvest 6th Oct 8th Oct 9th Oct 10th Oct
15th Oct*
BRM g/sq m 502 598 451 728 609*
CBD % w/w 6.0 8.8 _ 7.0 7.0*
CBD g/sq m 30 53 _ 50 33
Table 6.2 Summary of Agronomic and Yield Data from field trials performed between 2000 and
2006. *The 2000 trial was performed before commencement of this thesis, and the data is
included for comparison.
In all the field trials, when grown from cuttings, the CBD crop gained little in height in
the first four weeks after transplanting. Mid-June planting appeared to be sufficiently
early for good crop establishment and no obvious benefits were observed in planting in
May. In each year, irrespective of weather conditions and planting date, the maximum
rate of vegetative growth was observed in the period mid-July to end August. This is
150
demonstrated in Figure 6.1, which shows the results of the weekly monitoring plant
development in the 2006 trial. The sigmoidal growth curve is similar to that reported in
a monecious hemp crop grown from seed (BÓcsa and Karus, 1998), although flowering
occurred much earlier in hemp as the variety had been bred to mature earlier in the
season.
160
140
120
Height cm
100
80
60
40
20
0
16.38 16.32 16.22 16.08 15.51 15.27 15.08 14.44 14.19 13.53 13.26 12.59 12.31 12.04
23,Jun 2,Jul 9,Jul 16,Jul 23,Jul 31,Jul 06,Aug 13,Aug 20,Aug 27,Aug 03,Sep 10,Sep 17,Sep 24,Sep
Daylength(Hr.Mn)
Date
Figure 6.1 The mean height (± SD) of G5 M16 crop as observed at weekly intervals in 2005.
Thirty plants were measured on each occasion. Data points are shown as square symbols during
the establishment and vegetative phase. Data points are shown as triangle during the flowering
(generative) phase.
Observed during 2003-2006, flowering always commenced between 22nd August and
1st September (Table 6.2). As with the glasshouse crop, the increase in height almost
totally ceased seven to fourteen days after the appearance of the first stigmas. The
final height of the crop varied from year to year, being clearly influenced by the growing
conditions and soil nutrition. In 2005 the final crop height was 1.3 metres but the same
clones grown in the same location in 2006 reached 2.1 metres. This was likely due to
marked difference in mid-summer temperatures and sunshine-hours (Appendix 3)
experienced, the conditions in Jun and July being markedly brighter and warmer. In
four trials CBD yield was measured. Yields varied from 30 to 53 gm-2. However, none
of the trials were designed with the aim of achieving the maximum possible yield.
In 2006 floral development was observed in detail. Table 6.3 describes the stage of
inflorescence development and stigma senescence during the flowering period, as
recorded weekly between August 14th and October 8th. The table also compares the
151
pattern of development with three consecutive crops of the same cultivar when grown
in the glasshouse. As reported in Chapter 5, the proportion of senesced stigmas within
the inflorescence gives the most useful measure of the state of maturity. Figure 6.2
compares the pattern, speed and duration of inflorescence development and
maturation and showed that outdoor and indoor crops of variety G5 exhibited a very
similar pattern. This was despite the very different day lengths and growing conditions.
In each year’s trial, floral development was rapid from mid-September into the first days
of October, when development would cease. This cessation of growth was recorded in
some detail in the 2006 trial. Stigma senescence was observed to have commenced
just before 23rd September and to have been almost complete fourteen days later.
The 2006 outdoor crop yielded well, despite the fact that mean temperatures were
approximately 8°C lower (16.9°C – 17.3°C, Appendix 3) than the temperatures in the
glasshouse (25°C) during the August – September flowering period. Measured ten
kilometres away, at the glasshouse site, daily total radiation levels in autumn 2006
were close to those predicted for this latitude using a UK global radiation algorithm
(Hamer, 1999). This showed total daily radiation falling from 6.8 to 3.5 MJ m-2 d-1 over
the flowering period. This was slightly above that encountered by a glasshouse
production crop, which encountered a winter minimum 3.4 MJ m-2 d-1 rising to a
summer maximum 5.8 MJ m-2 d-1. Even though the indoor crop received supplementary
lighting, the glasshouse roof structure reduced the light transmission levels by
approximately a third. The glasshouse crop also experienced a shorter period of
daylength (12 hours) throughout the flowering period, whereas the field crop
experience longer daylengths during the first weeks of flowering. The higher radiation
level at the commencement of outdoor flowering may be especially pertinent.
Glasshouse studies (Chapter 5) showed that the greatest cannabinoid yields were
achieved in crops harvested in August. These had encountered the brightest light
conditions of the summer solstice during the very early phase of floral development.
152
Observation Timing and Pertaining Daylength Development stage of both indoor and
Indoor Crop Outdoor crop outdoor crops.
Week in Day Day

Date Inflorescence
short length Length Stigma Development
observed Development
days (Hr.mn) (Hr.mn)
0 12.00 13th Aug 14.40 Florets absent Stigmas absent

First florets
1 12.00 20th Aug 14.16 First stigmas visible
formed
Rapid formation Fertile stigmas
2 12.00 27th Aug 13.46
of new florets common
Rapid formation Prolific stigma
3 12.00 3rd Sep 13.23
of new florets formation
Rapid formation Prolific stigma
4 12.00 10th Sep 12.56
of new florets formation
Fewer fertile stigmas
Floret formation
5 12.00 17th Sep 12.28 observed. Older
slows
stigmas senesced.
Development Rapid stigma
6 12.00 24th Sep 12.01
almost ceased senescence
Development Senescence near
7 12.00 1st Oct 11.34
ceased complete
Development
8 12.00 8th Oct 11.11 Senescence complete
ceased
Table 6.3 A comparison of the pattern of inflorescence development and stigma senescence in
indoor and outdoor crops of cannabis chemovar G5. The stages of development of the
glasshouse crop are shown from the point at which the plants are moved into a 12hour
light/12hour dark until they are routinely harvested eight weeks later. The field crop
development is shown from mid-August, just before stigma formation commenced.
153
100
% Stigma Senescence 80
60 Glasshouse
Field
40
20
0
4 5 6 7 8 _
10th Sep 17th Sep 24th Sep 1st Oct 8th Oct 15th Oct
Weeks in Flower (Glasshouse)
Assessment Date (Field)
Figure 6.2 A comparison of the pattern of stigma senescence (%, ± SD, n = 7) in 2006 field trial
plants between 10th September and 15th October with that observed in five consecutively grown
routine indoor crops of the same variety (± SD, n = 5).

plant materials and enriched trichome preparations made from them
The 2005 trial incorporated a study to compare the essential oil profile of steam distilled
fresh mature cannabis and steam distilled ETP made from the same material. An
essential oil fraction, produced by steam distilling foliar and floral material from ten
plants taken at random from the crop on October 4th, yielded 7.7 ml m-2 (7.0 g m-2) The
approximate CBD potency of plants analysed from the same crop was 7% w/w,
indicating a CBD yield of 50 g m-2 (500 kg ha-1). It is useful to note that the essential
oil:CBD ratio was therefore approximately 1:7 (w/w).
The previous day (Oct 3rd) ETP was made from fresh G5 M16 plant material collected
from the same field trial. A similar batch was made from glasshouse plants at the same
state of maturity. These were immediately frozen and steam distilled alongside the
fresh plant material. The essential oils produced were analysed by GC at Botanix Ltd.
The results are shown in Table 6.4. The very similar terpene profiles of the fresh plant
material, and the ETP made from it, suggested that the complete range of
monoterpenes and sesquiterpenes had been equally well captured during the making
of the enriched trichome preparation. If the results of this single test were to be
routinely replicated, this would indicate that the terpene profile of a fresh cannabis
sample could be fairly assessed by studying an ETP made from that material. ETPs
154
made in this way are much more conveniently stored pending analysis. Being more
concentrated they enable more minor constituents to be quantified, as demonstrated by
trans-nerolidol in this test. Steam distillation further concentrates these terpenes.
Field Glasshouse
Fresh Leaf and Enriched Trichome Enriched Trichome
Flower Preparation Preparation
α-pinene 7.7 7.5 9.9
β-pinene 4.6 4.4 4.2
Myrcene 53.2 41.9 38.4
Limonene 6.9 7.9 8.3
β-ocimene 13.4 13.0 4.7
Linalol 1.5 1.1 2.6
α caryophyllene 2.6 4.3 5.4
β caryophyllene 10.3 17.0 22.7
Trans-Nerolidol <1.0 3.1 1.4
Table 6.4 A comparison of the terpene profile of freshly harvested fully mature field grown
cannabis leaf and flower material (cultivar G5 M16) and ETP made from the same fresh
material (2005 Field Trial). Also included is ETP made from similar mature glasshouse grown
material. The data show the relative peak areas when assessed by GC at Botanix Ltd. In each
case one batch of material was analysed.
The terpene profiles were seen to be dominated by the monoterpene myrcene and the
sesquiterpene trans-caryophyllene. Together they accounted for at least 60% of the
total peak area. The limited data looking at the cannabinoid, myrcene and trans-
caryophyllene ratios of cannabis inflorescences, and the enriched trichome
preparations made from them, indicated that the ratios were maintained during the
making of these preparations (Table 6.4). If enriched trichome preparations could be
made in bulk from freshly harvested field grown crops, they would avoid the
requirement for crop drying. They would also provide a greatly enriched feedstock for
drug production. A similar process is sometimes adopted in the hop processing
industry. Hop glandular trichomes (known as lupulin) are collected by agitating and
sieving deep-frozen material.
It is interesting to note that the essential oil yield of 7.7 ml m-2 achieved here is much
higher than the 1.8 ml m-2 previously recorded from unpollinated Cannabis sativa
(Meier and Mediavilla, 1998).
155
6.7.3 The effects of harvest timing and growth stage on yield and
cannabinoid profile
6.7.3.1 Botanical Raw Material Yield
G5 was believed to be a uniquely high-yielding high-CBD chemovar, derived from

Turkish parents naturalised at 41º N. When grown in the UK at 51º N this commenced
flowering in the fourth week of August, in response to a phytochrome-mediated
hormone switch, when the day-length was about 14 hrs:30 mns. At 41º N, this day-
length would have occurred in the last week of July. This suggests that, if grown in
Turkey, flowering would have commenced three weeks earlier. However, by the last
week of September, the day-length would have been 12 h in both locations. Research
findings described in Chapter 5 indicated that phytochrome-controlled systems also
appear to induce the cessation of floral development. Floral development would
therefore have probably ceased in both locations around the time of the autumn
equinox at the end of September. Outdoor growing of this variety might be expected to
be more successful if G5 production was moved closer to 41º N. The flowering period
would be extended and conditions would likely be less conducive to the proliferation of
Botrytis. However, the logistics of importing the crop would add to licensing difficulties
and costs.
In the 2006 trial, where five harvesting dates were compared (Figure 6.3), there was no
significant difference in the yields of raw material harvested weekly between 17th
September and 15th October (ANOVA). This appeared to be at least in part due to
early-harvested crops having a higher proportion of foliar material. Later-harvested
crops had a higher proportion of floral material but crop weight did not show a
corresponding increase. Much of the foliage in the older crops had senesced and
abscised by the time of harvest. It is likely that many of the primary metabolites within
the older foliage had been translocated to the developing inflorescences in the last
weeks of growth.
156
700
600
500
g / sq m
400
300
200
100
0
17th Sep 24th Sep 1st Oct 8th Oct 15th Oct
Harvest Date
Figure 6.3 Yield of Botanical Raw Material in the 2006 trial showing the effect of planting date
and harvest date. The results are the mean dry weights (gm-2 ± SD) harvested from seven
replicates.
6.7.3.2 CBD Concentration
Data comparing the CBD content of crops harvested between 2000 and 2005 showed
a range of approximately 6-9% w/w. The 2006 trial studied the effect of harvest date on
potency, and there was a clear linear upward trend from an initial content of 2.3% to
7.0% w/w (Figure 6.4). Analyses of variance showed highly significant weekly
increases in potency between 24th September and 8th October (p < 0.01). A regression
calculation, examining the correlation between days in flower and % CBD content
throughout the five assessments gave a highly significant upward linear trend,
p = 0.0028, R2 = 0.983.
157
9.0
8.0 y = 1.2106x + 0.8606
CBD Potency % w/w

7.0
6.0
5.0
4.0
3.0
2.0
1.0
0.0
Harvest Date
Figure 6.4 The effect of harvest date on cannabidiol concentration in Botanical Raw Material in
the 2006 trial. The results are the mean % CBD (± SD) content of samples from each of the
seven replicates, as measured by GC. Regression model, n = 7, p = 0.0028, R2 = 0.983.
6.7.3.3 CBD Yield
The yield of CBD, expressed as gm-2, is shown in Figure 6.5. A highly significant linear
upward trend is observed from 8gm-2 on 17th September to 33g m-2 on October 15th.
(Regression p = 0.0011, R2 = 0.994).
40
y = 6.0155x + 2.4709
35
30
CBD Yield g/sq m
25
20
15
10
5
0
Harvest Date
Figure 6.5 The yield of CBD in the 2006 trial, showing the effect of harvest date. The results
are the mean CBD yields (gm-2 ± SD) produced in each of the seven replicates. Regression
model, p = 0.0011, R2 = 0.994
158
Despite there being no clear increase in the dry matter yield from the mid-flowering
phase onwards, the 2006 trial showed cannabinoid yield increasing significantly in the
first two weeks of October (ANOVA, p < 0.01). Of the acetyl CoA, NADPH and ATP
generated by photosynthesis during this phase of growth, the plant appeared to be
diverting a greater proportion to secondary metabolite biosynthesis. Part of the slowing
in biomass increase during the flowering period is possibly due to the higher energy
requirement for the production of floral tissue. Hemp crops were calculated to have a
radiation efficiency of between 2.0 and 2.3 g/MJ PAR for vegetative growth, reducing to
between 0.6–1.2 g/MJ PAR after flowering (Struik et al., 2000). The biosynthesis of the
cannabinoids and terpenes would have a high energy requirement per unit weight. The
energy required to biosynthesise the monoterpenes myrcene, pinene and limonene
(C10H16) has been calculated to be a factor of 3.5 greater than that required for the
same weight of glucose (Gershenzon, 1994). Although the energy requirement for
CBDA and THCA biosynthesis is not known, metabolites with a similar C:H:O ratio
have an slightly lower energy cost than the monoterpenes, due to the lower state of
reduction. Additional energy would be utilized in the formation of the complex trichome
structure, which contains cellulose, cutin and a variety of enzyme proteins. Secondary
metabolites formed in the last days before harvest may have been increasingly
synthesised using metabolites released and translocated from older rapidly senescing
tissue.
Although the cannabinoid yield showed a continuous and significant increase up until
the final harvest date in mid-October (p = 0.0011), the quality of the material
deteriorated markedly as a result of fungal infection. Removal of all substandard
material would have negated much of this yield increase.
6.7.3.4. Effect of Harvest Date and Growth Stage at Harvest on Cannabinoid Profile
Analysis of the BRM by GC from the earliest harvest timings showed CBD to be the
only cannabinoid present in detectable quantities in plants of the G5 chemovar.
Meaningful comparison of cannabinoid profiles was only achievable by studying the
content of the much more potent enriched trichome preparations made at each
harvest-timing. CBD, THC and CBG were present in measurable quantities in all
samples. The ratios of these are shown in Table 6.5.
159
Harvest Date 2006 17th Sep 24th Sep 1st Oct 8th Oct 15th Oct
CBG as % of 1.59 ±
5.61 ± 0.11 5.31 ± 0.12 3.30 ± 0.18 2.31 ± 0.12
CBG+CBD (± SD) 0.01
THC as % of 3.85 ±
4.35 ± 0.20 4.27 ± 0.13 4.08 ± 0.11 3.93 ± 0.04
THC+CBD (± SD) 0.01
Table 6.5 The changing proportions of CBG and THC with respect to CBD (Mean ± sd) in
enriched trichome preparations produced from plants harvested at weekly intervals between 17th
September and 15th October. Just one ETP sample was made on each date after bulking together
one plant from each of the seven replicates. Three subsamples of each preparation were
analysed.
Results showed the importance of harvesting the crop at the correct time to achieve the
desired cannabinoid profile. Table 6.5 showed that the CBG:CBD ratio fell rapidly
during the last weeks of growth, the CBG proportion of the CBG + CBD total
decreasing steadily from 5.6% on September 17th to 1.6% on October 15th. The final
value was close to that typically found in glasshouse crops at harvest. Forty seven
consecutively processed batches of glasshouse-grown high-CBD chemovar were
analysed as part of the routine growing operation. The mean proportion of CBG in the
CBG + CBD total was 1.78% ± 0.21% SD. CBG is a cannabinoid with known
pharmacological activity (Formukong et al., 1988). Excessive quantities of this minor
cannabinoid in a CBD-based medicine would result in batches of feedstock being
rejected as having an unacceptable specification. To achieve the same low CBG:CBD
ratio as the glasshouse crop, the results of the 2006 trial suggested that the outdoor
harvest would have had to have been delayed until mid-October. However, Botrytis
rendered the quality of this crop unacceptable before this date. The THC:CBD ratio
also showed a downward but less pronounced trend over the harvest period. In the
above-mentioned batches of glasshouse-grown CBD chemovar the mean proportion of
THC, expressed as a percentage of the THC+CBD total, was 3.89 ± 0.16% SD
(n = 47). This is very close to the ratio found in the field grown plants in the later
harvests (3.85 ± 0.01, Table 6.5). Homozygous high-CBD chemovar plants like G5
produce only CBD synthase and no THC synthase (de Meijer et al., 2003). CBD
synthase however always appears to produce a small amount of THC, as well as CBD.
It is postulated that CBD synthase is present within CBD chemotypes in more than one
isoform, and these variably affect the final THC:CBD ratio. The ratio of these two
cannabinoids in commercial hemp varieties is reported to be typically around 1:20
(Mechtler et al., 2004, Hillig and Mahlberg, 2004). The parental lines used in the
breeding of G5 were selected on the basis of several characteristics including a
160
THC:CBD ratio as low as possible. As a consequence, the ratio in G5 plants was lower
than most commercial hemp varieties (deMeijer, pers comm), decreasing from 1:23 in
the early harvest to 1:26 over subsequent weeks (tentative n = 1 data only). This
apparent decrease in later harvested plants was possibly due to isoforms of CBDA
synthase being present which differed in their ability to function in the increasingly
autumnal growing conditions. However despite this, as the CBD content rose to 5.8 %
and 7.0 % w/w in the final weeks (Figure 6.4), the corresponding THC content rose to
0.23 % and 0.27 %.
As designated under the Misuse of Drugs Act Cannabis sativa can be licensed only for
research or "other special purposes" providing the Secretary of State is of the opinion
that it is in the public interest to do so. The only "special purpose" currently recognised
is the cultivation of EU-approved varieties of hemp for commercial and industrial
purposes. All such EU-approved varieties have THC contents below 0.2%. When this
program of field trials commenced, the maximum THC content permitted in floral
samples of licensed hemp crops was 0.3% w/w. This was subsequently reduced by
the European Union (EU regulations (VO (EG) Nr. 1420/98)) to 0.2%. Although a
minimum content of 0.3% was readily achievable by chemovar G5, the lower limit of
0.2% may be too low for the G5 chemovar to qualify for an EU subsidy payment, and
this would have a small effect on the cost of outdoor growing. Growing a variety
containing more than 0.2% THC is not illegal, but does necessitate a dispensation from
the Home Office Licensing. Enquiries to the Home Office (Drugs Licensing) indicate
that such an application would be met favourably (Evans, 2008).
6.7.4 Effect of Growth Stage and Harvest Date on Essential Oil Profile
The 2006 trial compared of the terpene profiles of field grown cannabis plants
harvested at weekly intervals between 17th September and 15th October. This was
achieved by first steam distilling an enriched trichome preparation made from fresh
botanical raw material collected on each date. Between 30 ml and 100ml of enriched
trichome preparation was produced from each, but this contained a large unmeasured
volume of residual water. It is not possible to quantify the essential oil concentration of
each sample on the basis of the non-aqueous fraction only. When distilled these
produced between 0.25 ml and 2.1 ml of essential oil and each was analysed by GW
Pharmaceuticals using GC. The relative peak areas for the twenty most abundant
terpenes detected are shown in Table 6.6. These twenty accounted for more than 97%
of the total peak areas. The monoterpene:sesquiterpene ratio of each of the essential
oil samples showed a weekly increase. A regression calculation showed the weekly
linear increase in the proportion of monoterpenes to be strongly correlated with time (p
161
= 0.0052, R2 = 0.973, y = 0.844x + 0.6219). All individual sesquiterpenes, of which

trans-caryophylene was the most dominant, showed a decreasing presence in the
overall terpenoids mixture.
The data show that of the monoterpenes, myrcene was by far the most dominant. It
was notable that when expressed as a % of the total peak area, the proportion of
myrcene more than doubled over the sampling period. Limonene and beta-ocimene
also showed clear though less dramatic increases. The pinenes conversely showed a
possible downward trend.
162
Harvest Date 2006

% of Peak Area
Monoterpenes
Alpha-pinene R 3.49 5.27 4.25 3.13 3.24
Alpha-pinene S 4.30 2.35 3.63 3.78 3.00
Beta-pinene 3.32 3.50 3.33 3.18 3.00
Beta-Myrcene 22.63 36.30 40.86 46.02 51.48
Limonene 2.75 3.91 4.31 4.68 5.07
Beta-ocimene 2.64 4.13 4.04 4.63 4.76
Sesquiterpenes
t-Caryophyllene 37.44 29.10 26.57 23.97 20.17
Bergotamene 0.34 0.18 0.16 0.12 0.08
Humulene 11.31 8.57 7.60 6.71 5.67
Aromadendrene 0.67 0.34 0.31 0.19 0.14
Selinene 1.17 0.63 0.52 0.36 0.29
Anon 0.84 0.46 0.41 0.26 0.21
Z,E Farnesene 0.18 0.09 0.08 0.06 0.07
alpha farnesene 1.70 1.77 1.44 1.25 1.11
alpha Gurjunene 0.17 0.15 0.11 0.09 0.07
Bisabolene 0.78 0.42 0.37 0.22 0.16
Nerolidol 0.66 0.62 0.38 0.26 0.23
Caryophyllene Oxide 4.13 1.70 1.29 0.91 1.05
Diepicedrene-1-oxide 0.89 0.38 0.28 0.17 0.19
Alpha-Bisabolol 0.59 0.12 0.07 0.02 0.02
Total Monoterpenes 39.12 55.46 60.41 65.42 70.55
Total Sesquiterpenes 60.88 44.55 39.59 34.58 29.45
Table 6.6 A comparison of the terpene profile of steam-distillates of enriched trichome

preparations made from freshly harvested plants on five dates between 17th September and 15th
October 2006. One bulked sample was analysed on each date.
Although the relative peak area data in Table 6.6 show very clear changing ratios
between individual terpenes, and between the overall ratios of monoterpenes to
sesquiterpenes, the true content of each of the twenty terpenes could not be accurately
gained from this data. Quantitive data was generated for eight terpenes for which
analytical standards were readily available. The results are shown in Table 6.7.
163
Harvest Date 17th Sep 24th Sep 1st Oct 8th Oct 15th Oct
α-Pinene 9.18 8.44 7.82 7.09 6.18
β-Myrcene 40.45 58.16 61.00 70.01 74.30
Limonene 2.89 2.61 3.84 2.82 2.84
Linalool 0.40 0.60 0.35 0.13 0.26
trans-
32.36 21.60 19.52 15.37 12.68
Caryophyllene
α-caryophyllene 9.18 5.93 5.27 4.02 3.30
Caryophyllene
4.14 1.52 1.46 0.02 0.03
Oxide
t-Nerolidol 1.39 1.13 0.74 0.53 0.41
Ratio Myrcene/
1.25 2.69 3.13 4.55 5.86
t-caryophyllene
Table 6.7 Ratio of eight terpenes in steam distilled enriched trichome preparations made from
freshly harvested field grown plants of cultivar G5 M16. The result for each terpene is
expressed as a weight percentage (% w/w) of the total within each column. The table also
shows the ratio of myrcene (the dominant monoterpene) and trans-caryophyllene (the dominant
sesquiterpene).
The qualitative peak area data in Table 6.6 gives a valuable though not accurate
measure of the ratios between the individual terpenes. Comparing quantitative data, as
in Table 6.7, is more accurate but limited by the availability of reference standards.
The two analytical methods are compared in Table 6.8, which shows the
myrcene/trans-caryophyllene ratios when calculated from Relative Peak Area and
quantitative w/w data. Each is expressed as a % of the October 15th value. Whether
determined from the peak area data or the w/w data, the ‘normalised’ values give very
similar results with the myrcene/trans-caryophyllene ratio doubling between September
24th and October 15th.
164
Harvest
Date 17th Sep 24th Sep 1st Oct 8th Oct 15th Oct
Myrcene/t-caryophylene
Actual Value
Peak Area 0.60 1.25 1.54 1.92 2.55
w/w 1.25 2.69 3.13 4.55 5.86
th
Normalized Value Shown as % of 15 Oct Value
Peak Area 24 49 60 75 100
w/w 21 46 53 78 100
Table 6.8 Comparison of myrcene/trans-caryophyllene ratios when calculated from Relative

Peak Areavalues and w/w data.
6.7.5 Comparison of the secondary metabolite content of glasshouse and

outdoor grown plants
As reported above in Section 6.7.6, a single field trial suggested that the THC:CBD
ratio of late harvested G5 CBD-chemovar plants was similar if plants were grown in
either environment. A more dramatic effect of outdoor growing was demonstrated by
five heterozygous BTBD genotypes derived from variety G159. These produced a
significantly lower mean THC:CBD ratio when grown outdoors (two-tailed t-test, p <
0.001, Table 6.9).
THC as % of THC+CBD (n = 4)
Clone G159/9 G159/11 G159/12 G159/13 G159/16 Mean
Field 33.1 ± 2.4 33.2 ± 0.2 48.9 ± 1.9 33.9 ± 0.0 33.3 ± 0.1 36.5
Glasshouse 38.8 ± 1.5 39.9 ± 0.6 57.4 ± 0.9 40.6 ± 0.6 38.6 ± 2.0 43.1
Table 6.9 The relative proportions of THC and CBD synthesised in heterozygous BTBD clones
derived from variety G159. The proportion of THC produced is shown as a % of the
THC+CBD total. Just one sample of dry inflorescence material was analysed from each plant.
In a separate study not reported here, where these clones were grown in a growth
room in identical conditions apart from contrasting growth temperature of 15ºC and
25ºC, the cooler temperature significantly increased the proportion of CBD within the
cannabinoid profile. A possible explanation is that in cooler conditions the CBD
synthase within these plants is able to compete more favourably than THC synthase for
the common precursor CBGA. No previous reports could be found describing this
temperature effect. As THC synthase and CBD synthase have been suggested to exist
165
in more than one isoform (de Meijer, 2003), other heterozygous mixed THC/CBD
chemotypes might exist which synthesise different ratios of one or both of these
synthases. These in turn may synthesise differently to temperature. It is also possible
that the increased proportion of CBD observed here is specifically induced by the
activation of an unidentified gene. In either case, the ability to produce proportionally
more CBD in cooler conditions may be an inherited trait which has at some point
improved the survival ability of Cannabis sativa.
Changing temperatures have been found to alter the ratios of terpenoids in glandular
trichomes of other species e.g. Pelargonium xhortorum. Walters and Harman (1991)
showed that growing temperature affected the ratio of C22 and C24 unsaturated
anacardic acids, both of which have similar molecular weights to the cannabinoids. It
was postulated that plants may be exhibiting a genetically controlled response to
temperature by altering the proportion of less viscous secondary metabolites within the
trichome, to maintain uniform viscosity. A potential benefit of doing so would be the
retention of the trichome’s ability to ensnare insects. Although the effects of
cannabinoid profile on trichome viscosity have not been tested, this explanation
seemed unlikely in cannabis. At room temperatures THCA is an oil and CBDA is a
crystalline solid. To retain stable viscosity as temperatures fell, it seems probable that
the trichome contents would require proportionally more of the oil. Observations in this
study demonstrated that the opposite had occurred. The cannabinoid profile may have
been affected by the higher proportion of UV light encountered outdoors. This part of
the spectrum is less able to penetrate a glasshouse and the amount produced by HPS
lamps is negligible. However, contrary to the proportional decrease in THC synthesis
observed in this outdoor crop, Lydon et al. (1987) showed that UVB radiation
significantly increased THC biosynthesis, whereas CBD biosynthesis did not show a
significant increase. However, Lydon et al. (1987) did not show the actual CBD data
and did not include a test of the statistical significance of the difference in biosynthesis
of THC and CBD when exposed to increasing UVB radiation.
To facilitate the comparison of the terpene profile of indoor and outdoor grown plants of
cultivar G5 M16, the mean peak area for each of the terpenes was calculated with each
of the five sampling dates being regarded as a single replicate. The results are shown
in Figure 6.6. To allow a clearer comparison between the contents of both major and
minor terpenes, the results are expressed as % of Total Peak Area using a logarithmic
scale. The terpene profile of these indoor and outdoor grown plants was seen to be
very similar, with no significant difference in mean % peak area in eight of the fourteen
terpenes (p > 0.05). Being based upon one indoor and one outdoor crop the relevance
166
of these differences has to be viewed with caution. The terpenes showing the greatest
significant difference in content were ocimene and nerilidol, which had proportionally
lower contents in outdoor-grown plants. This may have been a result of higher UV
light conditions outdoors. UVB radiation has been found to alter terpene profiles in other
species e.g. basil Ocimum basilicum (Johnson et al., 1999), thereby affecting the
flavour and commercial value of this culinary herb, but no obvious survival advantage
was attributed to this terpene profile change. The reason for the altered terpene profile
in outdoor grown cannabis plants may alternatively be a phytochrome-mediated
response, as observed in outdoor grown thyme, Thymus vulgaris (Tanaka et al., 1989).
Many of the terpenes found in glandular trichomes have been found to be repellant to
specific insects, and there may be an inherited benefit in only biosynthesizing certain
secondary metabolites to coincide with the life-cycle of a major predator.
100.0
10.0
% of Total Peak Area
1.0 Glasshouse
Field
0.1
Humulene
Cary Oxide†
β Pinene
β ocimene
Diepicedrene
β Myrcene
Limonene
t-Caryophyllene
Nerilidol
Selinene
Bisabolene
Aromadendrene
α farnesene
α Pinene
0.0
*** * * ** ** ***
Terpene
(ANOVA)
Figure 6.6 A comparison of the terpene profiles, as a proportion of the total peak area, of
enriched trichome preparations prepared from glasshouse and outdoor crops († = Caryophyllene
Oxide). The results are the mean of five samples produced at weekly intervals towards the end
of flowering (± SD). Glasshouse plants had been in a 12 hour day length for 6 to 10 weeks.
Field grown crops were sampled between September 17th and October 15th.
(ANOVA, Glasshouse v Field, * p < 0.05, ** p < 0.01, *** p < 0.001).
167
6.7.6 Summary of Pest and Disease Problems in the Field Trials

The main pest and disease problems incurred are summarised in decreasing order of
severity in Table 6.10.
Severity Score 1-5

(1-absent, 2-minor, 3-moderate, 4-severe, 5-fatal)
Year: - 2000 2003 2004 2005 2006

Pest
Grey Mould
1 3 3 3 1-5*
Botrytis cinerea
Caterpillars (Various spp) 2 3 3 3 2
Stemphylium Leaf Spot 1 1 1 1 3
Common Nettle Capsid †

- - - 2 2
Liocoris tripustulatus
Leaf Miner
1 1 1 1 2
Liriomyza strigata
Powdery Mildew
1 1 1 1 2
Sphaerotheca macularis
Black Bean Aphid
1 1 1 1 1‡
Aphis fabae
Table 6.10 Summary of pest problems experienced, in decreasing order of magnitude using a
subjective 1-5 score. * Botrytis initially minor but extremely severe in late harvested plots.
† Symptoms not recognized as pest damage until 2005. ‡ Aphids absent in trial plots but
moderate infestation of black bean aphid Aphis fabae observed on neighbouring cannabis seed-
crop with lower secondary metabolite content.
Grey Mould Botrytis cinerea was observed in all but the first year’s trial. The disease
regularly attacked the inflorescence tissue of mature plants (Fig 6.7) but was never
observed on the floral tissue of G5 plants before 1st October. Botrytis is reported to be
the most common disease of cannabis, proliferating in high humidity which encourages
conidia germination (McPartland at al., 2000). The disease would typically establish
deep within the inflorescence, where the humidity was especially high. In many cases
the disease was only discovered by manually parting the bracts. In the 2006 field trial
the level of B. cinerea was assessed on plants as they were harvested on five harvest
dates between 18th September and 17th October. On each occasion the number of
plants visibly infected in each plot was recorded. The results are shown in Table 6.11.
The first infected plants were observed on 2nd October. Although 12% of plants were
168
visibly spoiled, only a small proportion of each plant was affected. To comply with
Paragraph 3.1 of the GAP Guidelines the harvest should take place when the plants
are of the best possible quality according to the different utilizations. If harvested at this
stage, the crop would have to be laboriously graded to remove the poor quality
diseased material. By the 8th October a large proportion of the crop was affected and
grading the crop to only retain high quality material would have been especially difficult.
A week later on 17th October almost all of the plants were affected and the crop was
regarded totally unacceptable as a phytopharmaceutical feedstock.
Figure 6.7 Fungal damage of a cannabis inflorescence due to Botrytis cinerea.
Harvest Date 17th Sep 24th Sep 1st Oct 8th Oct 15th Oct
% of Plants Infected 0 0 12.2 30.6 98.0
Table 6.11 The level of infection with Botyrytis cinerea observed on plants harvested on five
dates between 17th September and 15th October 2006. Scores are the mean % infection rates in
the seven plots of each treatment
Other diseases experienced were Stemphylium Leaf and Stem Spot (Pleospora tarda
E.G. Simmons) and powdery mildew (Sphaerotheca macularis). Stemphyllium has
been recorded as a significant problem in hemp crops in Canada, Holland and Eastern
Europe (McPartland et al., 2000). In the field trial program the disease was only
observed in 2006. The identification was made by CABI Bioscience, Bakeham Lane,
Egham, Surrey, TW20 9TY, UK. The fungus was reported to be in its anamorphic state
169
- Stemphylium botryosum Wallr. Symptoms became widespread on foliage in August.

The discoloured leaf tissue within the spot necrosed and frequently disintegrated to
leave a characteristic ‘shot hole’ in the leaf. Towards late September these spots
would coalesce leading to a significant loss of leaf area. Throughout this latter phase
however, the bract tissue within the inflorescences was totally unaffected by the
disease.
Powdery mildew appeared in a field trial for the first time in 2006 as a minor disease.
No plants could be regarded as being of insufficient quality for harvest. Of the insect
pests to attack the crop caterpillars (various species) were present each year and
caused greatest damage, especially in late summer. However, damage was generally
restricted to the large leaves that subtended the top inflorescence. In each case the
capitate stalked trichomes on the bracts and bracteoles proved an excellent deterrent
to predation, as shown in Figure 6.9. Active caterpillars were often discovered within
the inflorescences but were observed not to be eating the resinous bracts and
bracteoles.
Figure 6.8 A cannabis plant at the late flowering stage. Resinous bracts are unaffected but
leaves below the inflorescence are heavily grazed. In some cases little more than the midrib of
the leaf remains.
Other insects pests observed included leaf miner Liriomyza (Agromyza) strigata
(Meigen), black bean aphid Aphis fabae and common nettle aphid (Liocoris
tripustulatus). In no year was more than an estimated 2% of the foliage lost to leaf
170
miner, the problem being most severe in 2006 when prolonged higher-than-average
temperatures favoured attack. Aphids were notable for their total absence in all seven
year’s crops of the G5 CBD chemovar. However, the pest was noted in 2006 on a
susceptible cannabis variety planted from seed alongside the G5 CBD trial. An
analysis of variance showed that the males were significantly more susceptible than
the females (males 11.9%, females 1.3%, p < 0.001). Cannabis appears therefore to
be typical of so many dioecious crops where the male is more susceptible to predators
– a clear example of sex-biased herbivory (Herms and Matson, 1992). The all-female
G5 crop appeared to be benefitting from both a varietal and sex-based resistance to
Aphis fabae.
Common nettle capsid is reported to feed on common nettle (Urtica spp.) and to
occasionally cause economic damage in a number of glasshouse crops (Malais and
Ravensberg, 1992). This was observed on regular occasions in outdoor crops during
August when larvae from the summer generation have reached the adult stage.
Commercial fibre and seed-hemp crops would typically be approaching their natural
harvest date in August and be beyond a stage at which serious damage could occur.
However, the all-female cannabis drug-crop would be producing ample succulent
foliage at this time and clearly more susceptible to attack. Also encountered but not
quantified were rabbits and arable weeds. After excessive grazing damage in
2001/2002 all subsequent trials were protected with rabbit proof fence. Weeds were
manually removed at intervals through the season.
6.7.7 Effect of Raised Temperatures on Crop Drying Time

The initial attempt to dry cannabis plants on the floor of an oast house drying chamber
was only partially successful. To maintain a uniform rate of drying through a hop crop,
the material needs to be spread evenly on the floor in sufficient depth to substantially
restrict air flow. Such a crop is typically dried within a few hours. The depth of the
cannabis (approximately 15 cm) was much less than the one metre depth of hops
traditionally processed in the oast, and the inlet temperature of 40ºC was at the lower
end of the typical range (Neve, 1991). To successfully dry the crop this way required
twenty four hours.
The crops in 2004-2006 were hung to dry on wires suspended within the oast house.
Cannabis crops typically feel sufficiently dry to be stripped when below 15% moisture.
Figure 6.10 suggests that in 30, 40 and 50ºC, the mean times taken to achieve 15%
moisture were approximately 36, 18 and 11 hours respectively. Plants of the same
cultivar, dried in the glasshouse-crop drying facility at 25ºC typically took 4.5 – 5.0 days
to reach this same moisture level. However, ventilation levels differed between the two
171
locations and the longer drying time cannot be entirely attributed to a lower drying
temperature. This prolonged drying time was highly favourable to the proliferation of
fungi and bacteria, and plants already showing low levels of disease were seen to
deteriorate in these conditions. Increased drying temperatures greatly reduced the
time within which such fungal spoilage could occur. It was accepted that proportionally
more of the volatile monoterpenes would be lost compared to the sesquiterpenes at the
higher drying temperatures. However, it was considered likely that the increased
monoterpene losses at 40ºC and 50ºC were minor compared to those that would have
been lost during the decarboxylation process, when milled BRM was heated at much
higher temperatures to convert the cannabinoid acids into the neutral forms.
80
70
60
% Moisture
50 30 C
40 40 C
30 50 C
20
10
0
0 3 6 9 12 15 18 21 24 27 30
Drying Time (Hours)
Figure 6.9 The rate of moisture loss of field grown cannabis, when dried at three temperatures
(30, 40 and 50ºC). The results are the mean of three crops dried in 2004-2006 (± SD) and show
the pattern of moisture loss until mean moisture content was <15%.
6.8 CONCLUSIONS
The research reported here indicated for the first time that the outdoor propagation of
Cannabis as a phytopharmaceutical in the UK was possible. However, the largest
problems to overcome were disease control and the logistical difficulties encountered in
harvesting a large volume of crop in a limited period of time. Cannabinoid and terpene
profiles of indoor and outdoor gown crops were shown to be very similar if both crops
were harvested at the end of floral development, as denoted by complete stigma
172
senescence. However, the desired cannabinoid profile was only achieved in mid-
October, when levels of Botrytis infection were high. The reliable outdoor production of
this crop appears impossible therefore without the use of fungicides. This would
necessitate harvested crops being tested for fungicide residues.
The cannabinoid profile of the outdoor crop appeared less stable than that of crops
grown in the glasshouse. As a result, the harvest would need to take place within a
very short period when cannabinoid levels were assessed as being within the agreed
specification. This could coincide with unfavourable weather conditions, which would
hamper harvest operations. However, the ability to rapidly dry the crop was
demonstrated. An alternative way of exploiting outdoor grown Cannabis may be to
collect the glandular trichomes from this material. The crop could conceivably be
harvested and promptly fresh frozen for processing at a later date. Similar harvest
operations have been used in the frozen pea industry for decades. As stated in the
previous chapter, the bulk removal of Humulus trichomes is also practised industrially.
The concept may warrant evaluation in forthcoming seasons.
173
Chapter 7 General Discussion

GW Pharmaceuticals plc, the sponsoring company, has been resolute in developing
cannabinoid-based phytopharmaceuticals from Cannabis sativa L. At the
commencement of this thesis in autumn 2003 the company was propagating two
chemotypes to produce the medicine Sativex®. This was undergoing Stage III clinical
trials for the treatment of symptoms of multiple sclerosis (MS) and for various forms of
pain relief. In Canada in 2005 Sativex® received provisional approval with conditions for
the treatment of central neuropathic pain in MS, and in 2007 for intractable cancer pain.
In 2008 GW Pharmaceuticals commenced Stage III clinical trials in the US to evaluate
the efficacy of this medicine for the control of pain in terminal cancer patients. Since
2005 Sativex® has been available in the UK on a ‘named-patient’ basis, with the Home
Office being notified of each recipient. Many other countries have since allowed
prescription of the medicine on the same basis. The medicine contains two major
cannabinoids, Δ9-tetrahydrocannabinol (THC) and cannabidiol (CBD), which both had
reported pharmacological properties. The choice of these two cannabinoids was based
upon evidence that, in addition to them individually being pharmacologically active, the
two acted together synergistically, with CBD abrogating the side effects of anxiety and
intoxication associated with THC alone (Fowler and Law 2006). Sativex® was a
botanical drug, i.e. a well characterised, multi-component standardised drug extracted
from plant sources. In addition to THC and CBD, the medicine also contained lesser
amounts of other cannabinoids, as well as monoterpenes and sesquiterpenes. In vivo
rodent studies showed cannabis extracts to have significantly increased
pharmacological activity over cannabinoids alone, (Williamson, 2001) and the
monoterpenes and sesquiterpenes were suspected of being at least partly responsible
for this increase. It was envisaged that these terpenes might also prove to be
advantageous ingredients in Sativex®.
To meet the quality and safety standards demanded of a medicine by the Regulatory
Authorities, the Botanical Raw Material used as the starter material for Sativex®, and
possible future botanical medicines, had to contain a minimum and maximum level of
each of several secondary metabolites. To increase the probability of each batch of
botanical raw material meeting this specification, the growing conditions initially
adopted were kept as uniform as possible and plants routinely harvested when of a
fixed age. Initial attempts to grow year-round uniform material were thwarted by a
seasonal fluctuation in yields. The cause of this was suspected to be seasonal variation
in irradiance levels within the glasshouse. Relatively little was known of how alterations
174
to the growing protocol would affect the secondary metabolite profile of the harvested
plants. It was possible therefore that any attempts to improve the growing conditions
might produce feedstock that failed the product specification. A series of tests was
performed to analyse the effect on plant development and secondary metabolite
content of altering growing conditions. As part of this, seed-sown plants were tested
alongside those derived from cuttings (clones). Plants were compared when grown in
varying daylengths and under differing irradiance conditions. A range of harvest timings
was also compared.
Most if not all other phytopharmaceuticals are produced from outdoor grown crops.
Growing cannabis indoors provided better growing conditions and higher level of
security. However, the process was costly and produced a large carbon footprint. The
energy consumed for lighting alone exceeded 0.5 kW hr per gram of dry feedstock for
glasshouse crops, and reached approximately 1 kW hr per gram in a totally enclosed
environment. Applying equally high levels of security to the growing of CBD chemovar
crops was arguably excessive. It was recognised that outdoor growing of cannabis
would reduce energy costs, but the effects on secondary metabolite profile and plant
quality might be unacceptable. The vagaries of the weather would be expected to
influence yields but how this would affect secondary metabolite profile was not known.
Pest and disease problems would likely be different to those encountered by a
glasshouse crop and this might unacceptably affect quality. Drying the crop would also
present logistical difficulties. Prior to commencing the work of this thesis, one CBD crop
had been grown successfully outdoors and the cannabinoid yields achieved were
similar to those achieved with a single glasshouse crop. For this thesis a sequence of
CBD-chemotype crops was grown outdoors and plant development studied. Pest and
disease levels were monitored. Crops were harvested over a range of dates from mid-
September to mid-October and the effect of harvest date on cannabinoid and terpene
profile compared to that of glasshouse crops. In an attempt to reduce the degree of
crop spoilage during the drying process, trials were performed over three seasons to
compare crop drying regimes at a range of temperatures.
When characterising any phytopharmaceutical feedstock, a systematic and illustrated

report of the microscopical details is generally required. In Cannabis sativa L, the
trichomes were perceived to be the most important parts of the plant, as the
cannabinoids and terpenes are biosynthesised and sequestered within these
structures. A review of the form, function and distribution of these constitutes Chapter
Three of this thesis. As the chapter shows, there are differing trichome forms. The
cannabinoid and terpene profile of the plant changes through its life and this can often
175
be at least partly explained by the presence of changing proportions of so-called

sessile and capitate stalked trichomes. Most notable examples are the higher
proportions of CBC and sesquiterpenes within the secondary metabolite profile of
foliage compared to floral material, this being attributable to the absence of capitate
stalked trichomes on foliage. In potent female floral material the total volume of
secondary metabolites in sessile trichomes is almost negligible compared to that in the
larger capitate stalked trichomes. Consequently, although they have a different profile,
the sessile trichomes have little influence on the overall secondary metabolite profile of
the plant. So-called ‘manicuring’ of inflorescence, to remove bract tissue lacking
capitate stalked trichomes, is time consuming. Although favoured by producers of
recreational cannabis, seeking to improve potency and flavour, the process has little
effect on the profile of the remaining feedstock.
The removal of glandular trichomes from cannabis floral material produces an ‘enriched
trichome preparation’ that has a very similar cannabinoid and terpene profile to that of
the plant material from which it came, but is much more potent. One major exception is
the diterpene phytol, which is associated with the biosynthesis and catabolism of
chlorophyll and Vitamin E, and this is predominantly found outside the trichome. Due to
the naturally small volume of the sessile trichome population, the collection of these
from aerial tissues produces very small quantities of material. However, the secondary
metabolite profile of the collected material can be of great potential interest. The
removal and sieving of sessile trichomes from aerial parts of a CBC-rich chemovar
containing 1.4% w/w CBC produced a preparation that contained 44% w/w CBC. The
stage of development of these sessile trichomes can greatly affect their cannabinoid
profile. In the example just quoted CBC accounted for 61% of the total cannabinoids
present in the raw material. However, the agitating and sieving process predominantly
captured the older more mature sessile trichomes and the CBC purity level of the
resultant filtrate was 94% w/w. This method of producing an enriched source of CBC,
by selectively capturing mature sessile trichomes, was submitted for patent protection.
Enriched trichome preparations, predominantly containing capitate stalked trichome
resin heads, were produced which contained up to 67% w/w THC. This material has
the clear advantage of being less bulky than feedstocks containing botanical raw
material. It also lacks the chlorophyll-based pigments and other substances found in
foliage, which may be regarded as undesirable in some botanical medicines. When
considering possible ways of mechanising the removal and collection of glandular
trichomes, it was recognised that harvesting of cannabis glandular trichomes is also the
basis of illicit cannabis resin manufacture. The efficiency of this was assessed in
176
studies to characterise the illicit cannabis materials typically used in England in

2005/2005.
Resin is just one form of illicit cannabis circulating in the UK. Also commonly found are
imported outdoor-grown ‘herbal cannabis’ and a more potent intensively indoor-grown
all-female cannabis product called sinsemilla - or more colloquially ‘skunk’. Illicit
cannabis is used for recreational and medicinal purposes. Indeed, it was partly on the
basis of increasing evidence of the efficacy of this material that GW Pharmaceuticals
was given Home Office permission to evaluate cannabis as a phytopharmaceutical.
However, little was known of the cannabinoid profile of this material and the impact that
this would have on its efficacy and safety. For this thesis, the first survey of UK
cannabis cannabinoid content was performed.
The study showed that herbal, sinsemilla and resin contained very variable quantities
of THC and herbal and sinsemilla cannabis were almost devoid of CBD. Resin
contained extremely variable quantities of CBD, and along with herbal cannabis often
contained high levels of CBN. The ratio of these cannabinoids in resin varied greatly,
much of this due to the fact that THC catabolised much faster than CBD. In theory, a
variable content of one cannabinoid could be overcome by patients if they self-titrated
to achieve the desired dose level. When using resin, the only significant source of
CBD, self-titration did not offer the ability to correctly dose both THC and CBD.
Although the only significant source of CBD, previously unseen survey data analysed
for this thesis showed that resin was unfavoured amongst the majority of medicinal
cannabis users. Whereas the appearance of resin gave no indication of its likely
cannabinoid content, simple visual organoleptic assessments did give a significant clue
as to the THC content of sinsemilla. Part of the preference for sinsemilla amongst
some medicinal users is the ability (albeit illegal) to grow the material at home.
Although seed is commercially available, it is almost always of the high-THC genotype.
The study showed that the resin’s share of the illicit cannabis market was in decline,
with sinsemilla becoming more dominant. The THC content of sinsemilla in the UK was
showing a significant upward trend. The psychoactive potential of the average
cannabis sample was thus increasing. This was due to the increasing THC content,
combined with a decline in the presence of the antipsychotic cannabinoid CBD. This
data was used by the Advisory Council on the Misuse of Drugs before making
recommendations to the UK Government on the possible reclassification of cannabis
(ACMD, 2008).
Chemical characterisation involved the analysis of whole cannabis plant tissues and
glandular trichomes collected from them. Unpollinated female floral tissues were known
177
to be the greatest source of cannabinoids and terpenes and studies were focused here.
Most cannabis varieties showed a strong photoperiodic response and, after an initial
few weeks of vegetative growth in continuous lighting, plants generally commenced
flowering within ten days of being switched to a daylength of 12 hours. Inflorescences
produced increasing numbers of florets over subsequent weeks but after eight weeks in
short days, growth was slowing rapidly and few additional fertile stigmas were formed.
The cessation in floral development was mirrored by a slowing in cannabinoid
biosynthesis. Over this period the cannabinoid profile changed, with the proportion of
CBG decreasing significantly. Outdoor grown CBD chemovars showed a very similar
pattern of floral development and cessation of flowering was matched by the same
decrease in the proportion of CBG in the cannabinoid profile. The terpene profile of the
plant also changed during plant development and, whether grown in stable glasshouse
conditions or outside, the pattern of change was similar with the profile stabilising as
the formation of new florets ceased.
The cannabinoid profile of other chemovars also showed a marked change during the
flowering process. This was most pronounced in heterozygous chemovars producing
more evenly mixed cannabinoid profiles. One of these (variety G159) was the source
of several clones that synthesised THC and CBD in approximately equal quantities,
and the ratio was observed to change significantly during the flowering process. These
clones all produced a significantly higher proportion of CBD when grown outdoors in
cooler average temperatures. Growth room tests confirmed that cooler propagation
temperatures significantly increased the proportion of CBD within the cannabinoid
profile. This could be simply due to the CBD synthase enzyme having a proportionally
greater efficiency than THC synthase at lower temperatures, or to the plant switching
on some mechanism to enhance CBD synthesis at the expense of THC. This finding
has a wider implication for the understanding of how THC and CBD chemovars
perhaps evolved and were exploited by man. THC-dominant chemovars are typically
associated with latitudes of 30º or less (Small, 1976) where the plant is most capable of
producing potent material and the local culture enables its production. CBD-dominant
chemovars are associated with temperate latitudes, and this is heavily influenced by
licensing restrictions that only allow the exploitation of this chemotype for fibre or seed
production. This observation with heterozygous cannabis suggests that in a landrace
population, which would include homozygous THC and CBD chemovars and
heterozygous mixed THC+CBD chemovars (de Meijer, 2003), a higher proportion of
THC would be synthesised in the warmer climates irrespective of the involvement of
man.
178
Although the glasshouse used for this research had sophisticated environmental
management and supplementary lighting systems, as stated earlier there was a large
seasonal fluctuation in light conditions within the building. This was the presumed
explanation for large seasonal variations in crop yield. Although the existing
supplementary lighting system was capable of adding 16 W m-2 of photosynthetically
active radiation (PAR) to the natural light, and originally considered adequate to
support the healthy growth of many plant species, this proved inadequate for uniform
year-round growth of Cannabis. Winter yields of botanical raw material were
significantly reduced and there was an even greater reduction in the yield of
cannabinoids. Following growth room tests, this was overcome by the provision of an
improved supplementary lighting system capable of adding photosynthetically active
radiation levels of 53 W m-2 to that provided naturally by daylight. This level of lighting is
more than twice that typically used in the production of UK glasshouse-grown food-
crops. It was discovered that high irradiance levels appear to be most essential at the
beginning of the flowering process and greater efficiency and reduced costs could
perhaps be achieved by apportioning more of the light energy to crops at this growth
stage. It was also shown that by using lamps to provide irradiance levels of 75 W m-2,
year-round crops of cannabis could be grown in a totally enclosed environment with no
natural lighting. These showed no significant difference in yield compared to those
grown in the glasshouse under supplementary lighting.
It was found that the yield of botanical raw material produced per unit area was linearly
proportional to the average irradiance level in that growing environment. Raising the
irradiance level also caused the plant to divert a higher proportion of energy to the
biosynthesis of cannabinoids. The amount of cannabinoid produced was therefore
strongly related to the consumption of electrical lighting energy. In a drive to produce
flowering cannabis crops with a uniform secondary metabolite content throughout the
year, irradiance levels were kept as uniform as practicable throughout the twelve hour
day. In mid-summer, this could mean that supplementary lighting would be used to
achieve target irradiance levels during dull periods of the day, but on clear days
glasshouse roof shades would be closed to reduce the ingress of free sunlight around
noon. This was not necessarily the most energy efficient policy. It is possible that the
allocation of carbohydrate to primary or secondary metabolite biosynthesis is less
dependent on the irradiance level (i.e the light energy per unit area per unit time) and
more closely related to the total quantity of light energy falling on the crop during any
one twelve hour day - or possible an even longer period. As observed in the study here
with variegated cannabis, cannabinoid biosynthesis continues in tissue that is not
179
photosynthesising. Indeed Crombie (1977) reported that cannabinoid biosynthesis

continues during the night, the carbohydrate used therefore having been formed many
hours earlier. Had uniformity of crop been less important, the glasshouse
supplementary lighting system could have been switched on continuously on summer
days and this would have boosted yields of raw material and cannabinoids even
further, but possibly altered the secondary metabolite profile. The most likely changes
within the secondary metabolite profile would have been alterations to the ratio of
monoterpenes and sesquiterpenes, as a consequence of a predicted altering in the
ratio of foliar and floral material.
Field trials over several seasons showed that the Sativex-dependent CBD crop could
also be propagated outdoors in the UK. The costs were greatly reduced but a range of
problems were encountered, of which fungal infection with Botrytis cinerea caused the
greatest crop losses. Initially low levels of infection would frequently ruin crops within
the first 48 hours of the drying process, if plants were dried at 30ºC or below.
Increasing crop drying temperatures up to 50ºC considerably reduced crop drying
times and offered some reduction in the level of fungal spoilage. Over a five year
period, the crop regularly commenced floral growth in the last ten days of August and it
was judged ready for harvest in the second week of October, when environmental
conditions favoured fungal attack. Future outdoor growth appears unlikely without the
use of fungicides to control disease. Sativex-dependant THC varieties commenced
flowering too late in September to produce a satisfactory crop, but it was shown that
other high yielding THC genotypes could produce healthy high-yielding crops outdoors
in Southern England.
A major first step in the optimisation of cannabis as a phytopharmaceutical was to

check that botanical raw material produced from clones was more uniform than that in
plants grown directly from seeds, even in a highly inbred variety exhibiting pronounced
phenotypic uniformity. It is common practise amongst licensed and illicit growers to
grow plants from seed and to make clones from the highest yielding or most potent
progeny. The plants grown to maturity from these clones typically maintain the
desirable trait. Research here showed that cloned plants exhibited a significantly more
uniform cannabinoid profile as well as a higher cannabinoid yield.
Grown indoors or out, the cannabinoid and terpene profiles were shown to change
throughout the female floral development stages, with both tending to stabilise once the
formation of new florets within the inflorescence ceased. The near or complete
absence of newly formed stigmas was a visible conformation of the readiness of plants
for harvest. Grown in tightly controlled indoor conditions, this growth stage could be
180
routinely achieved after a fixed time and harvest times could be planned in advance. In
outdoor growing conditions plant development was under strict photoperiodic control.
Over five consecutive seasons crops of the CBD chemovar G5 were seen to
commence flowering within a nine day period between the 22nd August and 1st
September, when daylength was approximately 14 hours. In all trials the crop finished
flowering in the second week of October, just two weeks after the autumnal equinox,
and this date could be predicted as optimum for future harvests of this variety at this
latitude. Studies of the cannabinoid profile and terpene profile in both environments
showed that by the end of the flowering period the cannabinoid and terpene profiles
were most stable and the proportion of CBG within the cannabinoid profile was at a
minimum. This was the ideal time to harvest the plant for economic and quality
reasons. The proportion of propyl (THCV, CBDV etc.) and pentyl side-chained
cannabinoids (THC, CBD etc.) were seen to change through the flowering period and
hence the ratio of these would be affected by growth stage at harvest.
As recommended in many growing guides, illicit cannabis is typically induced to flower

by exposing it to a steady twelve hour daylength. In an outdoor setting this is the
daylength naturally occurring at the autumn equinox when outdoor grown crops were
seen to finish flowering. The CBD chemovar, grown outdoors, was seen to commence
flowering when daylength fell to approximately fourteen hours per day, and this would
be defined as its ‘critical daylength’. At this point a phytochrome-controlled hormone
release mechanism induces flowering. The use of a twelve hour day in the glasshouse
is thus shorter than that required to induce flowering in this chemovar and denies a
flowering crop access to an additional two hours light exposure and associated
photosynthetic activity. This in turn might reasonably be expected to affect raw material
and cannabinoid yields. However, glasshouse studies showed this prediction to be
incorrect. The studies in question compared the development of plants grown in
eleven, twelve or thirteen hour daylengths. In all regimes the plants were observed to
commence flowering equally quickly. However, those maintained in thirteen hour days
were slower to cease vegetative growth. This resulted in significant changes in the
cannabinoid profile, a significant increase in height and no increase in the feedstock
yield, despite a proportional increase in electrical lighting energy costs. This finding
suggests that there is more than one critical daylength in cannabis, one of which
induces flowering and a shorter daylength at which vegetative growth is hormonally
inhibited. Further reducing the daylength below twelve hours resulted in significant
decreases in raw material and cannabinoid yield. Consequently, it is recommended
181
that a twelve hour daylength should continue to be used as the standard daylength in
which to maintain the cannabis crops through the flowering stage.
Looking ahead, the survey of the cannabinoid content of illicit cannabis in the UK in
2005 was clearly ‘a snapshot in time’. The results showed that the potency and market
share of cannabis resin, herbal cannabis and sinsemilla were changing and this
warranted a future study to monitor these changes. Indeed, such a survey was
performed by the Home Office Scientific and Development Branch in 2008, and
acknowledged assistance was given (Hardwick and King 2008). A further Home Office
study is planned to monitor on-going changes in the potency of street cannabis in late
2009.
Within the GW Pharmaceuticals glasshouse a number of additional studies are

foreseen. More research is recommended to investigate ways of further improving the
winter and summer yield of glasshouse grown crops by alterations to light energy
levels. In addition to placing greater emphasis on total the light energy received by the
plant, rather than the irradiance conditions prevailing, future studies should be
performed to research the correlation between the total quantity of light energy
received and the relative allocation of carbohydrate to primary or secondary metabolite
biosynthesis. The observation that high irradiance levels appear to have greatest
impact on cannabinoid yields when applied at the beginning of flowering requires
further investigation. If confirmed, supplementary lighting may be selectively
concentrated on plants at this growth stage. The analytical studies, in which terpene
profiles were monitored during plant development, should be repeated in view of the
small number of crops assessed to date.
To enable CBD chemovar crops to be cultivated outdoors, the application of fungicides

seems increasingly likely. A range of fungicides should be tested and residues within
the crop assessed. Alternatively, a more southerly growing location should be
evaluated. Plant breeding of earlier flowering varieties may overcome disease
problems, but it may be impossible to breed such varieties with sufficiently similar
cannabinoid and terpene profiles to be regarded as being of the same chemotype.
Rather than relying on a mixed feedstock of dried foliar and floral material, the
collection and storage of enriched trichome preparations offers distinct advantages.
Further evaluation of trichome collection methods is highly recommended. The
technique is already performed on an industrial scale with the closest relative genus -
Humulus (the hop) and further work is recommended to see if the available machinery
could be adapted for use on cannabis.
182
As the site of the biosynthesis of cannabinoids, monoterpenes and sesquiterpenes, the

glandular trichomes are arguably the most important part of the plant. A greater
understanding of the activities within these trichomes would possibly enable the grower
to identify when these structures were operating at their optimum. The grower would
then be able to assess how difference in growing conditions altered the level of
activities within the trichome. As part of a general study of these structures, the
capitate stalked form was seen to be the most important type, producing the bulk of the
cannabinoids on female floral material. Using relatively simple light microscopes, the
study showed that the terpenoids are sequestered in a resin head. Within a mature
resin head a disk of secretory cells is visible, occupying about 20% of the volume. Vital
stains were perceived as a tool to locate and highlight those tissues where respiratory
activity was taking place within the trichome, and to indicate when this was most active.
One such stain ‘tetrazolium red’ suggested that almost all the metabolic activity took
place within the secretory cells, even though the site of cannabinoid biosynthesis is
purported to be within the storage space above the cells (Sirikantaramas et al., 2005,
Taura et al., 2007). The secretory cells would indeed consume much energy in the
biosynthesis of cannabinoid precursors and cannabinoid synthase enzyme. However,
the lack of apparent activity within the storage space may be purely an anomaly, due to
the inability of these water based stains to penetrate this area. The use of lypophylic
alternatives may improve the assessment of metabolic activity within the secretory
head.
Enriched trichome preparations, almost entirely containing capitate stalked trichome

resin heads, were found to have a wide range of cannabinoid contents from
approximately 30% up to more than 67% % w/w (dry). This may be due to varying
quantities of membrane tissue, enzymes and terpenes. If enriched trichome
preparations were to be used as phytopharmaceutical feedstock, rather than floral and
foliar material, it would be important to gain further understanding of the cause of this
variability.
In September 2008, GW Pharmaceuticals plc reported positive results from a placebo-

controlled randomized withdrawal study of Sativex in patients with neuropathic pain
due to MS. In February 2009 the company reported positive results in a similar trial with
patients experiencing spasticity due to MS. These studies showed that the efficacy of
Sativex in the treatment of both neuropathic pain and spasticity due to MS is
maintained in long-term use. A few days before completion of thesis GW
Pharmaceuticals plc announced positive results from a pivotal Phase III double-blind
randomised placebo-controlled study of Sativex® in patients with spasticity due to MS,
183
who have achieved inadequate spasticity relief with existing therapies. This Phase III
study used an enriched design whereby 573 patients initially received Sativex for four
weeks in a single blind manner (Phase A), following which Sativex responders (n =
241) were randomized to continue on Sativex or switch to placebo for a further twelve
weeks in a double-blinded manner (Phase B). The prospectively defined primary
efficacy endpoint of the study - the difference between the mean change in spasticity
severity of Sativex vs Placebo in Phase B - was highly statistically significantly in favour
of Sativex (p = 0.0002). The difference between Sativex and placebo was also
significant for a number of secondary endpoints. 74% of Sativex patients achieved an
improvement of greater than 30% in their spasticity score over the entire study versus
51% on placebo (p = 0.0003). These three studies was performed following regulatory
guidance from the UK regulatory authority (MHRA) and provided evidence of long term
efficacy to be included as part of a forthcoming European regulatory submission
planned for mid 2009. If successful this would require a large increase in the quantity
of crop grown. The research performed for this thesis aids the reliable propagation of
that crop. In the UK alone, MS is the most common disabling neurological disease of
young adults, affecting approximately 85000 people. Approximately 2% of the entire
UK population experiences neuropathic pain. Regulatory approval of Sativex® would
lead to the medicine being widely available on prescription. For many of these
patients, illicit cannabis would have been the only form of medicinal cannabis available
hitherto.
In early 2009, the company was performing Phase I trials to evaluate the cannabinoid
THCV for appetite control in treating obesity. Other cannabinoids were also undergoing
a range of in-vitro and in-vivo studies. A range of different Cannabis chemovars had
been specifically bred to be dominant in cannabinoids other than THC and CBD. The
parent plants used to breed these chemovars differed in provenance, and the optimum
growing conditions for these had not been determined as fully as those used to
produce THC and CBD. More horticultural research would be required to optimise their
propagation.
This thesis has emphasised many times the fundamental importance of the glandular
trichome on Cannabis sativa L. It is here, within the vesicles of the glandular trichome
that the enzymes and precursors of cannabinoid biosynthesis are secreted and the
phytocannabinoid story starts. Further research will aim to improve the understanding
of the biosynthetic activities within these trichomes and take this story forward.
184
References
REFERENCES
Acamovic, T. and Brooker, J.D., 2005. Biochemistry of plant secondary metabolites
and their effects in animals. Symposium on 'Plants as animal foods: a case of catch
22?’, Proceedings of the Nutrition Society, 64, 403-412.
ACMD (Advisory Council on the Misuse of Drugs). 2008. Cannabis: Classification and
Public Health. London, Home Office. http://drugs.homeoffice.gov.uk/ [accessed 01 Jan
2009].
Akers, C.P., Weybrew, J.A. & Long, R.C.1978. Ultrastructure of glandular trichomes of
leaves of Nicotiana tabacum L., cv Xanthi. Am. J. Bot., 65, 282-292.
Alberts, B., Johnson, A., Lewis, J., Raff, M., Roberts, K. and Walter, P. Molecular
Biology of The Cell (4th Edn.). 2002. Garland Science, New York. p. 854.
Albuisson, M., Lefevre, M., Wald, L. 2006. Averaged Solar Radiation 1990-2004. Paris:
Ecole des Mines de Paris/Armines. http://www.soda-is.com/eng/map/index.html#monde
[accessed on 21 Mar 2008].
Anathakrishnan, T.N., 1993. Bionomics of thrips. Annu. Rev. Entomol., 38, 71-92.
Andre, C. and Vercruysse, A., 1976. Histochemical study of the stalked glandular hairs
of the female Cannabis plants, using fast blue salt. Planta. Medica., 29, 361-366.
Baier, C. and Bomme, U., 1996. Veruneinigung von Arzneidrogen: aktuelle Situation
und Zukunftsperspektiven. Zeitschrift fur Arznei- und Gerwurzpflanzen: 1 (4), 40-48.
185
References
Baker, R., 2002. Commerce in crude drugs. In: Evans, W.C. (Ed.), Trease and Evans
Pharmacognosy, 15th Ed, Saunders, London, pp. 57-66.
Bang, M-H., Choi, S.Y., Jang, T.O., Kim, S.K., Kwon, O-S., Kang, T-C., Won, M.H.,
Park, J., Baek, N-I., 2002. Phytol, SSADH Inhibitory Diterpenoid of Lactuca sativa. Arch.
Pharm. Res, 25 (5), 643-646.
Barber, A., Corkery, J., Ogunjuyigbe, K., 1996. Statistics of drugs seizure and offenders
dealt with, United Kingdom, 1995. Home Office Statistical Bulletin: 25-96; Government
Statistical Service, London.
Barnes, M. P., 2006. Sativex®: clinical efficacy and tolerability in the treatment of
symptoms of multiple sclerosis and neuropathic pain. Expert. Opinion. Pharmacother.,
7 (5), 607-615.
Beutler, J.A. and Der Manderosian, A.H., 1978. Chemotaxonomy of Cannabis, 1.
crossbreeding between Cannabis sativa and Cannabis ruderalis, with analysis of
cannabinoid content. Econ. Bot., 32, 387-394.
Bócsa, I. and Karus, M., 1998, The Cultivation of Hemp. Botany, Varieties, Cultivation
and Harvesting. California: Hemptech. p. 44.
Bonnie, R and Whitbread, C. 1970. The forbidden fruit and the tree of knowledge: An
enquiry into the legal history of American marijuana prohibition. Virginia Law Review.
Vol. 56, No. 6.
186
References
Boucher F., L. Cosson, J. Unger.,Paris, M.R., 1974. Le Cannabis sativa L.; races
chemiques ou varietes. Pl. Med. Phytotherap. 8, 20-31.
Briosi, G. and Tognini, F. 1894. Intorna alla anatomia deila canapa Cannabis sativa L.
Parte prima: Organi sessuali. Atti 1st Bot. Pavia, 2 (3), pp. 91-209.
BMA (British Medical Association). 1997. Therapeutic uses of cannabis: Harwood
Academic Press, Amsterdam, pp. 7-20.
Brown, D.T. 1998. The Therapeutic potential for cannabis and its derivatives. Cannabis,
The Genus Cannabis. Harwood Academic Publishers. Amsterdam, pp. 115-124.
Bryant, J.P., Chapin III, F.S., Klein, D.R.,1983. Carbon/nutrient balance of boreal plants
in relation to vertebrate herbivory. OIKOS, 40, 357-368.
Byers, J.A., 2005. The cost of alarm pheromone production in cotton aphids, Aphis
gossypii. Naturwissenschaften. 92, 69-72.
Caldwell, M.M., Robberecht, R., Flint, S.D., 1983. Internal filters: prospects for
UV-acclimation in higher plants. Physiol. Plant, 58, 445-450.
Callaway, J.J. 2002. Hemp as food at high latitudes. Journal of the International Hemp
Association, 7 (1), p. 105.
187
References
Cerniak, L., 1985. The Great Books of Hashish, Vol 1: Book 1. Kulu Trading, Bussum,
Netherlands, pp. 81-96.
Cichewicz, D.L. 2004. Synergistic interactions between Cannabinoid and opiod
analgesics. Life. Sci., 74, 1317-1324.
Clifford, P. 2004. Teaching the pressure flow hypothesis of phloem transport. J. Biol. Ed,
39 (1) 35-39.
Clarke, C. 2006. Regulation of Cannabis: House of Commons Debate 19th January
2006. Hansard: Column 982.
Clarke, R.C. 1981. Marijuana Botany. Ronin Publishing. San Francisco.
Clarke, R.C. 2001. Sinsemilla heritage – What’s in a name? In: King J, (Ed.),The
Cannabible.: Ten Speed Press, Berkely, California, pp. 1-24.
Clarke, R.C. and Watson, D.P., 2007. Cannabis and Natural Cannabis Medicines. In:
ElSohly, M.A. (Ed), Marijuana and the Cannabinoids.: Humana Press. Totowa, New
Jersey. pp. 1-16.
Comelli, F., Giagnoni, G., Bettoni, I., Colleoni, M., Costa, B., 2008. Antihyperalgesic
effect of a Cannabis sativa extract in a rat model of neuropathic pain: mechanisms
involved, Phytother. Res. doi: 10.1002/ptr.2401
188
References
Conrad, C., 2004. Cannabis Yields and Dosage – A Guide to the Production and Use of
Medical Marijuana., www.safeaccessnow.net [accessed 27 Feb 2008].
Consroe, P., Musty, R., Rein, T., Tillery, W., Pertwee, R., 1997. The perceived effects of
smoked cannabis in patients with multiple sclerosis. Eur Neurol, 38, 44-48.
Costa, B., Trovato, A.E., Comelli, F., Giagnoni, G., Colleoni, M., 2007. The
non-psychoactive cannabis constituent cannabidiol is an orally effective therapeutic
agent in rat chronic inflammatory and neuropathic pain. Eur. J. Pharmacol., 556, 75-83.
Crombie. W.M.L. 1977. The influence of photosynthesis and SKF inhibitors on
cannabinoid production in Cannabis sativa. Phytochemistry. 16, 1369-1371.
Croteau, R. and Martinkus, C., 1979. Metabolism of monoterpenes: demonstration of
(+)-neomentyl-β-D-glycoside as a major metabolite of (-) –menthone in peppermint
(Mentha piperita). Plant. Physiol., 64, 169-175.
Croteau, R. and Johnson, M.A. 1984. Biosynthesis of Terpenoids in Glandular
Trichomes. In: Rodriguez, E.P., Healey, L., Mehta. I., (Eds.), Biology and Chemistry of
Plant Trichomes. Plenum Press: London. pp. 133-185.
Culpeper N (1653).The Complete Herbal, Petert Cole, London. pp. 128-129.
189
References
Dayanandan, P. & Kaufman, P.B. 1976. Trichomes in Cannabis sativa L.
(Cannabaceae). Am. J. Bot ., 63 (5), 578-591.
Dell, B. and McComb, A.J., 1978. Plant Resins - Their Formation, Secretion and
Possible Functions. Advances in Botanical Research., Academic Press, San Diego,
pp. 278-317.
Delly, J.G. 1988. Photography through the microscope.: Kodak Publications, Rochester,
NY. pp. 3-47.
Devane, W.A., Hanus, L., Breuer, A., Pertwee, R.G., Stevenson, L.A., Griffin, G.,
Gibson, D., Mandelbaum, A., Etinger, A., Mechoulam, R. 1992. Isolation and structure
of a constituent that binds to the Cannabinoid receptor. Science. 258 (5090),
1946-1949.
Dickison, W.C. 1974. Trichomes. In: Radford, A.E., Dickison, W.C., Massey, J.R,. Bell,
C.R. (Eds), Vascular Plant Systematics.. Harper and Row, New York, pp. 198-202.
Dietmar-Behnke, H. 1984. Plant trichomes – structure and ultrastructure: General
terminology. Taxonomic applications and aspects of trichome-bacteria interactions in
leaf tips of Dioscorea. Biology and Chemistry of Plant Trichomes., Plenum Press,
London. pp. 1-21.
190
References
ElSohly, M.A., Holley, J.H., Lewis, G.S., Russell, M.H., Turner, C.E., 1984. Constituents
of Cannabis sativa L. XXIV: The potency of confiscated marijuana, hashish, and hash
oil over a ten-year period, J. Forensic. Sci, 29 (2), 500-514.
El Sohly, M. A., Ross, S.A., Mehmedic, Z., Arafat, R., Yi, B.,Banahan III, B.F. 2000.
Potency trends of delta9 - THC and other cannabinoids in confiscated marijuana from
1980-1997. J. Forensic. Sci., 45(1), 24-30.
EMEA 2006 Guidelines for Good Agricultural Practice (G.A.P.) of Medicinal and
Aromatic Plants. Working Copy no. 7.3, released 3rd April 2006.,
http://www.europam.net/GAP.htm [accessed on internet 9th February 2009].
Europam 2006 Guidelines for Good Wild crafting Practice (GWP) of Medicinal and
Aromatic Plants. Working Copy no. 5.3, released 3rd April 2006
http://www.europam.net/GWP.htm [accessed on internet 9th February 2009].
Evans, F. 1991. The separation of central from peripatetic effects on a structural basis.
Planta. Med., 57, 560-567.
Evans, W.C. 2002. Techniques in Microscopy. In: Evans, W.C. (Ed), Trease and Evans
Pharmacognosy. 15th Ed., Elsevier, London, pp. 538-547.
Fahn, A., 1988. Secretory tissues in vascular plants. New. Phytol., 108, 229-257.
191
References
Fairbairn, J.W., 1972. The trichomes and glands of Cannabis sativa L. Bull. Narc., 4
(004), 29-33.
Fairbairn, J.W., 1976. The Pharmacognosy of Cannabis. In: Graham, J.D.P. (Ed.),
Cannabis and Health. Academic Press, Amsterdam, pp. 3-19.
Farnsworth, N.R. and Morris, R.W. 1976. Higher plants the sleeping giant of drug
development. Am. J. Pharm. Educ. 148, 46-52.
Fellermeier, M., Eisenreich, W., Bacher, A., Zenk, M.H., 2001. Biosynthesis of
cannabinoids. Incorporation experiments with 13C labelled glucoses. Eur. J. Biochem.
268, 1596-1604.
FDA (Food and Drug Administration) 2004 Guidance for Industry – Botanical Drug
Products. http://www.fda.gov/cder/guidance/index.htm [accessed 13 Feb 2008].
Formukong, E.A., Evans, A.T. and Evans, F.J. (1998) Analgesic and anti-inflammatory
activity of constituents of Cannabis sativa L. Inflammation. 12 (4), 361-371.
Fournier, G., Richez-Dumanois, C., Duvesin, J., Mathieu, J-P.N., Paris, M., 1987.
Identification of a new chemotype in Cannabis sativa: Cannabigerol dominant Plants,
biogenetic and agronomic prospects. Planta. Med. 53 (3), 277-280.
192
References
Fowler, M.W. and Law, I., 2006. Plant-based pharmaceuticals – A strategic study
relating to UK activity and interests, [accessed 28th Feb 07],
http://www.nnfcc.co.uk/metadot/index.pl?id=3146;isa=DBRow;op=show.
Frank, M., 1997. Marijuana Grower’s Guide. Red Eye Press Inc., Los Angeles.
Fuchs L. (1542) De Historia Stirpium Commtarii Insignes. Octavo 2003 CD-Rom Edn,
Oakland, California.
Gang, D.R., Wang, J., Dudareva, N., Nam, K.H., Simon, J.E., Lewinsohn, E., Pichersky
E., 2001. An investigation of the storage and biosynthesis of phenylpropenes in sweet
basil. Plant. Physiol., 125, 539-555.
Gershenzon, J., 1994. Metabolic costs of terpenoid accumulation in higher plants.
J. Chem. Ecol., 20 (6), 1281-1328.
Gershenzon, J., McConkey, M.E., Croteau, R.B., 2000. Regulation of monoterpene
accumulation in leaves of peppermint. Plant. Physiol., 122, 205-214.
Gertsch, J., Leonti, M., Raduner, S., Racz, I., Chen, J-Z., Xie, X-Q., Altmann, K-H.,
Karsa, M., 2008. Beta-caryophyllene is a dietary cannabinoid. PNAS, (105) 26,
9099-9104.
193
References
Ghosal, S., Singh, S., Bhattacharya, S.K., 1971. Alkaloids of Mucuna pruriens:
chemistry and pharmacology. Planta. Med., 19 (3), 279–284.
Glover, B.J., 2000. Differentiation in plant epidermal cells. J. Exp. Bot., 51 (344)
497-505.
Green, G. 2003. The Cannabis Grow Bible., Green Candy Press, San Francisco.
Grinspoon, L. and Bakalar, J.B., 1995. Marijuana as a medicine – a plea for
reconsideration. J. Am. Med. Assoc. 273 (23), 185-186.
Hallahan, J. C. and Gray, D. J., 2000. Monoterpenoid Biosynthesis in Glanduar
Trichomes of Labiate Plants. In: Hallahan, D.J. and Gray J.C. (Eds.), Advances in
Botanical Research incorporating Advances in Plant Pathology: Vol. 31. Plant
Trichomes. Academic Press, San Diego. pp. 77-120.
Halliday, K.J. and Fankauser, C., 2003. Phytochrome-hormonal signalling networks.
New. Phytol., 157, 449-463.
Hamer, P.J.C., 1999. An algorithm to provide UK global radiation for use with models.
J. Compag., (22), 41-49.
Hammond, C.T. and Mahlberg, P.G. 1973. Morphology of glandular hairs of Cannabis
sativa from scanning electron microscopy. Am. J. Bot., 60, 524-528.
194
References
Hammond, C.T. and Mahlberg, P.G. 1977. Morphogenesis of capitate glandular hairs of
Cannabis sativa (Cannabaceae), Am. J. Bot. 64 (8), 1023-1031.
Hampson, A.J., Grimaldi, M. Axelrod, J., Wink, D., 1998. Cannabidiol and
Δ9-tetrahydrocannabinol are neuroprotective antioxidants. USA: Proc. Natl. Acad. Sci.
95, 8268-8273.
Hansard (Australia), 1996. Drugs of Dependence (Amendment) Bill. 15th May 1996,
p. 1238.
Hardwick, S. and King, L., 2008. Home Office Cannabis Potency Study 2008.
http://drugs.homeoffice.gov.uk/publication-search/cannabis/potency?view=Binary
[accessed on internet 8th May 2008].
Harlan J.R. and de Wet J.M.J., 1971, Towards a rational classification of cultivated
plants. Taxon. 20, 509 – 517.
Hawksworth, G. 2004. Metabolism and Pharmacokinetics of Cannabinoids. In: Guy,
G.W., Whittle, B.A., Robson, P.J. (Eds.), The Medicinal Uses of Cannabis and
Cannabinoids. Pharmaceutical Press, London, pp. 205-228.
Heinrich, M., Barnes, J. Gibbons, S., Williamson, E., 2004. General principles of botany:
morphology and systematics. In: Fundamentals of Pharmacognosy and Phytotherapy,
Churchill Livingstone, Edinburgh, pp. 23-31.
195
References
Henderson, S.M., 1973. Equilibrium conditions for hops when drying., J. Agric. Eng.
Res. 18, 55-58.
Herms, D.A. and Mattson, W.J., 1992. The Dilemma of Plants: To Grow or Defend. The
Q. Rev. Biol., 67(3), 283-335.
Heuvelinke, E., Batta, L.G.G., Daqmen, T.H.J., 1995. Transmission of solar radiation by
a multispan Venlo-type glasshouse : validation of a model. J. Agrformet., 74 (1-2),
41-59.
Hillig, K.W., and Mahlberg, P.G., 2004. A chemotaxonomic analysis of cannabinoid
variation in Cannabis (Cannabaceae). Am. J. Bot., 91, 966-975.
Hooke, R. 1665. Micrographia or some physiological description of minute bodies made
by magnifying glasses with observations and enquiries thereon.
http://www.gutenberg.org/ebooks/15491 [Accessed 17th Sep 2007].
Hough, M., Warburton, H., Few, B., May,T., Man, L-H., Witton, J., Turnbull, P.J., 2003.
A Growing Market. The Domestic Cultivation of Cannabis, UK: York Publishing
Services, York. p.7.
HLSCST (House of Lords Select Committee on Science and Technology) 1998a
Cannabis the Scientific and Medical Evidence, HL Paper 151. The Stationary Office:
London.
196
References
HLSCST (House of Lords Select Committee on Science and Technology) 1998b
Cannabis the Scientific and Medical Evidence, HL Paper 151-I. The Stationary Office:
London.
Howard, G.A., 1970. The determination of hop oil. J. Inst. Brew., 76, 381- 386.
Huestis, A.M., Henningfield, J.E., Cone, Z.C. 1992. Blood cannabinoids. I. Absorption
of THC and formation of 11-OH-THC and THC-COOH during and after smoking
marijuana., J. Anal. Tox., 18, 276-282.
Huffman, M.A., Page, J.E., Sukhedo, M.V.K., Gotob, S., Kalunde, M.S., Chandrasiri, T.,
Towers, G.H.N.. 1996. Leaf –swallowing by chimpanzees : A behavioural adaptation for
the control of strongyle nematode infections. Acta. Oecol., 17 (4), 475-503.
Iversen, L., 2007. The Science of Marijuana, Oxford University Press, Oxford.
pp. 52-57.
Jansen, M. and Terris, R., 2002. One woman’s work in the use of hashish in a medical
context. In: Russo, E. Dreher, M. Mathre, M.L., (Eds.), Women and Cannabis, Medicine,
Science and Technology, Haworth Press, New York, pp. 135 – 143.
Johnson, C.B., Kirby, J., Naxakis, G., Pearson, S., 1999. Substantial UV-B-mediated
induction of essential oils in sweet basil (Ocimum basilicum L.). Phytochemistry, 51,
507–510.
197
References
Johnson, J.R. and Potts, R., 2005, Cannabis-based medicines in the treatment of
cancer pain: a randomised, double-blind, parallel group, placebo controlled,
comparative study of the efficacy, safety and tolerability of Sativex® and Tetranabinex®
in patients with cancer-related pain. British Pain Society Annual Scientific Meeting
Poster Abstract, Edinburgh, Scotland, March 2005, p. 35.
Joy, J.E., Watson, S.J., Benson, J.A., 1999. Marijuana and Medicine Assessing the
Science Base. National Academy Press, Washington. pp. 137-192.
Kelsey, R.G., Reynolds, G.W., Rodriguez, E., 1984. The Chemistry of Biologically
Active Constituents Secreted and Stored in Plant Glandular Trichomes. In: Rodriguez,
E.P., Healey, L., Mehta, I. (Eds.), Biosynthesis of Terpenoids in Glandular Trichomes.
Plenum Press, London. pp. 187-241.
Kim, E.S. and Mahlberg, P.G.,1991. Secretory cavity development of glandular trichome
of Cannabis sativa L. (Cannabaceae). Am. J. Bot., 78, 142-151.
Kim, E.S. and Mahlberg, P.G., 2003. Secretory vesicle formation in the secretory cavity
of glandular trichomes of Cannabis sativa L. (Cannabaceae). Mol. Cells., 15 (3),
387-395.
King, L.A., Carpentier, C., Griffiths, P., 2004. European Monitoring Centre for Drugs and
Drug Addiction. Insights No. 6. An overview of cannabis potency in Europe.
Luxembourg: Office for the Publications of the European Communities.
198
References
http://www.emcdda.europa.eu/attachements.cfm/att_33985_EN_Insight6.pdf
[Accessed 30th Jan 2009].
Krings, M., Kellog, D., Derek, W., Kerp, H., Taylor, T.N., 2003. Trichomes of the seed
fern Blanzyopteris praedentata: implications for plant-insect interactions in the Late
Carboniferous. Bot. J. Linn. Soc, 141 (2), 133-149.
Langton, A. and Fuller, D., 2001. Supplementary Lighting of Pot Chrysanthemums.
Horticultural Development Council Technical Publication, pp. 1-40.
Ledbetter, M.C. and Krikorian, A.D., 1975. Trichomes of Cannabis sativa L as viewed
with scanning electron microscope. Phytomorphology, 25, 166-176.
Lehmann, T. and Brenneisen, R.J., 1995. High Performance Liquid Chromatographic
Profiling of cannabis Products. J. Liq. Chromatogr., 18 (4), 689-700.
Lenton, S. (2004), Pot, Politics and the Press – reflections on cannabis law reform in
Western Australia. Drug. Alcohol. Rev. 23, 223-233.
Lydon, J., Teramura, A.H., Coffman, C.B., 1987. UV-B radiation effects on
photosynthesis, growth and cannabinoid production of two Cannabis sativa
chemotypes. Photochem. Photobiol. 46 (2), 201-206.
199
References
Mahlberg, P.G., Hammond, C.T., Turner, J.C., Hemphill, J.K., 1984. Structure,
development and composition of glandular trichomes of Cannabis sativa L. In:
Rodriguez, E.P., Healey, L., Mehta. I., (Eds.), Biology and Chemistry of Plant
Trichomes. Plenum Press, London, pp. 23-51.
Malais, M.H. and Ravensberg, W.J., 1992. Knowing and Recognising . The Biology of
Glasshouse Pests and their Natural Enemies. Reed Business Information. Doetinchem,
Netherlands.
Malingré, T., Herndriks, H., Battermann, S., Bos, R., Visser, J., 1975. The essential oil
of Cannabis sativa L. Planta med, 28, 56-61.
Mannische, L., 1989. An Ancient Egyptian Herbal. British Museum Press. London.
Martone, G. and Della Casa, E., 1990. Analysis of the ageing processes in hashish
samples from different geographic origin. Forensic Sci. Int., 47, 147-155.
Matsuda, L.A., Lolait, S.J., Brownstein, M.J., Young, A.C., Bonner, T.I., 1990. Structure
of a cannabinoid receptor and functional expression of the cloned cDNA. Nature. 346
(6284), 561-564.
May, T., Warburton, W., Turnbull, P.J., Hough, M., 2002. Times they are a changing,
policing of Cannabis, York Publishing Services, York.
200
References
May, T., Duffy, M.., Warburton, W., Hough, M., 2007. Policing Cannabis as a Class C
drug. An arresting change? York Publishing Services, York. pp. 5-51.
McPartland, J.M., Clarke, R.C., Watson, D.P., 2000. Hemp Diseases and Pests
Management and Biological Control. CABI Publishing. Oxford. pp. 93-95.
McPartland, J. M. and Russo, E.B., 2001. Cannabis and Cannabis Extracts: Greater
Than The Sum Of Their Parts. Cannabis Therapeutics in HIV/AIDS: Howarth Press,
New York, pp. 103-132.
Mead, A., 2004. International control of cannabis: changing attitudes. In: Guy, G.W.,
Whittle, B.A., Robson, P.J. (Eds.), Pharmaceutical Press, London, pp. 369 – 426.
Mechoulam, R., 1986. The Pharmacohistory of cannabis sativa. In: Mechoulam, R.
(Ed.), Cannabinoids as Therapeutics Agents, CRC Press, Boca Raton, Florida.
Mechtler, K., Bailer, J., de Hueber, K., 2004. Variations of THC content in Single Plants
of Hemp Varieties. Ind Crop Prod, 19 (1), 19-24.
Mediavilla,V. and Steinemann, S., 1997. Essential Oil of Cannabis sativa L. strains.
J.I.H.A., 4 (2), 80-82.
Meier, C. and Mediavilla, V., 1998. Factors influencing the yield and the quality of hemp
(Cannabis sativa L.) essential oil. J.I.H.A., 5 (1), 16-20.
201
References
de Meijer, E.D.M., 1994. Diversity in Cannabis. PhD Thesis. University of Wageningen,
Netherlands.
de Meijer, E.P.M.,1994. Cannabis germplasm resources. In: Ranalli, P., (Ed.),
Advances in hemp research. Haworth Press, New York, pp.133-151.
de Meijer, E.P.M.., Bagatta, M., Carboni, A., Crucitti, P., Cristiana Moliterni, V.M.,
Ranalli, P., Mandolino, G., 2003. The inheritance of chemical phenotype in Cannabis
sativa L. Genetics, 163, 335-346.
Merck’s Manual, 1899, Merck, New York.
Merck Index, 1996, Twelfth Edition on CD-Rom, Chapman and Hall, London.
MHRA. 2007. Rules and Guidance for Pharmaceutical Manufacturers and Distributors.
Revised edition (27 April 2007). Pharmaceutical Press, London.
Miller, J.N. and Miller, J.C., 2005, Statistics and Chemometrics for Analytical Chemistry,
Pearson, Harlow, UK, pp. 186-187.
Morimoto, S., Tanaka, Y., Sasaki, K., Tanaka, H., Fukamizu, T., Shoyama, Y.,
Shoyama, Y., Taura, F. 2007. Identification and characterization of cannabinoids that
202
References
induce cell death through mitochondrial permeability transition in cannabis leaf cells.
J. Biol. Chem., 282 (28), 20739-20751.
Musty, R.E., 2004. Natural Cannabinoids: interactions and effects. In: Guy, G.W.,
Whittle, B.A., Robson, P.J. (Eds.) The Medicinal Use of Cannabis and Cannabinoids,
Pharmaceutical Press, London, pp. 165 – 204.
Mwenda, L., Ahmad, M., Kumari, K., 2005. Seizures on drugs in England and Wales,
2003. Findings 265. Home Office Research, Development and Statistics Directorate,
Office for National Statistics, http://www.homeoffice.gov.uk/rds/pdfs05/r265.pdf
[accessed 24th Jan 2007].
Neve R.A., 1991a. Hops.,Chapman and Hall, London, p. 4.
Neve, R.A., 1991b. Hops. Chapman and Hall, London, pp. 89-100.
National Institute of Agricultural Botany, 2007. Pocket Guide to Cereals, Oilseeds and
Pulses, N.I.A.B., Cambridge.
Ohlsson, A., Lidgren J.E., Wahlen A., Agurell S., Hollister L.E., Gillespie H.K. 1980.
Plasma delta-9-tetrahydrocannabinol concentrations and clinical effects after oral and
intravenous administration and smoking. Clin. Pharmacol. Ther., 28, 409-416.
203
References
Page, C.P., Hoffman, B., Curtis, M. and Walker, M (Eds). 2006, Integrated
Pharmacology, Mosby, Edinburgh.
Pate, D., Chemical Ecology of Cannabis. Journal of the International Hemp
Association., 2, 29, 32-37.
Payne, W.W., 1978. A glossary of plant hair technology. Brittonia, 30, (2), 239-255.
Pertwee, R.G., 1997. Pharmacology of cannabinoid CB1 and CB2 receptors.,
Pharmacol. Ther., 74, (2), 129-180.
Pertwee, R.G. 1998. Cannabis: The Scientific and Medical Evidence. House of Lords
Select Committee on Science and Technology. London: The Stationery Office:
pp. 70-83.
Pertwee, R.G. 2004. Receptors and pharmacodynamics: natural and synthetic
cannabinoids and endocannabinoids. In: Guy, G.W., Whittle, B.A., Robson, P.J., (Eds.),
Pharmaceutical Press, London, pp. 103-139.
Petersen, R.L. and Vermeer, J., 1984. Histochemistry of Trichomes. In: Rodriguez, E.,
Healey, P.L., Mehta, I., (Eds.), Biology and Chemistry of Plant Trichomes. Plenum
Press, London. pp. 71-94.
204
References
Phillips, G.F. 1998. Analytical and Legislative Aspects of Cannabis 1998 The
Cannabis Plant: Botany, Cultivation and Processing for Use. In: Brown, D.T., (Ed.),
Cannabis: the genus Cannabis.: Harwood Academic Publishers, Amsterdam.
pp. 71-114.
Pijlman, F.T.A., Rigter, S.M., Hoek, J., Goldschmidt, H.M.J., Niesink, R.J.M., 2005.
Strong increase in total delta-THC in cannabis preparations sold in Dutch coffee shops.
Addict Biol, 10, (2), 171–180.
Pliny: Natural History Vol. 6 (circa 60 AD). Translated by Jones, W.H.S., 1951. Harvard
University Press, Cambridge, Massachusetts.
Pongprayoon, U., Baeckstrom, P., Jacobsson, U., Lindstrom, M., Bohlin, I., 1991.
Antispasmodic Activity of ß-Damascenone and E-Phytol isolated from Ipomoea
pes-caprae. Planta Med. 58, 19-21.
Potter, D.J., 2004. Growth and Morphology of Medicinal Cannabis. In: Guy, G.W.,
Whittle, B.A., Robson, P.J. (Eds.), The Medicinal Uses of Cannabis and Cannabinoids.
Potter, D.J., Clark, P., Brown, M.B., (2008). Potency of delta 9-THC and other
cannabinoids in Cannabis in England in 2005: Implications for psychoactivity and
pharmacology. J. Forensic. Sci. 53, (1), 90-94.
205
References
Raman, A.R., 1998. The Cannabis Plant: Botany, Cultivation and Processing for Use.
In: Brown D.T., (Ed.), Cannabis The Genus Cannabis, Harwood Academic Publishers,
Amsterdam. pp. 29-54.
Raman, A.R. and Joshi, A., 1998. The Chemistry of Cannabis, In: Brown, D.T., (Ed.),
Cannabis The Genus Cannabis, Harwood Academic Publishers, Amsterdam, pp. 55-70.
Rhodes, D.F. 1977. Integrated antiherbivore, antidesiccant, and ultraviolet screening
properties of creosote bush resin. Biochem. Syst. Ecol. 5, 281-290.
Rigby, F.L., 2000. Process and apparatus for obtaining lupulin from hops. Patent No
WO/2000/006691.
Roberecht, R. and Caldwell, M.M., 1980. Leaf ultraviolet optical properties along a
latitudinal gradient in the arctic-alpine life zone. Ecology, 61, 612-619.
Robson, P., 1999a. Forbidden Drugs. 2nd Ed. Oxford University Press, Oxford. pp.
66-86.
Robson, P., 1999b. Forbidden Drugs. 2nd Ed. Oxford University Press, Oxford. pp. 1-4.
Room, R., Fischer B., Hall, W., Lenton, S., Reuter, P., Cannabis Policy: Moving Beyond
Stalemate. The Beckley Foundation Global Cannabis Commission Report. 2008.,
206
References
http://www.beckleyfoundation.org/pdf/BF_Cannabis_Commission_Report.pdf,
[accessed 1 Dec 2008].
Rosenthal, E., 1998. The Marijuana Grower’s Handbook (paperback). Quick Trading
Co. Oakland, California.
Rosenthal, E. 2001. The Big Book of Buds. In: Newhart, S. (Ed.), Quick American
Archives, Oakland, California, p. 214
Ross, R.A., 2007. Allosterism and cannabinoid CB1 receptors: the shape of things to
come. Trends Pharmacol. Sci., 28, 11.
Ross, S.A. and ElSohly, M.A., 1997. CBN and ∆9-THC concentration ratio as an
indicator of the age of stored marijuana samples. Bull. Narc., 1, 008.
Rowan, M.G. and Fairburn, J.W., 1977. Cannabinoid patterns in seedlings of Cannabis
sativa. J. Pharm. Pharmac. 29, 491-494.
Russo, E.B., 2003, From Pariah to Prescription. Howarth Press, New York, p. 3.
Russo, E.B. 2004. History of cannabis as a medicine. In: Guy, G.W., Whittle, B.A.,
Robson, P. J., (Eds.), The Medicinal Uses of Cannabis and Cannabinoids.:
207
References
Russo, E. B., Jiang, H. E., Li, X., Sutton A., Carboni, A.. del Bianco F, Mandolino, G.,
Potter, D. J., Zhao Y.-X., Bera, S. Zhang Y.- B., Lü, E.-G., Ferguson, D. K., Hueber, F.,
Zhao, L.-C., Liu C.-J., Wang, Y.-F., Li. C.-S., 2008. Phytochemical and genetic analyses
of ancient cannabis from Central Asia. J. Exp. Bot. 59, 15, 4171-4182.
Ryan, D., Drysdale, A.J., Pertwee, R.G., Platt, B., 2007. Interactions of cannabidiol
with endocannabinoid signalling in hippocampal tissue. Eur. J Neuroscience. 25,
2093-2102.
Samuelsson, G., 1999. Drugs of Natural Origin, A Textbook of Pharmacognosy, 4th Ed.,
Swedish Pharmaceutical Press, Stockholm, p. 31.
Schultes, R.E., 1970, In: Joyce, C.R.B. and Curry, S.H., (Eds.), The Botany and
Chemistry of Cannabis. J and A Churchill, London. p.11.
Schultes, R.E., Klein W.M., Plowman T., Lockwood T.E., 1974. Cannabis: an example
of taxonomix neglect. Harvard University Botanical Museum Leaflets 23, 337–367.
Shade, R.R., Thompson, T.E. and Campbell, W.R., 975. An Alfafa Weevil Larval
Resistance Mechanism Detected in Medicago., J. Econ. Entomol., 68, 399-404.
Shoyama, Y., Yagi, M., Nishoika, I., Yamouchi, T., 1975. Biosynthesis of cannabinoid
acids, Phytochemistry, 14, 2189-2192.
208
References
Sirikantaramas, S., Taura, F., Tanaka, F., Ishikawa, Y., Morimoto, S., Shoyama, S.,
2005. Tetrahydrocannabinolic acid synthase, the enzyme controlling marijuana
psychoactivity is secreted into the storage cavity of the glandular trichomes. Plant. Cell.
Physiol., (46), 9, 1578-1582.
Small, E. and Beckstead, H.D.,1973a, Common cannabinoid phenotypes in 350 stocks
of Cannabis, Lloydia, 36,144-165.
Small, E. and Beckstead, H.D., 1973b. Cannabinoid phenotypes in Cannabis sativa L.
Nature, 245, 147-148.
Small, E., 1976., The forensic taxonomic debate on Cannabis: semantic hokum.
J. Forensic. Sci. 21, 239 – 251.
Small, E. and Cronquist, A., 1976. A Practical and Natural Taxonomy for Cannabis,
Taxon, 25, (4), 405-435.
Smith, F.E., 1951. Tetrazolium Salt. Science, 113, 751-754.
Smith, N., 2005. High potency cannabis: the forgotten variable. Addiction, 100:
1558–9.
Snoeijer, W. 2002. A checklist of some Cannabaceae cultivars, Part a, Cannabis.,
Division of Pharmacognosy, Amsterdam Centre for Drug Research, Leiden, pp. 5-7.
209
References
Struik, P.C., Amaducci, S., Bullard, M.J., Stutterheim, N.C., Venturi G., Cromack,
H.T.H., 2000. Ind. Crop. Prod., 11 (2-3), 107-118.
Sumner, J., 2000. The Natural History of Medicinal Plants.: Timber Press, Portland,
Oregon. pp. 145-161.
Systma, K.J., Morawetz, J. C., Pires, M., Nepokroeff, E., Conti, M., Zjhra, J. C., Hall, M.
Chase, W., 2002. Urticalean rosids: circumscription, rosid ancestry, and phylogenetics
based on rbcL, trnL-F, and ndhF sequences. Am. J. Bot, 89, 1531-1546.
Szendrei, K. 1997. Cannabis as an illicit crop: recent developments in cultivation and
production quality. Bull. Narc., XLIX and L (1/2) 1-21.
United Nations Office on Drugs and Crime, World Drug Report 2006. Chapter 2,
Cannabis: Why we should care. [accessed 30 Jul 2007].
http://www.unodc.org/pdf/WDR_2006/wdr2006_chap2_why.pdf.
Tanaka, S., Yamamura, T., Shigemoto, R., Tabata, M., 1989. Phytochrome mediated
production of monoterpenes in thyme seedlings. Phytochemistry. 28, 2955–2957.
Taura, F., Morimoto, A., Shoyama, Y. 1996. Purification with characterization of
cannabidiolic acid synthase from Cannabis sativa L. J. Biol. Chem., 271 (29):
17411-17416.
210
References
Taura, F., Sirikantaramas, S., Shoyama, Y., Morimoto, S., 2007. Phytocannabinoids
in Cannabis sativa : Recent Studies on Biosynthetic Enzymes. Chem and Biodivers, 4,
1649-1663.
Theis, N. and Lerdau, M., 2003. The evolution of function in plant secondary
metabolites. Int. J. Plant. Sci., 164 (3), S93-S102.
Theobald, W.L., Krahulik, J.L., Rollins, R.C., 1980. Trichome Description and
Classification In: Metcalfe, C.R. and Chalk, L. (Eds.), Anatomy of Dicotyledons,
Clarendon Press, Oxford, pp. 40-53.
Thompson, W.W. and Healey, P. L., 1984. Cellular Basis of Trichome Secretion. In:
Biology and Chemistry of Plant Trichomes. Rodriguez, E., Healey, P.L., Mehta, I.,
Plenum Press, London. pp. 95-111.
Turner, A.E., El Sohly, M.A., Boeren, E.G., 1980. Constituents of Cannabis sativa L.
XV11. A Review of the Natural Constituents. J. Nat. Prod., 169-233.
Turner, J.C., Hemphill, J K. , Mahlberg, P.G., 1977. Gland distribution and cannabinoid
content in clones of Cannabis sativa L. Am. J. Bot., 64 (6): 687-693.
Tyler V.E., Brady, L.R., Robbers, J.E., 1988. Pharmacognosy, 9th Ed., Lea and
Febiger, Maryland, USA. p. 7.
211
References
UNODC. United Nations Office on Drugs and Crime, 2006 World Drug Report., United
Nations Publications, New York, Vol. 2. p. 105.
Uphof, J.C.T., 1962. Plant Hairs. Encyclopedia of Plant Anatomy IV., Gebruder
Borntraeger, Berlin, 5, pp. 1-206.
Vale, S., 2008. More yield less work. The Commercial Greenhouse Grower. ACT
Publishing, Maidstone, pp. 31-33.
Van der Werf, H.M.G., Brouwer, K., Wijlhuizen, M., Withagen, J.C.M., 1995. The effect
of temperature on leaf appearance and canopy establishment in fibre hemp (Cannabis
sativa L.), Ann. Appl. Biol. 126, 551-561.
Vincent, R.M., Lopez-Meyer, M., McKnight, T.D., Nessler, C.L., 1997. Sustained harvest
of camptothecin from the leaves of Camptotheca acuminata. J. Nat. Prod. 60, 618-619.
Wade, D.T., Makela, P., Robson, P., House, H., Bateman, C., 2004. Do cannabis-based
medicinal extracts have general or specific effects on symptoms in multiple sclerosis? A
double-blind, randomized, placebo-controlled study on 160 patients. Multiple Sclerosis.
10 (4), 434-441.
Wagner, G.J., 1990. Secreting Glandular Trichomes: More Than Just Hairs. Plant
Physiol., 96, 675-679.
212
References
Wagner, G.J., Wang, E., Shepherd, W., 2004. New Approaches for Studying and
Exploiting an Old Protuberance, the Plant Trichome. Ann. Bot. (London), 93. 3-11.
Wall, P. 1998. Memorandum to Science and Technology Committee. House of Lords
Select Committee on Science and Technology. Cannabis: The scientific and medical
evidence., H.M.S.O., London, 151 (1) 31-32.
Walters, D.S., Harman, J., Craig, R., Mumma, R.O., 1991. Effect of temperature on
glandular trichome exudate composition and pest resistance in geraniums. Entomol.
Exp. Appl. 60, 61-69.
Ware, M.A., Adams, H. , Guy, G.W., 2005. The medicinal use of cannabis in the UK:
results of a nationwide survey. Int. J. Clin. Pract., 59 (3), 291-295.
Werker, E. 2000., Trichome Diversity and Development. In: Hallahan, D.L.. Gray, J.C.,
Callow, J.A. (Eds.), Advances in Botanical Research incorporating Advances in Plant
Pathology Plant Trichomes, Academic Press, London, 31, pp. 1-35.
Whittle, B.A., 2004. The Medicinal Uses of Cannabis and Cannabinoids. In: Guy, G.W.,
Whittle, B.A., Robson, P.J. (Eds.), Pharmaceutical Press, London, p XII (Preface).
Wilkinson, J.D., Whalley B.J., Baker, D., Pryce, G., Constanti, A., Gibbons, S. and
Williamson, E.M., 2003. Medicinal cannabis: is – delta 9 - tetrahydrocannabinol
necessary for all its effects? J. Pharm. Pharmacol., 55, 1687-1694.
213
References
Wilkinson, T. J., 2006. Cannabis Growth Medium. Confidential GW Pharmaceutical Ltd
Standing Operating Procedure.
Williamson, E.M. and Evans, F.J. 2000. Cannabinoids in Clinical Practice. Drugs, 60 (6),
1303-1314.
Williamson, E.M., 2001. Synergy and other interactions in phytomedicines.
Phytomedicine 8 (5), 401-409.
Wills S., 1998. Cannabis use and abuse by man: a historical perspective. In: Brown D.T.
(Ed.), Cannabis: The Genus Cannabis, Harwood Academic Publishers. Amsterdam,.
pp. 1-28.
Yamaura, T. 1992. Localisation of the biosynthesis and accumulation of
monoterpenoids in glandular trichomes of thyme. Planta. Medica. 58 (2), 153-158.
Zajicek, J., Fox, P., Sanders., Wright, D., Vickery, J., Nunn, A., Thompson, A., 2003.
Lancet, 362, 1517-1513.
Zias, J., Stark., Levy, R., Werker, E., Breuer, A., Mechoulam, R., 1993. Nature, 363,
215.
214
Appendix 8.1 Analysis of cannabinoid content using HPLC and GC
Appendix 8.1 Analysis of cannabinoid content using HPLC

and GC
8.1.1 HPLC Analysis
8.1.1.1 HPLC and GC Equipment
Apparatus Source
Sonicator Model UR-324T Gemini B.V., Elsweg 57,
7311 GV, Apeldoorn, Netherlands
Hewlett Packard/Agilent 1100 HPLC Agilent Technologies UK Ltd.
incorporating: - South Queensferry,
West Lothian.
Autosampler DE594901133
Diode Ray Detector DE91605830
Column Oven DE91609412
Mobile Phase Degaser Unit DE9605830
Discovery C8 150 x 4.6 mm column and

30 x 4.6 mm precolumn packed with 5μm
Kingsorb ODS packing material.
Personal computer with HP Chemstation

software.
Sonicator Model UR-324T Gemini B.V., Elsweg 57, 7311
GV, Apeldoorn, Netherlands
Hewlett Packard 6890 GC with FID incorporating Pre-owned item
an HP-5 320 μm x 30 m column with a 0.25 μm
film and an autoanalyser
Table 8.1.1 Analytical Equipment
215
8.1.1.2 Analytical Materials
Materials Source
11-Nor-Δ9-tetrahydrocannabinol-9- Sigma Aldrich Ltd., Fancy Road, Poole,
carboxylic acid 50 μg/mL in methanol. Dorset
CBGA Applied Analysis Ltd. Rowley House,

CBDA Tokenspire Business Park,
CBCA Beverley, Yorkshire.
Chloroform HPLC Grade Ultrafine Ltd. Marlborough House, 298
Regents Park Road, London
Methanol HPLC Grade Fisher Scientific UK Ltd. Bishops
Meadow Lane, Loughborough, Leics.
Table 8.1.2 Analytical Materials
8.1.2 HPLC determination of cannabinoid content
Previous forensic analysts have reported that investigations into cannabis potency are
made difficult by the inhomogeneous nature of herbal cannabis. Within a well-mixed
single large batch of crude material, and following removal of unwanted matter,
different aliquots could lead to quite different analytical results (King 2004). Batches of
the ‘naturally inhomogeneous material’ were well mixed before sampling. A minimum of
three subsamples was analysed if sufficient material was available.
The botanical raw material is thoroughly mixed. Five small samples of approximately 1
g are taken at random from the mixture and blended. From this, a single sub-sample of
100 mg is extracted with 1.0 ml methanol-chloroform (9:1 v/v) by sonication for fifteen
minutes. 100 μl of the filtered extract is diluted with 300 μl of methanol and aliquots of
1 μl used for HPLC.
Using a Discovery C8 150 x 4.6 mm column and a 30 x 4.6 mm precolumn containing

5μm Kingsorb ODS packing material, an operating temperature of 25ºC and a UV
wavelength of 220 nm were adopted. The run time was 23 minutes and THC the
internal standard.
Apart from THC, which was purchased from Sigma-Aldrich, most cannabinoid
analytical standards were not available at the commencement of this study.
Cannabinoids such as CBGA were identified and quantified in the HPLC trace by
216
comparing chromatograms with those produced by Lehmann and Brenneisen (1995).

CBGA, CBDA and CBCA were later purified and identified at Applied Analysis Ltd
(Flockhart pers comm.) and the original identification of their HPLC peaks confirmed.
A chromatogram produced when analysing an extract of Clone Line G1 M1 is shown in

Figure 8.1.1. The CBGA and THCA peaks are observed at 6.56 and 15.45 minutes
respectively. Following the later acquisition of a CBGA standard it was ascertained
that when equimolar preparations of THCA and CBGA are evaluated by HPLC using
this method, CBGA produces a peak area greater than THCA by a factor of 1.2.
Concentrations of the two in this thesis have been calculated accordingly. Corrections
for the peak characteristics of other cannabinoids have been similarly implemented.
217
Figure 8.3.1 Chromatogram of a preparation from Clone G1 M1.
218
8.1.3 GC determination of cannabinoid content
The botanical raw material was thoroughly mixed. Five small samples are normally
taken at random from the mixture and blended. From this, a single sample 50 mg was
taken and prepared for analysis.
The samples were then prepared and analysed as developed by de Meijer et al 2003. I
ml of ethanol (>99.7%) was added to the filtration tube and the sample sonicated for 15
minutes and the extract then centrifuged at 4000 rpm for 10 minutes. This procedure
was then repeated a further three times and the resultant 4 ml of ethanol containing the
cannabinoid extracts then transferred to a five ml volumetric flask. 0.25 ml of a
phenanthrene stock solution (10mg/ml) in ethanol was added as an internal standard
and adjusted to 5 ml with ethanol. Extracts were homogenised and transferred to GC
vials.
Gas-chromatographic analyses were performed on a Hewlett Packard 6890 GC

equipped with an autoanalyser, a flame ionization detector and an HP-5 320 μm x 30 m
column with a 0.25 μm film.
219
Appendix 8.2 Analysis of terpene content using GC
8.2.1 Analysis Equipment
Apparatus Source
Gas Chromatograph with split/splitless capillary Agilent Technologies UK Ltd.
injector (Agilent Technologies 5890 or 6890) South Queensferry,
Flame ionisation detector (FID) and/or Electron West Lothian.

Ionisation Mass Spectrometer (EI-MS)
Autosampler DE594901133
Diode Ray Detector DE91605830

Column Oven DE91609412
Mobile Phase Degaser Unit DE9605830
Discovery C8 150 x 4.6 mm column and

30 x 4.6 mm precolumn packed with 5μm Kingsorb
ODS packing material.
Personal computer with HP Chemstation software.
Table 8.2.1 Analytical equipment
8.2.2 GC Analysis Method
The botanical raw material was thoroughly mixed. 1 g samples are taken at random
and extracted with approximately 40 ml of methanol:chloroform (9:1 v/v) in a 50 ml
volumetric flask by sonication for thirty minutes at 25ºC. The solution is allowed to cool
and the volume made up to 50ml with methanol:chloroform (9:1 v/v). An aliquot of each
extract is centrifuged at 3000 rpm for 2 minutes. 1.5ml of the supernatant is added to a
2ml auto-sampler vial, capped and securely crimped prior to analysis. Injector and
Detector temperatures are 250ºC and 325ºC respectively. The injection volume is 1 µl
with a split ration of 5:1. The FID fuel gases were Hydrogen 40 ml min-1, Air 50ml
min-1, Helium 45 ml min-1.
220
Groups of terpenes may be identified at retention times of approximately 4-5 minutes

for monoterpenes and 14-20 minutes for sesquiterpenes. Peak identifications may be
made by retention time comparisons with certified standards and by using gathered
mass spectra. For generation of accurate quantitative data the losses and uncertainties
of the injection and volatilisation process can be reduced by using an internal standard
- Phenanthrene at 1 mg ml-1.
The assay methods described in the above two appendices have been validated in
accordance with ICH Guidelines and have subsequently been included in product
license applications. The exact details of this validation remain the intellectual property
of GW Pharmaceuticals Ltd.
221
Appendix 8.3 A Comparison of the Structure of CB1 and CB2 Receptors
Appendix 8.3 A Comparison of the Structure of

CB1 and CB2 Receptors
a)
b)
The structure of CB1 (a) and CB2 receptors (b) in human tissue. Above the cell
membrane is the extra-cellular N-terminal and below the intracellular C-terminal.
222
Appendix 8.4 Meteorological Data pertaining to UK Field Trials
Appendix 8.4 Meteorological Data pertaining to UK Field

Trials
2000 2003 2004 2005 2006 Mean
May 12.7 12.2 12.5 11.8 8.4 11.5
Jun 15.3 16.3 15.9 16.0 16.2 15.9
Jul 15.7 17.9 16.5 17.3 20.3 17.5
Aug 17.2 19.2 18.0 16.7 16.9 17.6
Sep 15.3 14.7 15.4 16.0 17.3 15.7
8.4.1 Mean daily temperatures in the field trial region as recorded by the Meteorological
Service.
2000 2003 2004 2005 2006 Mean
May 196 202 206 224 168 192
Jun 174 224 232 215 274 224
Jul 174 206 189 202 311 216
Aug 215 240 202 242 181 216
Sep 128 194 173 158 164 163
8.4.2 Total number of sunshine hours each month in the field trial region as reported by the
Meteorological Service.
The above data was downloaded from the Meteorological Office web page: -
http://www.metoffice.gov.uk/climate/uk [accessed on 6th February 2009].
223
Appendix 8.5 Definitions of Social Grade as defined by National Readership Survey
Appendix 8.5 Definitions of Social Grade as defined by

National Readership Survey
The Social Grade classification used by NRS was developed for the Survey over 50
years ago. It remains a highly effective way of classifying readers of different
publications, and is widely used in planning advertising, not just in newspapers and
magazines, but in other media too. The grades were also used by the Government
when compiling the UK Population Census.
Social Grade is determined by the occupation of the Chief Income Earner (CIE) in each
household. Additional criteria such the size of the organisation, and the number of
people for which the CIE is responsible, are used to refine the process.
A brief description of the six grades is as follows:
Social Status Occupation
A Upper Middle Class Higher managerial, administrative or professional

B Middle Class Intermediate managerial, administrative or professional
Supervisory or clerical and junior managerial,
C1 Lower Middle Class
administrative or professional
C2 Skilled Working Class Skilled manual workers

D Working Class Semi and unskilled manual workers
Those at the lowest levels Casual or lowest grade workers, pensioners and others
E
of subsistence who depend on the state for their income
224

Potter Thesis 2009

Uploaded by

Copyright:

Available Formats

Potter Thesis 2009

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Potter Thesis 2009

Uploaded by

Copyright:

Available Formats

THE PROPAGATION, CHARACTERISATION AND

OPTIMISATION OF CANNABIS SATIVA L

In fulfilment of the requirement

Department of Pharmaceutical Science Research

King’s College London

In response to known pharmacology, and an increasing weight of anecdotal evidence of

effective alternatives often unavailable, otherwise law-abiding UK patients have regularly

knowledge of the form and function of trichomes in Cannabis sativa L. Supporting

chemical analyses ascertain what secondary metabolites are biosynthesised within

demands of the pharmaceutical industry, and in marked contrast to illicit cannabis, a

phytopharmaceutical feedstock must meet high expectations regarding the minimum

and maximum content of a range of compounds. Specific studies are performed to

environmental expectations. This involves studying plant development and secondary

cannabis-based botanical medicine supported by this research is Sativex®. This

Awarded European Plant Variety Rights EU 16301, 10th Oct 2005

To my best mate John Pook

I express my sincerest gratitude to GW Pharmaceuticals Ltd, particularly Dr Geoffrey

especially grateful to Kathy Hammond for diligently analysing so many of my samples

and Heather North for administrative assistance and pastoral care.

co-speaker at several inspirational Multiple Sclerosis Group Meetings. My thanks go to

give a presentation at a memorable Yorkshire Pain Management Group Meeting at

St Stephens College – an experience that opened my eyes to the dedicated work of

Cannabinoids. Pharmaceutical Press, London. pp. 17-54.

in cannabis in England in 2005: Implications for psychoactivity and pharmacology.

J Forensic Sci 53 (1): 90-94.

Chapter 3 Cannabis trichome form, function, and distribution ...............................47

Chapter 4. The Function and Exploitation of Secondary Metabolites from

Chapter 6 The Outdoor Propagation of Phytopharmaceutical Cannabis .............143

Chapter 7 GENERAL DISCUSSION....................................................................174

chemotype G5 M16 cv Gill, (b) THC chemotype G1 M3 cv Guinevere (c) Afghan

landrace clone M146 (Illustrations by Valerie Bolas, commissioned by GW

reflexed to expose the anthers. 5

Geranylpyrophosphate; 2, Divarinic Acid (R1) or Olivetolic Acid (R2); 3,

Cannabigerovarin (CBGV) (R1) or Cannabigerol (CBG) (R2);. 4, Δ9-

tetrahydrocannabivarin (R1) or Δ9-tetrahydrocannabinol (R2). R1 (-C3H7) and

R2(-C5H11) indicate the propyl or pentyl forms of the metabolites; Enzyme I:

geranylpyrophosphate:olivetolate geranyltransferase(GOT); Enzyme II: THC(V)

found in the plastid and cytosol respectively. 20

(Alberts et al., 2002) 22

Compressed herbal cannabis material with selection of removed seeds. 31

separate packets, each containing approximately one gram. 32

CBD and CBN resin (n = 169). 37

constabularies in 2004/5 (n = 169). 38

al. (2008) (n = 33). 39

by police in Derbyshire (n=15), Kent (n=58), London Metropolitan (n = 96), Merseyside (n

= 44) and Sussex (n = 34) in 2004/5 (total n = 247). 40

stalked trichomes to be present only within the proximal region. 55

symmetrically opposite sides of the midrib. 57

using incident light. 59

low-power microscope. (b) closer view of antherial trichomes. 61

Figure 3.8. (a) A temporarily dry-mounted capitate stalked trichome viewed in

transmitted light. An irregular arrangement of poorly-defined secretory cells is visible at

trichome on Cannabis sativa by Briosi and Tognini (1894). 64

specimen is temporarily mounted in a 70% v/v aqueous solution of glycerol and

illuminated from below. 66

cystolythic trichomes, viewed after twelve hours in 0.1% tetrazolium solution. 67

on (M280). The specimen was viewed in incident light. 68

temporarily mounted in oil. 70

photographed with a tripod-mounted camera incorporating a macro lens. The

of capitate stalked trichomes. 72

red). (* p < 0.05, p < 0.01, * p < 0.001). 128