Euphytica (2020)216:6

https://doi.org/10.1007/s10681-019-2544-9 (0123456789().,-volV)
(0123456789().,-volV)

RNA-Seq analysis of Orobanche resistance in Nicotiana

tabacum: development of molecular markers for breeding
recessive tolerance from ‘Wika’ tobacco variety
E. Julio . A. Malpica . J. Cotucheau . S. Bachet . R. Volpatti .
C. Decorps . F. Dorlhac de Borne

Received: 29 July 2019 / Accepted: 5 December 2019

 Springer Nature B.V. 2019

Abstract Orobanche spp. (broomrape) is an obli- markers including SNPs or genes differentially
gate root parasite that can attack a wide spectrum of expressed between susceptible and resistant lines
plants, including tobacco. It has been responsible for were identified. An F2 population segregating for
economic losses in Europe since 2002 and its ‘Wika’ recessive tolerance was then used for marker
incidence in many tobacco-growing countries is validation and mapping. All candidates were situated
increasing. Preventive and curative methods exist, on chromosome 14 of the tobacco genetic map. The
including the use of agrochemicals, however efficacy Nicotiana variety collection from Imperial Brands was
is limited and pest dissemination remains important also tested for these markers, highlighting or confirm-
due to a high rate of multiplication of the parasite and ing other potential tolerant lines. KASPTM genotyping
very small long lasting seeds. The tobacco variety or markers for conventional gel electrophoresis are
‘Wika’ induces lower or delayed germination of now available to drive the transfer of ‘Wika’ recessive
Orobanche seeds. This seems to be conditioned by a tolerance into elite lines. RNA-Seq technology com-
single recessive gene (Cailleteau et al. in: CORESTA bined with sound experimental testing has again
Congress, Paris, 2006). Artificial testing in Petri dishes proven its high efficiency to identify useful markers
was developed to evaluate the ability of tobacco for tobacco breeding.
plantlets to stimulate seed germination. Different lines
derived from ‘Wika’, with susceptible control lines, Keywords Orobanche  Parasite tolerance 
were tested and studied by RNA-Seq. Candidate Tobacco  RNA-Seq  Germination inhibition

Electronic supplementary material The online version of

this article (https://doi.org/10.1007/s10681-019-2544-9) con-
tains supplementary material, which is available to authorized Introduction
users.
Plant-parasite species from Orobanche and Pheli-
E. Julio (&)  J. Cotucheau  R. Volpatti 
panche (broomrapes) genera belong to the Oroban-
C. Decorps  F. Dorlhac de Borne
Leaf Research, SEITA - Imperial Brands, La Tour, chaceae family, and are responsible for major food
24100 Bergerac, France crop losses worldwide (Parker 2009). Broomrapes are
e-mail: Emilie.julio@fr.imptob.com holoparasites meaning total absence of chlorophyll
rendering the parasite completely dependent on host
A. Malpica  S. Bachet
Bergerac Seed and Breeding, La Tour, 24100 Bergerac, water and nutrients (Aly 2007).
France

123
6 Page 2 of 18 Euphytica (2020)216:6

The life cycles of these parasitic plants share Buschmann et al. 2005; Cardoso et al. 2011). Within
several features, namely a profuse production of Solanaceae family, branched broomrape (O. ramosa)
robust and long-lasting seeds, known to remain viable is a particularly problematic for crop culture of
for over a decade of dormancy; detection mechanisms eggplant, tomato and tobacco, accounting for an
to synchronize germination with host root develop- average yield loss of 30% (Parker 1994), and reaching
ment (Xie et al. 2010; Waters et al. 2017; Cardoso up to 50–100% for tomato cultures in some countries
et al. 2011) and formation of haustoria to attach and (Parker and Riches 1993). In the case of tobacco,
penetrate the host root system, thus establishing additional losses due to effects on quality have been
connections by which they obtain water and nutrients reported. In France, hemp, winter rape and tobacco
to thrive (Brun et al. 2017; Fernández-Aparicio et al. cultures are increasingly affected by broomrape; and
2016; Yoshida and Shirasu 2012; Yoshida et al. 2016; growing geographical infestation and virulence vari-
Cardoso et al. 2011). As a result, host plant develop- ability are also observed, which presage further
ment is stunted, resulting from biomass reduction agricultural difficulties ahead (Brault et al. 2007;
(affecting reproductive tissue most significantly) Gibot-Leclerc et al. 2003). However, precise estima-
(Barker et al. 1996; Manschadi et al. 1996; Lins tions of current economic impact of O. ramosa require
et al. 2007), and is directly correlated to parasite more reliable survey data, which could help assess
biomass accumulation (Cardoso et al. 2011). In local and global infestation rates and crop damage
addition, the particular mechanism of root parasitism caused by this particular broomrape species (Parker
involved enables the plant to evolve underground, 2009), in view of developing pest control strategies
therefore undetected by farmers. By the time shoot and crop protection.
development and flowering occur, irreversible damage Host plants have nonetheless developed rudimen-
to the host plant has already been caused. The broad tary resistance or tolerance mechanisms to reduce
spectrum of hosts, seed durability and parasitic sink parasite invasion, acting at different levels and stages
metabolism sustained by a strong osmotic pressure all of parasite development to impede germination,
contribute to strengthen the parasite on cultivated attachment or establishment, for example (Labrousse
plants. Moreover, broomrape genetic diversity and et al. 2001; Eizenberg et al. 2003; Yoder and Scholes
adaptation mechanisms have been documented, both 2010; Rubiales 2003, 2014; Fernández-Aparicio et al.
of which reinforce overall parasite virulence (Brault 2016). Host plant features such as very early or very
et al. 2007). Finally, the physical proximity of the late root growth, reduced biomass or modifications of
parasitic plants with their hosts and their common root exudates may confer infection escape mecha-
physiological pathways imply that mechanical and nisms. Resistance or tolerance strategies to thwart
chemical interventions are not viable solutions for pest parasite establishment may also include combinations
eradication. Cultural, mechanical, biological and of root cell wall modifications, extensive deposition of
chemical strategies aiming to reduce the agricultural callose or suberin, lignification and production of toxic
impact of the Orobanchaceae exist, however none of metabolites (Reactive Oxygen Species, for example)
the current approaches have demonstrated sufficient at the penetration site, sink-strength competition to
efficiency in the face of such invasive parasitism reduce nutrient outflow through parasite attachment
(Fernández-Aparicio et al. 2016; Habimana et al. sites, among other means that have been reviewed
2014). The particular evolutionary advantages that elsewhere (Fernández-Aparicio et al. 2016 and refer-
Orobanchaceae have acquired therefore present sig- ences therein). Reduction of parasite seed germination
nificant challenges for infestation control. is of particular interest, as it may be a natural way of
Roughly 170 species of Orobanche exist, seven of avoiding parasitism altogether. Orobanche spp. seeds
which are particularly specialized crop parasites (O. germinate in response to the detection of chemical
aegyptiaca, O. crenata, O. cernua, O. cumana, O. signals emanating from developing roots of target
foetida, O. minor and O. ramosa), possessing high plants; a vast majority of these molecules belong the
parasitic potential enabling them to parasitize a broad strigolactone (SL) family, which are primarily
spectrum of hosts from Asteraceae, Brassicaceae, involved in promoting the development of beneficial
Apiaceae, Fabaceae and Solanaceae genera in central fungal symbiosis within the rhizosphere (Cavar et al.
and Eastern Europe, Africa and Asia (Joel et al. 2007; 2015; Brun et al. 2017). The structural diversity of SL

123
Euphytica (2020)216:6 Page 3 of 18 6

molecules is circumvented by Orobancheae ability to markers were identified to support the mechanism of
respond to numerous molecular cues by adapting SL ‘Wika’ Orobanche spp. resistance (Brault-Hernandez
detection mechanisms to allow for lower specificity 2006). Little is known regarding ‘Wika’ resistance,
(for review see Cardoso et al. 2011; Brun et al. 2017; except that it has been demonstrated as recessive
Fernández-Aparicio et al. 2016). Lower germination (Cailleteau et al. 2006; Brault-Hernandez 2006).
induction has been shown to be as a resistance In the current study, we set out to identify genetic
mechanism in certain varieties of faba bean, tomato, sequences involved in ‘Wika’ resistance mechanisms
rice and pea (Fernández-Aparicio et al. 2011, 2014; and develop molecular markers for screening and
Rubiales et al. 2016; Abbes et al. 2010; Trabelsi et al. breeding of broomrape-resistant tobacco varieties. To
2016, 2017; Dor et al. 2009, 2011; Koltai et al. 2010; do so, we compared RNA-Seq data from O. ramosa-
Torres-Vera et al. 2016; Umehara et al. 2008; Bardaro resistant and –susceptible varieties enabling the iden-
et al. 2016, Pavan et al. 2016; for review see Brun et al. tification of differentially expressed sequences, and
2017; Samejima and Sugimoto 2018) uncovering mapped the areas of interest on Nicotiana tabacum
evidence linking this property to modifications of root reference genome. Identified sequences were con-
exudate SL content: some accessions induced lower firmed by linkage mapping on an F2 segregating
parasite seed germination or tubercle necrosis. How- population. We confirmed the recessive nature of
ever, apart from these cases, it remains unclear ‘Wika’ resistance, and offer molecular and biological
whether lower parasite infection rates observed with evidence in favour of a deletion mutation spanning a
other resistant or tolerant species (such as sunflower, large area of NT14 underlying this trait. Further
cowpea, sorghum or maize) are indeed linked to analyses to identify potential resistant varieties among
modified SL production by host roots, or to the release Imperial Brands collection were carried out using
of inhibitory molecules as defence mechanisms. Given biological and molecular screening approaches. Fol-
the important roles of SL and rhizosphere interactions lowing extensive analysis, we developed PCR/KASP
for plant growth, crop fitness may be compromised by (Kompetitive allele specific PCR) markers which we
resistance and tolerance mechanisms involving SL have been made available to directly detect and
content reductions in root exudates, and must also be implement our findings for breeding purposes.
considered during breeding process. Indeed, tolerance
mechanisms are generally complex, multifactorial
and/or polygenic phenomena; with low heritability Materials and methods
and are thus unstable (Rubiales 2014; Rispail et al.
2007). Plant material
Indeed, naturally occurring resistant and tolerant
varieties in crop species have received increasing Plant material is listed in Online Resource 1. All
attention and significant efforts to identify genetic and breeding lines (ORO2 to ORO11 KYR, V4K,
molecular bases for Orobanchaceae resistance and ITB31612 and BY02) and F2 population come from
tolerance have been made recently (Yang et al. 2017; Bergerac Seeds and Breeding Company (BSB, http://
Letousey et al. 2007; Dı́az-Ruiz et al. 2010; Darvish- www.bergeracsb.com/). All the other tobacco acces-
zadeh 2016; Ocaña-Moral et al. 2017). With regards to sions (named IT ? number) are part of the Imperial
tobacco resistance or tolerance to root parasitism, two Brands germplasm collection as found on the cata-
distinct Orobanche spp.-tolerant varieties of Tobacco, logue available on http://www.imperialbrandsscience.
BC60 FC (Virginia tobacco) (Covarelli 2002) and com/en/our-science/plant-science.html.
‘Wika’ (Imperial Tobacco Collection) (Cailleteau For RNA-Seq experiment, twelve breeding lines
et al. 2006; Brault-Hernandez 2006), have been were used for RNA-Seq analysis, including seven
described and have raised hope of breeding tolerance tolerant (ORO2;ORO3;ORO4; ORO5; ORO6; ORO7;
to the parasite. It appears that the tolerance mechanism ORO8) and five susceptible (ORO9; ORO10; ORO11;
involves absence of parasite seed germination induc- V4K; KYR), as well as one susceptible control VD and
tion correlated with absence of active molecules in ‘Wika’ as the resistant reference. ORO2, ORO3,
root exudates, as compared to sensitive varieties of ORO3, ORO6, ORO10 and ORO11 are lines devel-
Virginia tobacco, however no genetic or molecular oped from an initial cross between V4K and ‘Wika’.

123
6 Page 4 of 18 Euphytica (2020)216:6

ORO5, ORO7, ORO8 and ORO9 are lines developed or accession, 4 plantlets were scored and the mean was
from an initial cross between KYR and ‘Wika’. calculated. Plants with a percentage of germination
Two F2 populations were used for linkage map- below or equal to 10% and with a number of fixation
ping: the first F2 population resulting from the cross below or equal to 1 were considered as tolerant, others
between ‘Wika’ and V4K (susceptible) and a second were considered to be susceptible.
F2 population resulting from the cross between ‘Wika’
and KYR (susceptible). RNA-Seq analysis
For re-sequencing experiment, ‘Wika’ (tolerant)
and BY02 (susceptible) varieties were used. The cDNA libraries were prepared by mixing equal
For validation on modern breeding lines, Bergerac quantities of total RNAs extracted using the Qiagen
Seeds and Breeding varieties were used: susceptible (F RNeasy Kit from leaves and roots from a pool (6
2 fertile line, MS 4, MS 5 MS 6 Male sterile lines and plants) of 8-weeks-old plantlets for tobacco lines (8
BSB 6199 and ITB188 commercial varieties) or tolerant, 7 susceptible and the F1).
resistant (F 10 and F 11 fertile lines, MS 12, MS 13, The concentration of all RNA samples was checked
MS 14, MS 15 male sterile lines, HF1-17, HF1-18, using a NanoDrop ND-1000 Spectrophotometer and
HF1-19 hybrids and BSB 6190 and BSB 6191 RNA quality was assessed on an Agilent RNA 6000
commercial hybrids). BSB 6199 hybrid has only one Nano LabChip (Agilent Technology 2100 Bioana-
tolerant parent, and BSB 6190 and BSB 6191 have lyzer). Samples with an RNA Integrity Number (RIN)
both parents tolerant. Both are susceptible for ITB188 value greater than 7.5 were used to prepare mRNA
and ITB 31612. libraries according to the TruSeq RNA Sample Prep
Kit v2 (Illumina Inc., San Diego, CA, USA) protocol
Orobanche tolerance testing and the quality of libraries was assessed using the
Agilent Bioanalyzer. Libraries were sequenced on an
This test is based on the ability of Orobanche ramosa Illumina Hi-Seq 3000 (2 9 150 nt) with barcodes.
seeds to germinate in presence of tobacco roots.
Orobanche spp. seeds from O. ramosa species were Re-sequencing analysis of ‘Wika’ and BY02
collected during field trials in France. Petri dishes
were filled with wet mineral wool covered with a ‘Wika’ and BY02 DNA was extracted independently
Whatmann fibreglass paper sheet, and pierced with a from 8 plantlets using Qiagen Dneasy Kit according to
small hole on one side of the edge. O. ramosa seeds manufacturer instructions, and checked for concen-
(500-1000) were sprayed on the paper sheet. Four tration. DNA were mixed together for each variety.
week old tobacco plantlets were placed in the upper Libraries were prepared with Illumina TruSeq PCR
half of the plates. Plantlet roots were cleaned, placed free Kit and sequenced on a HiSeq 3000, one lane per
on the fibre paper and the leaves were placed so as to DNA (paired-end 2x150 pb).
protrude from the hole. Petri dishes were closed with
the lid which was maintained by an elastic band. Petri Genetic mapping of candidate genes
dishes were suspended vertically in boxes containing
2 cm of water. Boxes were placed in a greenhouse All the primers are available in Online Resource 3.
with natural day/night photoperiod and constant
temperature of 22 C. Interpretation was done after Conversion of candidate sequences into PCR
4 weeks by scoring the % of O. ramosa seed markers
germination in the vicinity of the roots and also the
number of germinated seeds that were attached to the Candidate sequences from differential expression
roots (named star-stage tubercle): tubercle develops a analysis were subjected to various BLAST searches
crown of adventitious roots shaped as a star. An against Nicotiana tabacum sequences (nr, gss,
illustration is available in Online Resource 2. An ESTs…) to define copy numbers of genes and to help
additional scoring was performed one week later to in the design of specific primers. PCR amplifications
confirm the phenotype when percentage of germina- were carried out with the HotStart polymerase (Euro-
tion was below 10% at the first scoring. For fixed lines bio) in a 10 ll volume containing 0.5 ll DNA, 0.4 lM

123
Euphytica (2020)216:6 Page 5 of 18 6

of each primer, 2.5 ll of HotStart Taq mix and water Amplification of KASPTM markers
qsp 10 ll. PCR was conducted using a SimpliAmp
thermal cycler from Applied Biosystems. as follows, Sequences used for KASPTM design are available in
with melting temperature adjustment according to Online Resource 3.
primers used: 95 C 2 min, 35 of 95 C 15 s, Two KASPTM markers were designed according to
60–64 C for 15 s, 72 C for 1 min. Amplification the Tobacco Infinium 30 K-2015 genetic map avail-
products were analysed by non-denaturing elec- able on Sol Genomics Network (https://solgenomics.
trophoresis on 1.5% agarose gels. Primers were first net/cview/map.pl?map_version_id=178). These
tested on 8 tolerant and 8 susceptible plants. In case of markers (Nt1AE9764; and Nt1AA1967) were chosen
positive scoring (presence in susceptible and absence according to the efficient design of KASP primers, on
in tolerant), 50 -FAM or 50 -VIC labelled primers were the distal end of Nt14 on the genetic map. Three SNPs
used for further analyses on ABI3130xl. already mapped in this region of chromosome 14 in
PCR markers directly designed on Nt14 sequence another study were added for KASPTM design
of the reference genome (Edwards et al. 2017; (T34981; T19942 and T55624). Candidate sequences
available on https://solgenomics.net/organism/ with the SNP position were sent to LGC for KASPTM
Nicotiana_tabacum/genome) were also designed, in design. KASPTM markers were then validated on the
order to have markers distributed between 0 and 14 parent of the genetic map by amplification with KASP
million bases (Marker names beginning with Nt14). genotyping chemistry from LGC, with KASP Master
The same protocol was used. mix and KASP Assay mix containing the target
specific primers, according to manufacturer instruc-
Amplification of SSR markers tion. Markers were amplified on QuantStudio 3 (Ap-
plied Biosystems) and analysed with QuantStudio
Amplification of SSR markers from Bindler et al. Design and Analyses software v1.4.3.
(2011) and Tong et al. (2016) were used. SSRs were
amplified with the Type-it Microsatellite PCR Kit Linkage mapping
from Qiagen in multiplex, according to the manufac-
turer’s instructions. SSR primers were prepared as Markers were amplified on F2 plants from ‘Wika’ x
follows: the sequence-specific forward primer conju- KYR or ‘Wika’ x V4K. Segregation and grouping of
gated with 50 -TGA CCG GCA GCA AAA TTG-30 tail SSR markers, KASPTM markers, Nt14 and candidate
at its 50 end or the sequence-specific reverse primer markers were performed with the JoinMap 3.0
conjugated with 50 -TGT AAA ACG ACG GCC AGT- program (Van Ooijen 2006) with the default param-
30 tail at its 50 end. The PCR contains these primers in eters, using regression mapping and the Kosambi
addition to the FAM labelled 50 -TGA CCG GCA GCA mapping function.
AAA TTG-30 primer or the VIC labelled 50 -TGT AAA
ACG ACG GCC AGT-30 primer. The PCR was Bioinformatic analyses on CLC Genomic
conducted on SimpliAmp thermal cycler with the Workbench
following program, with melting temperature adjust-
ment, according to the multiplex: 95 C for 5 min, 35 RNA-Seq and t test analysis
to 40 cycles of 95 C for 30 s, 58-64 C for 90 s,
72 C for 1 min, followed by 30 min at 60 C. The same reference transcriptome as described in Julio
et al. (2015) was used, which was developed from
Marker analyses tobacco ancestors, with a total of 32,852 contigs for N.
sylvestris and 29,543 contigs for N. tomentosiformis,
SSRs and PCR markers were analysed on ABI3130xl with a median length of respectively 1283 bp and
Genetic Analyzer with POP-7TM Polymer (Applied 1277 bp. To quantify the expression of each transcript
Biosystems) with software GeneMapper v4.0. in samples and identify differentially expressed genes,
the reads from each sample were mapped onto the N.
sylvestris and N. tomentosiformis reference transcrip-
tomes using the CLC Genomics Workbench 12

123
6 Page 6 of 18 Euphytica (2020)216:6

software as described in Julio et al. (2015). Differen- A mean of 67% of reads were successfully mapped
tial expression analyses were performed with a t-test on the reference transcriptome, with 51.5% of reads
with default parameters, assuming groups were homo- mapped in pairs. Differential expression was then
geneous in terms of variance. examined with the t-test suited to a two-group
experiment procedure, with a normalization step.
Mapping of candidate markers onto the Nicotiana Contigs were sorted by P-value of tolerant vs suscep-
tabacum reference genome tible normalized values. As a first step of the analysis,
it was decided to only show contigs with a P-value less
The N. tabacum genome was obtained on Sol Genomics than or equal to 10-6. This represented 89 contigs.
Network (SGN; ftp://ftp.solgenomics.net/genomes/ These results are available in Online Resource 4. Four
Nicotiana_tabacum/edwards_et_al_2017/assembly/) and of the 89 were preferentially expressed in tolerant
used as a reference. Candidate genes were mapped on the plants, and more than 95% were expressed in suscep-
N. tabacum genome with the Blast function of CLC tible plants, as expected for a recessive origin of the
genomic Workbench. Alternatively, the Blast function tolerance. Only 5 contigs were of a N. sylvestris origin,
directly available on Sol Genomics Network was used showing that the origin of the tolerance is probably
(https://solgenomics.net/organism/Nicotiana_tabacum/ related to a N. tomentosiformis chromosome. The 89
genome). corresponding sequences were mapped by Blast
Search on the N. tabacum genome available on Sol
Mapping of ‘Wika’ and BY02 re-sequencing reads Genomics Network. The reference genome is K326, a
onto the Nicotiana tabacum reference genome flue-cured variety susceptible to Orobanche. 87.6% of
the contigs (78 of the 89) were mapped on the
Re-sequencing data were aligned to the N. tabacum chromosome 14 (Nt14). By selecting the best match
reference genome. CLC Genomic Workbench was for each contig on Nt14, we defined an area of 14 Mb
used with the function ‘‘Map Read to Contigs’’. on the distal end of Nt14 where all these contigs
Parameters used were as default, with Length frac- seemed to be located (Online Resource 4). This
tion = 1 and Similarity fraction = 0.98. Tracks func- represents approximately 220 genes according N
tion was used to compare the reads coverage of ‘Wika’ tabacum gene models annotated on the reference
and BY02 against the reference. genome available on Sol Genomics Network JBrowse.
These first results show that most of the differen-
tially expressed contigs are absent in tolerant plants,
Results and preferentially map on the chromosome 14 of the
reference genome, the origin of which is N. tomen-
RNA-Seq analysis of tolerant vs susceptible tosiformis (Sierro et al. 2014; Edwards et al. 2017).
tobacco varieties identify differential expression
patterns ramosa-susceptibility phenotype is recessive,
and tolerant plants lack genomic sequences located
Phenotype of varieties was checked with a biological on Nt14
test, confirming a percentage of germination less than
10% for the eight resistant lines and greater than 50% A first population resulting from the cross between
for the six susceptible lines. RNAs from O. ramosa- ‘Wika’ and V4K was generated, and biological
resistant lines (ORO2 to ORO8 and ‘Wika’) and O. broomrape resistance was tested on 105 F2 plants.
ramosa-susceptible lines (ORO9 to ORO11 and KYR, Seven plants of each parent variety (‘Wika’ and V4K)
V4K and VD) were sequenced on two Illumina lanes, were used as positive and negative controls. Of the F1
with multiplex barcoding. After quality control, generation, all plants were classified as susceptible
adapter trimming and filtering, a mean of 51,586,108 according to assessment criteria, thus confirming the
and 45,957,777 paired reads was obtained for O. recessive nature of ‘Wika’ tolerance mechanisms. Of
ramosa-resistant lines and of O. ramosa-susceptible the 105 F2 plantlets, the biological test identified 23
lines, respectively. (21.9%) of them as tolerant, and 82 (78.1%) as

123
Euphytica (2020)216:6 Page 7 of 18 6

susceptible, roughly fitting the 1:3 ratio expected for a 2016) was performed. Five KASP markers were also
recessive gene inheritance pattern (P value = 0.464). amplified; two coming from the Tobacco Infinium
DNA was extracted on 83 of the F2 plants, of which 30 K-2015 genetic map (NT1AE9764 and
20 were classified as tolerant and 63 as susceptible. Nt1AA1967) and three previously mapped in our
Primers were designed to amplify several candidate laboratory in another study (T34981; T19942 and
contigs obtained by RNA-Seq: contigs T002229, T55624) on the distal end of Nt14. Markers were
T018112, T003744, T005035 and T003437 were chosen so as to be distributed across a large portion of
tested for amplification in these plants. Contigs were Nt14. Polymorphism obtained on parents is given in
chosen according to expression differences observed Online Resource 5. For each F2 individual, markers
in the RNA-Seq experiment, and the ease with which were noted as either dominant or codominant. The
PCR primers could be designed. Their theoretical same results was obtained with DNA markers from
position (by BLAST on Nt14 of the reference genome) Nt14 and markers derived from contigs, with an
is between 695,000 bp and 8,000,000 bp (Online absence of amplification in tolerant lines. In order to
Resource 4). DNA corresponding to the five selected simplify the figure, only one contig marker (T003444)
contig sequences could not be amplified in 19 tolerant and one DNA marker (Nt14-2721371) were selected
plants out of the 20 that were tested. The tolerant plant for the mapping, with a dominant scoring. Nine SSR
for which an amplification was obtained was observed were scored as codominant (PT53142, TM57910,
to have a small root system, which is not favourable to PT61499, PT53687, TM54360, PT54242, TM66147,
O. ramosa seed germination. They were amplified in TM59844 and TM59852) and three were scored as
all the 63 susceptible plants except one. dominant markers (PT52334, TM60319 and
These experiments confirm the strong correlation TM60323). Surprisingly, some KASP markers were
(97.6%) between tolerance to broomrape, i.e. plants expected not to be amplified in this area, however an
inducing very low germination rates of O. ramosa amplification was observed for T55624, T19942,
seeds, and the absence of genomic sequences corre- T34981, for which three groups of F2 plants were
sponding to loci situated on chromosome 14. observed. NT1AE9764 an NT1AA1967 were scored
A second F2 population resulting from the cross as dominant, with two groups observed on F2 plants.
between ‘Wika’ and KYR was used for linkage Locus genotype frequency was as expected with a
mapping analysis. 144 F2 individual plants were ratio of 1:3 for dominant markers (‘Wika’ allele:KYR
tested for O. ramosa tolerance. Ten plants displaying allele), or 1:2:1 for codominant markers except for
minimal development or with a small root system were three markers; a strong segregation distortion was
discarded to reduce the impact of false negatives. observed for T34981 (35:85:7, X2 = 26.9) with only 7
Scoring was then performed on 134 plants, with parent plants recorded with KYR allele. It was removed from
plants as controls. As for ‘Wika’ x V4K1 F2 plantlets, mapping. However, F2 scored with the ‘Wika’ allele
34 (25.4%) of ‘Wika’ x KYR F2 individuals were were resistant, and F2 scored with KYR allele or
classified as resistant, and 100 (74.5%) were suscep- heterozygous were susceptible. TM59852 and
tible according to scoring criteria (Fig. 1); a 1:3 ratio TM59844 were slightly distorted with respectively
indicating recessive inheritance was again observed (P X2 = 6.05 (25:79:26) and X2 = 5.1 (28:80:26). The
value = 0.336). linkage mapping of the tolerance locus (BIO) show
that differentially expressed genes are genetically
Differentially expressed contigs are located linked to SSR markers previously located on Nt14, as
on Nt14, a large region of which is missing illustrated in Fig. 2. As for the ‘Wika’ x V4K F2 cross,
in tolerant plants the tolerance phenotype (low germination of parasite
seeds and attachment of tubercles) was strongly
Following DNA extraction of the 134 ‘Wika’ x KYR correlated to the absence of genetic sequences located
F2 plantlets (34 Tolerant and 100 Susceptible), PCR on Nt14 (95%) thus confirming that a majority of
amplification of 18 DNA markers (7 contigs from differentially expressed contigs are located on Nt14,
RNA-Seq and 11 DNA markers located on Nt14) and and that the missing sequences in O. ramosa tolerant
13 microsatellite markers (SSR) located between 0 plants account for the observed phenotype. The
and 38 cM on Nt14 (Bindler et al. 2011 and Tong et al. missing chromosome fragment is situated on the distal

123
6 Page 8 of 18 Euphytica (2020)216:6

100 35
90
30
80
70 25

60
20
50
15
40
30 10
20
5
10
0 0

1 2 3 4 5
Fig. 1 Results of the biological testing of the F2 population number of germinated seeds fixed on tobacco roots (right scale).
Wika x KYR. Black bars represent the percentage of Orobanche 1: 134 F2 Wika x KYR plants; 2: KYR parent (4 plants); 3: Wika
seeds germinated (left scale) and the grey bars represent the (4 plants); 4:F1 KYR x Wika; 5: susceptible controls (4 plants)

end of Nt14, and spans roughly 10 million bp, based on showed amplification bands for all investigated PCR
the markers we selected. primers; whereas tolerant varieties displayed disparate
amplification profiles. Indeed, as shown in Online
A vast majority of O. ramosa-tolerant tobacco Resource 6, IT 1036, IT 167, IT 450, IT 453, IT 367
varieties from Imperial Brands collection lack and IT 377 varieties have amplification profiles which
portions of Nt14 are identical to ‘Wika’ (except for SSR marker
PT53142), and confirm our previous RNA-Seq data.
Following the RNA-Seq experiments showing that IT 234, IT 998, IT 175, IT163, IT 241 and IT 1191
recessive O. ramosa tolerance is due to a large deletion varieties showed amplification bands for 17 or more
of Nt14, we sought to determine whether other non-consecutive markers. IT 447 variety, despite
sensitive and tolerant varieties from collection retaining the O. ramosa tolerance phenotype, dis-
libraries carry the same genomic hallmarks, namely played amplification bands for all the markers on
that tolerant varieties have missing sections of their Nt14. Thus resistant phenotypes do not reflect anal-
genome on Nt14 as compared to sensitive ones. In a ogous genotypes, instead a spectrum of genomic
first screening experiment, we used 41 markers deletions exists among these resistant varieties. The
selected among the contig sequences we identified, data reveals that certain deletion profiles are discon-
DNA markers (Edwards et al. 2017) and SSR markers tinuous; some varieties show amplification bands over
(Bindler et al. 2011; Tong et al. 2016) all located along large spans of Nt14 followed by missing areas for
Nt14 and spanning over 14,000,000 bp. Markers were which no amplification signals were found for markers
chosen so as to cover Nt14 relatively evenly over the located therein. Surprisingly, again for KASP markers
whole section identified as missing in tolerant plants. a polymorphism was obtained, with two groups
Five KASP markers used for the genetic mapping on corresponding to ‘Wika’ profile or susceptible profile,
the F2 were used too. We conducted PCR amplifica- although tolerant plants should not be amplified.
tion on DNA material from 19 varieties; 5 susceptible A second screening experiment that included 818
and 13 tolerant cultivars (see Online Resource 6) varieties from Imperial Brands collection was per-
preliminary phenotyped for tolerance to broomrape. formed, using two markers located on Nt14:Nt14-
We found that, as expected, susceptible varieties 2721371 and SSR TM60323. 108 varieties for which

123
Euphytica (2020)216:6 Page 9 of 18 6

mechanisms may be involved despite the deletion of

identical areas of Nt14.
To further examine the phenotypic and genomic
correlations, allelism testing were performed to either
confirm or exclude whether individual resistance
phenotypes involve reduced germination of parasite
seeds (i.e. whether the tolerance mechanism is iden-
tical to that of ‘Wika’). Candidate varieties were
crossed with ‘Wika’, and O. ramosa resistance
phenotype of the F1 plantlets was examined using
the biological test previously described. Biological
test result is given in the last column of Table 1 and
detailed in Table 2. Among tolerant varieties with
identical Nt14 deletion patterns to those of ‘Wika’,
one would expect that the recessive O. ramosa-
tolerant phenotype would be maintained in F1
plantlets after crossing with ‘Wika’; which is the case
for IT 377, IT 175, IT 163, IT 1191, IT 1048, IT 975,
IT 1064, IT 864 and IT 156. Results are less clear for
IT 65, IT 60 and IT 965, where one plantlet show a less
tolerant phenotype with up to 10% of germination or
10 attached Orobanche. These observations are
strongly in favour of the initial hypothesis that
genomic deletions on Nt14 observed in ‘Wika’ are
responsible for the broomrape-tolerant recessive phe-
Fig. 2 Linkage mapping of the tolerance loci on Nt14 obtained notype characterized by parasite seed germination
by analyzing 134 F2 plants from the cross Wika x KYR.
Distance are given in centimorgans (cM) on the left. BIO is the inhibition and radicle fixation inhibition. Among the
biological testing recorded as a dominant marker tolerant varieties we tested, some had different Nt14
deletion patterns but either retained the recessive O.
Nt14-2721371 was absent (no amplification), and 102 ramosa resistance phenotype in the F1 plantlet after
varieties that had no amplification signal for TM60323 crossing with ‘Wika’ (IT 163) or gave rise to
were identified. Among these both markers could not susceptible F1 plantlets (IT 241). On the one hand,
be amplified in 104 varieties. These plants were this seems to indicate that the Nt14 deletion patterns,
therefore expected to present an O.ramosa-tolerant although different, included genomic regions which
phenotype in the biological resistance test. Eight determine the resistance phenotype; and on the other
candidates were selected among the potentially resis- hand, that in some cases, parent resistance phenotypes
tant varieties, together with 5 others varieties and rely on mechanisms which are different from those of
screened for 26 markers distributed on Nt14 with ‘Wika’. Also, a F1 resulting from the cross between IT
‘Wika’ as control. Table 1 presents the results 447 and BY02 (susceptible) gave rise to susceptible
obtained on these 13 new varieties and compared to plants, confirming the hypothesis of a recessive
previous. The biological test results showed that the resistance in IT 447, probably allelic to Wika’s
absence of the selected Nt14 markers was correlated resistance. The results described here warrant further
with O. ramosa-tolerant phenotypes according to our experiments and underscore the complexity of O.
biological testing protocol. Not all the candidate ramosa-tolerance phenotypes that have not been
varieties had the same deletion profiles as ‘Wika’, completely unravelled.
but retained O. ramosa resistance, confirming the
genomic heterogeneity that was previously suggested
and implying that several different resistance

123
 6
Table 1 results of amplification of markers from different type on susceptible (phenotype S) and tolerant (phenotype T) varieties
Variety Mean % Mean Phenotype NT14- T003447 NT14- PT53078 NT- PT61529 T002229 TM57910 PT60462 PT60586 Nt14- T009443
germination star 196093 2721371 3398235 4910540

123
level
Page 10 of 18

KYR 45 1 S 1 1 1 212 1 240 1 232 186 192 1 1

V4K 75 6.4 S 1 1 1 214 1 242 1 232 186 192 1 1
WIKA 1.25 0 T 0 0 0 0 0 0 0 236 184 192 0 0
IT 377 0 0 T 0 0 0 0 0 0 0 242 184 192 0 0
IT 175 0.75 0.25 T 0 0 0 0 0 240 1 242 184 192 0 0
IT 998 0 0 T 0 0 0 0 0 240 1 242 184 192 0 0
IT 1191 0 0 T 0 0 0 0 0 240 1 242 186 192 0 0
IT 241 0 0 T 0 0 0 0 0 240 1 238 184 192 1 1
IT 447 0 0.37 T 1 1 1 210 1 240 1 238 184 192 1 1
IT 156 6.25 0.5 T 0 0 0 0 0 0 0 242 184 192 0 0
IT 1048 0 0 T 0 0 0 0 0 0 0 242 184 192 0 0
IT 60 3.38 2.8 T 0 0 0 0 0 0 0 242 184 192 0 0
IT 65 0 0 T 0 0 0 0 0 0 0 242 184 192 0 0
IT 975 0.375 0 T 0 0 0 0 0 0 0 242 184 192 0 0
IT 965 5.25 1.25 T 0 0 0 0 0 0 0 242 184 192 0 0
IT 1064 0.375 0.125 T 0 0 0 0 0 0 0 230 186 192 1 1
IT 864 0.75 0.125 T 1 1 1 208 1 0 0 242 184 192 0 0
IT 1123 60 22 S 1 1 1 214 1 244 1 230 186 192 1 1
IT 897 43.75 18.63 S 1 1 1 208 1 240 1 230 186 192 1 1
IT 910 21.88 3.75 S 0 0 1 208 1 240 1 242 184 192 1 1
IT 86 80 22.1 S 1 1 1 214 1 244 1 230 186 192 1 1
ITB 31612 72.5 22.5 S 1 1 1 214 1 244 1 230 186 192 1 1
Theoritical 0.2 0.7 2.7 3.1 3.3 3.8 3.8 4 4.1 4.3 4.9 5
position
on Nt14
(Mb)
Euphytica (2020)216:6

Variety PT52334 Nt14- TM60319 NT14- PT60250 PT60170 NT14- T010110 T010253 PT53142 CCD4 PT53687 PT61107 PT54896 Bio.
5411837 5632510 7938801 testing
of F1
with
Wika

KYR 195 1 276 1 159 151 1 1 1 210 1 226 222 243 S

 Table 1 continued
Variety PT52334 Nt14- TM60319 NT14- PT60250 PT60170 NT14- T010110 T010253 PT53142 CCD4 PT53687 PT61107 PT54896 Bio.
5411837 5632510 7938801 testing
of F1
with
Wika
Euphytica (2020)216:6

V4K 195 1 288 1 159 149 1 1 1 210 1 256 218 241 S

WIKA 0 0 0 0 0 0 0 0 0 208 1 254 218 241 –
IT 377 0 0 0 0 0 0 0 0 0 208 1 254 218 241 T
IT 175 0 0 0 1 159 153 1 0 0 208 1 254 218 241 T
IT 998 0 0 0 1 159 153 1 0 0 208 1 254 218 241 T
IT 1191 0 0 0 1 159 153 1 0 0 212 1 254 224 241 T
IT 241 nd 1 290 1 159 155 1 0 0 208 1 256 218 241 S
IT 447 195 1 290 1 159 155 1 1 1 208 1 256 218-222 241 T
IT 156 0 0 0 0 0 0 0 0 0 208 1 254 218 241 T
IT 1048 0 0 0 0 0 0 0 0 0 210 1 224 220 241 T
IT 60 0 0 0 0 0 0 0 0 0 210 1 254 218 241 T-
IT 65 0 0 0 0 0 0 0 0 0 210 1 258 218 241 T-
IT 975 0 0 0 0 0 0 0 0 0 210 1 254 218 241 T
IT 965 0 0 0 0 0 0 0 0 0 210 1 256 218 245 T-
IT 1064 0 0 0 0 0 0 0 0 0 212 1 224 222 243 T-
IT 864 0 0 0 0 0 0 0 1 1 210 1 224 220 243 T
IT 1123 195 1 286 1 159 149 1 1 1 212 1 222 220 243 nd
IT 897 195 1 284 1 159 149 1 1 1 214 1 222 218 243 nd
IT 910 197 1 286 1 159 149 1 1 1 220 1 222 220 243 nd
IT 86 195 1 286 1 159 149 1 1 1 212 1 224 220 241 nd
ITB 31612 195 1 286 1 159 149 1 1 1 212 1 224 220 241 nd
Theoritical 5.1 5.4 5.5 5.6 6.7 7.4 7.9 12.5 14 14.8 15.3 16.7 16.7 19.4
position
Page 11 of 18

on Nt14
6

(Mb)
Results of broomrape tolerance testing is given as a mean of 4 plants for the percentage of broomrape seeds germination and the number of germinated seeds fixed on roots.
Theoritical position was directly determined on chromosome 14 for Nt14 markers or by blast alignment of primers (SSR) or sequences (contigs) on tobacco reference genome on
Solgenomics. Only the best alignment result was reported when several matching were retrieved. Allelism tests was performed by crossing Wika with candidate varieties, and F1
plants were tested for Orobanche tolerance. Results is given in the last column. (T Tolerant; T-Tolerant with one plantlet recorded with a % of germination [ 10% of more than 10
seeds fixed; S susceptible; Nd not determined)

123
6 Page 12 of 18 Euphytica (2020)216:6

Table 2 Detailed results of F1 cross % of germination Number of germinated seed fixed

Orobanche tolerance tests
on F1 crosses P1 P2 P3 P4 P1 P2 P3 P4

F1 Wika x IT 377 0 0 0 0 0 0 0 0
F1 Wika x IT 175 0 0 0 0 0 0 0 0
F1 Wika x IT 163 0 0 0 0 0 0 0 0
F1 Wika x IT 1191 0 0 0 0 0 0 0 0
F1 Wika x IT 241 5 5 10 5 13 22 25 13
F1 Wika x IT 447 0 0 0 0 0 0 0 0
F1 Wika x IT 156 0 0 0 0 0 0 0 0
F1 Wika x IT 1048 1 0 0 1 1 1 0 3
F1 Wika x IT 60 0 0 0 5 2 1 0 12
F1 Wika x IT 65 30 1 1 3 1 0 1 1
F1 Wika x IT 975 1 0 0 0 3 0 0 0
F1 Wika x IT 965 0 0 0 10 0 0 0 13
F1 Wika x IT 1064 5 1 0 1 8 2 1 1
F1 Wika x IT 864 1 0 1 1 1 0 5 3
Four plants were tested for
each cross. Wika was used F1 By02 x IT 447 50 50 75 40 15 30 23 21
as tolerant control and Wika 0 0 0 0 0 0 0 0
ITB31612 as susceptible ITB 31612 30 60 10 15 6 18 2 0
control

Re-sequencing of ‘Wika’ and comparison absence of genetic material. Reads mapping between
with BY02, a susceptible tobacco variety 3866 Kb and 4905 Kb in ‘Wika’ are consistent with
amplification of markers in Online Resource 6, where
Genomic re-sequencing of ‘Wika’ and BY02, a markers TM57910, PT60462 and PT60586, which
susceptible variant, was then performed, and reads have a theoretical position between 3900 and 4900 Kb
were mapped on the reference genome of N. tabacum. are amplified in ‘Wika’. However, this is not consis-
An illustration of the results is given in Fig. 3. tent with amplification of KASP markers Nt1AE9764,
Compared to BY02 (susceptible), where a full cover- T34981, Nt1AA1967, T55624 and T19942 that should
age of Nt14 is obtained, no aligned reads were be in the deleted area. Taken together, these findings
observed for ‘Wika’ between 0 and 3866 Kb and underscore the discrepancies between the genetic re-
between 4905 Kb and 14227 Kb, thus confirming the sequencing data and the KASP marker results. Indeed,

Fig. 3 Schematic representation of genomic reads mapping on sequence in the reference genome; vertical dashes: probable
Nt14. The figure shows the results of reads mapping analyses of deleted region in Wika; white arrows: potential duplication of
Wika and BY02 Illumina re-sequencing on the N. tabacum Nt14 genomic sequences. Compared to BY02 (susceptible), where a
reference genome available on SolGenomic Network, between full coverage of Nt14 is obtained, no read alignment is obtained
positions 0 and 16000000 bases. Black area: high density of for Wika between 0 and 3.866 Mb and between 4.905 Mb and
genomic reads mapping against the reference genome; white 14.227 Mb
area: no read mapping; horizontal dashes: undetermined

123
Euphytica (2020)216:6 Page 13 of 18 6

these incongruent results point to insufficient or deletions of large portions of Nt14 and O. ramosa-
erroneous knowledge of NT14 structure. tolerant phenotype. In addition, we show that other
resistant or tolerant varieties display resistance mech-
Validation of markers on BSB modern breeding anisms which are probably distinct from those
lines expressed by ‘Wika’: in IT 241, despite Nt14 deletions
partly similar, recessive inheritance is not conserved
The biological and molecular data are in agreement, when these varieties are crossed with ‘Wika’; whilst
thus allowing the DNA markers to be predictive of the IT 447 display tolerance phenotypes but do not
presence of the Nt14 deletion(s) and broomrape harbour identified genomic deletions. This particular
tolerance. In order to improve breeding and to variety maybe helpful in the future to localize the gene
characterize heterozygous plants, the aim was to responsible for broomrape susceptibility on Nt14.
develop KASP codominant markers. With regards to More importantly, we have tested and validated
results of genetic mapping, the KASP T34981 is a specific PCR markers for dominant genes and codom-
reliable marker for ‘Wika’ resistance breeding, as it is inant KASPTM markers for the development of
strongly correlated with the biological testing. To tolerant tobacco varieties through assisted selection
validate this potential on commercial lines, four methods for breeders using ‘Wika’. And finally, the
markers were chosen for ease of scoring and tested first set of DNA markers for broomrape tolerance was
on 18 modern varieties, male sterile, fertile lines or reported: our work has established several candidate
hybrids belonging to the breeding pool of BSB. This markers (DNA and SSR) which are readily available
experiment also included two susceptible control VD for O. ramosa-tolerance breeding in tobacco.
and ITB31612 and the resistant control ‘Wika’. Three The genomic deletions which we have character-
dominant markers were used (T002229, T002264 and ized in our work were based on the theoretical
T003447) as well as one KASP T34981. Results are positions of DNA and SSR markers aligned on the
given in Online Resource 7. All these markers reference genome of N. tabacum published previously
succeeded in differentiating tolerant and susceptible (Edwards et al. 2017), and our findings regarding the
lines. The KASP marker was also useful to confirm the deletion profile of ‘Wika’ emphasize the complexity
heterozygous genotype of BSB 6199, a hybrid of these particular genomic areas. According to the
between a susceptible line and a tolerant line devel- reference genome, our findings suggest a large,
oped from ‘Wika’. discontinuous deletion of genetic material at one end
of Nt14 in ‘Wika’ and other tobacco varieties from our
collection, as shown by the amplification data using
Discussion PCR and SSR markers, and the re-sequencing data.
The hypothesis of a partial and interrupted deletion, is
In this study, biological and molecular evidence of highly unlikely; and what we observed would rather
extensive chromosome 14 deletions underlying O. indicate an imperfect assembly of the reference Nt14.
ramosa-tolerant phenotype in tobacco varieties was As mentioned above, the tobacco genome is
provided. The recessive nature of ‘Wika’ parasite complex, but also very large and allotetraploid, arising
tolerance was confirmed, based on the inhibition of O. from the combination of two genomes N. sylvestris
ramosa seed germination and radicle attachment as and N. tomentosiformis, and a modest size reduction
demonstrated in biological testing experiments. The with loss of some ancestral sequences (Lim et al. 2004;
development of the biological test we used here Bindler et al. 2011; Leitch et al. 2008; Renny-Byfield
provides a specific tool for tolerant tobacco variety et al. 2011; for review see Strickler et al. 2012).
selection based on the ability to inhibit broomrape Although the coverage of N. tabacum genome was
seed germination and attachment. We also show that expanded recently, current sequencing data does not
for naturally occurring resistant and tolerant varieties, provide consistent and comprehensive genomic loca-
a vast majority have Nt14 deletions identical or greatly tions for gene placement attribution, with only 64% of
similar to ‘Wika’ deletion profile; and that susceptible genes successfully anchored on chromosomes. Indeed,
tobacco varieties do not have Nt14 deletions, thus large spans of the N. tabacum genome remain
establishing a strong correlation between genomic unknown, as shown in Fig. 3. Indeed, gaps between

123
6 Page 14 of 18 Euphytica (2020)216:6

known sequences of the tobacco genome, owing to (same strand, 97%). This finding was represented by a
marker distribution, sequence inversions or mis- probable duplicated zone in Fig. 3. There may be
matched alignments are to be expected, given the several duplications and/or inversions in this area,
slight imprecision with which some chromosome which could explain codominant markers amplifica-
markers have been placed on draft assembly and tion in unexpected areas. These phenomena may
possible recognition of two identical loci due to explain some of the puzzling results we report here,
allotetraploidy mentioned above. In our experiments, namely that for ‘Wika’ and other varieties, long
we observed amplification signals for codominant stretches of deleted genomic material appear to be
KASPTM markers which are, theoretically, located in interrupted by what seem to be long sections of
the deleted area for varieties harbouring large dele- genomic material, evidenced by RNA-Seq reads and
tions on Nt14. Closer scrutiny revealed that the PCR amplification. For example, sequences corre-
sequences used to design these markers match at least sponding to SSR markers TM57910, PT60462 and
on two loci on Nt14 (sense or antisense orientations PT60586, which the theoretical reference map places
were observed). For example, the sequence of marker at respectively 4, 4.2 and 4.3 million bp on Nt14, are
T19942 aligned both at 696,004 bp (reverse strand) followed by deleted sections evidenced by the absence
and 12,964,658 bp (same strand) on Nt14, with 97% of PCR amplification for markers which the reference
and 95% identity respectively. The sequence of map places further on Nt14. These discontinuous
T34981 aligned both at 5,378,807 bp (reverse strand, deletion profiles were consistently shown by both
94%) on Nt14 and on Nt17 (92%). Whereas genomic and transcriptomic sequencing approaches
Nt1AE9764 sequence aligned only at 1,477,831 for several varieties of tobacco, and therefore highlight
(reverse strand, 99%) on Nt14, Nt1AA1967 sequence the discrepancies between theoretical marker mapping
aligned at 6,932,760 bp (same strand, 98%) and and molecular evidence. Also, it is impossible, despite
6,752,758 bp (reverse strand, 97%) on Nt14. Finally, the sequencing coverage of Nt14 we offer here, to
the sequence of T55624 aligned at 1,406,048 bp (same pinpoint the exact location of the chromosome break
strand, 99%) and 12,598,081 bp (reverse strand, 98%) or deletion for ‘Wika’. Finally, there are several
on Nt14. This indicates putative duplications in this portions of Nt14 for which mapping data is missing
area, corroborating the synteny between pseudo- with no determined genomic sequences, and the
molecules on the distal end of Nt14 evidenced in ‘Wika’ variety deletion of roughly 14 million bp of
Edwards et al. (2017). To support this finding, we Nt14 may contain up to 200 genes, and determining
performed an amplification of some sequences tar- which one is directly responsible for the O. ramosa-
geted by KASP design, with a classical AmpliTaq resistance phenotype warrants further studies.
DNA polymerase and KASP PCR program on a According to the literature, potential candidate
SimpliAmp thermalcycler. The three primers (‘Wika’ genes to explain ‘Wika’ tolerance to broomrape
allele and BY02 allele forward primers, and reverse included those involved in SL synthesis. However,
primer) were used to amplify DNA of ‘Wika’ and among possible gene candidates, a carotenoid cleav-
BY02, and sequences were cloned and sequenced age dioxygenase ccd4, situated at 15Mbp (Id: Nitab
(data not shown). For Nt1AE1964 and Nt1AA1967, 4.5_0005280g0010.1; 15,273,177–15,276,966 bp) on
we could find sequences corresponding to both alleles Nt14 seems to be situated just after putative chromo-
in ‘Wika’, and only one allele in BY02, supporting the some breaks in varieties lacking genetic material in
hypothesis of a duplicated region with slight differ- our study. This gene, may be involved in carotenoid
ences that allow KASP markers amplification. cleavage dioxygenase pathway leading to SL synthesis
Moreover, we looked in details the alignment of (like ccd7 and ccd8 in tomato SL synthesis, in Vogel
contigs differentially expressed obtained by RNA- et al. 2010) and would be potential target proteins
Seq. The same findings than for KASP markers was involved in broomrape-tolerance. However this locus
observed, with frequent alignment on two position of is not affected by the deletion present in ‘Wika’.
Nt14. For example, contig T002264 aligned at Indeed, both susceptible and tolerant tobacco varieties
357,000 bp (reverse strand, 99%) and 13,441,727 bp have reads which seem to indicate presence of ccd4,
(reverse strand, 96%). Contig T003447 aligned at and no SNP has been detected in this area by re-
695,027 bp (reverse strand, 100%) and 12,964,947 bp sequencing experiment. Other candidate genes include

123
Euphytica (2020)216:6 Page 15 of 18 6

cytochrome P450 (CYP450) that the reference genome Despite these shortcomings, throughout our
places within the deleted area in tolerant varieties sequencing and biological experiments, we provide
(Booker et al. 2005). The involvement of these robust evidence for the validation of selected DNA
candidate genes could be investigated using EMS and SSR markers for characterization of Nt14 dele-
mutants as described in Julio et al. (2015). tions; and consistent data in favour of a strong
Naturally occurring parasite resistance is, more correlation between the absence of Nt14 deletions
often than not, a polygenic trait (Darvishzadeh 2016; and O. ramosa susceptibility phenotype. Overall, our
Rubiales 2003; Pérez-de-Luque et al. 2005a, b) given findings point to the potential use of the PCR and
the adaptive capacity of broomrape (Rubiales KASP markers located roughly on the distal end of
2014, 2018; Rispail et al. 2007). However the Nt14 for screening of tobacco varieties in view of
inheritance patterns we uncovered are at odds with determining broomrape susceptibility or tolerance.
this statement: our data indicate a monogenic resis- In the biological test we used to characterize
tance mechanism. There is little scientific literature broomrape-susceptibility, the medium (mineral wool
dedicated to ‘Wika’ tobacco variety. Indeed, ‘Wika’ and water) is presumed to model soil parameters for
O. ramosa-tolerance has been characterized biologi- diffusion phenomena which are paramount to the
cally (Brault-Hernandez 2006; Cailleteau et al. 2006), establishment of rhizosphere molecular interactions
showing that this phenotype is linked to modified root and allow the diffusion of signals that O. ramosa seeds
exudates, but the exact chemical and molecular pick up to initiate germination (Westwood 2000). The
pathways are as yet undetermined. The recessive biological assessment of O. ramosa tolerance pheno-
nature of the phenotypic trait was ascertained in type we used here was adapted from tests developed
accordance with previous reports (Brault-Hernandez for sunflower susceptibility or resistance characteris-
2006; Cailleteau et al. 2006), and appears to be tics (Eizenberg et al. 2003; Labrousse et al. 2001), and
monogenic, given the 1:3 heritability ratio we enables the discrimination between parasite germina-
observed. For susceptible varieties with no Nt14 tion and attachment. The exact characteristics of the
sequence deletions such as V4K and KYR, the F2 natural phenomenon have not been quantified or
plantlets (cross with ‘Wika’) and the allelism tests on measured in the field, owing to difficulties of SL
susceptible collection varieties thus confirmed previ- detection, very small size of O. ramosa seeds, plantlet
ous findings. Some tolerant varieties, despite harbour- variabilities and the influence of environmental fac-
ing Nt14 deletions which appear to be different or not tors; therefore the biological correlates our O. ramosa
as extensive as those observed in ‘Wika’ variety, show resistance phenotyping test is based on are not fully
allelism results indicating that their resistance mech- defined, but allow for sound interpretation of the
anisms is identical to that of ‘Wika’. These varieties biological trait that we score in our assay, thus
are particularly interesting for narrowing down poten- providing a first description of a phenotyping tool
tial genomic areas for gene candidates underlying the for tobacco plantlets. The O. ramosa seed density in
O. ramosa-tolerance phenotype in future work (as- the test Petri dishes could be considered as high even
suming Nt14 assembly can be either consolidated or though we did not measure it during the test; however
improved). There are, however, unclear results among it may be considered as an accurate representation of
the findings we presented here. For example, IT 241 the extent of infestation in areas highly affected by the
tobacco variety has Nt14 deletions which are different parasite (Manschadi et al. 2001). In this test, ‘Wika’
from ‘Wika’s, is characterized by a strong tolerant plantlets do not induce any O. ramosa seed germina-
phenotype, but which probably involves an alternative tion, and consistently inhibit radicle attachment,
tolerance mechanism, as the phenotype is lost when IT illustrated by the total absence of star-shaped tubercles
241 is crossed with ‘Wika’. Overall, despite correla- of developing O. ramosa rootlets. Throughout our
tions between the absence of Nt14 sequence deletions biological experiments, susceptible varieties were
and susceptibility to O. ramosa, we must underscore observed, with several dozens of germinating O.
the ambiguity of both the intensity of O. ramosa ramosa seeds and over 10 star-stage tubercles. How-
tolerance and the monogenic model of the phenotype; ever, some varieties had intermediate profiles, illus-
elucidating the genetic causes of ‘Wika’ O. ramosa trating relative biological variability, and these
tolerance calls for further investigations. profiles were consistently observed. In addition,

123
6 Page 16 of 18 Euphytica (2020)216:6

although ‘Wika’ resistance is clear-cut, some tobacco parasitism by Orobanche foetida. Phytopathol Mediterr
varieties which had the same Nt14 deletion profile as 49:239–248
Aly R (2007) Conventional and biotechnological approaches for
‘Wika’, induced slight O. ramosa seed germination control of parasitic weeds. Vitro Cell Dev Biol Plant
(IT 156 and IT 60 varieties for example). This 43:304–318
illustrates the fact that O. ramosa tolerance, as well Bardaro N, Marcotrigiano AR, Bracuto V et al (2016) Genetic
as being influenced by environmental factors, may analysis of resistance to Orobanche crenata (Forsk.) in a
pea (Pisum sativum L.) low-strigolactone line. J Plant
exist over a continuum of resistance or tolerance of Pathol 98:671–675
varying intensity, arising from different mechanisms Barker ER, Press MC, Scholes JD, Quick WP (1996) Interac-
which are in favour of diverse resistance traits, which tions between the parasitic angiosperm Orobanche
are important to maintain as monogenic resistance is aegyptiaca and its tomato host: growth and biomass allo-
cation. New Phytol 133:637–642
more likely to be overcome by the parasite through Bindler G, Plieske J, Bakaher N, Gunduz I, Ivanov N, Van der
broomrape adaptability. Hoeven R, Ganal M, Donini P (2011) A high density
Future directions for our work includes the investi- genetic map of tobacco (Nicotiana tabacum L.) obtained
gation of ‘Wika’ tolerance to other Orobanche vari- from large scale microsatellite marker development. Theor
Appl Genet 123:219–230
eties. Indeed, O. ramosa is the predominant N. tabacum Booker J, Sieberer T, Wright W, Williamson L, Willett B,
parasite in Europe, however the resistance mechanisms Stirnberg P et al (2005) MAX1 encodes a cytochrome P450
involved could open new avenues for developing family member that acts downstream of MAX3/4 to pro-
parasite-resistant crops and thus provide new breeding duce a carotenoid-derived branch-inhibiting hormone. Dev
Cell 8:443–449
tools and scientific insight to manage crop parasitism by Brault M, Betsou F, Jeune B, Tuquet C, Sallé G (2007) Vari-
species in the Orobancheae. Indeed, the mechanisms ability of Orobanche ramosa populations in France as
underlying naturally occurring parasite resistance are of revealed by cross infestations and molecular markers.
particular interest in agriculture, and implementing Environ Exp Bot. https://doi.org/10.1016/j.envexpbot.
2007.06.009
biotechnological approaches to enhance such pheno- Brault-Hernandez (2006) L’orobanche rameuse en France, Le
types represents a particularly attractive means of couple Orobanche ramosa L./Nicotiana tabacum L.:
fighting rampant infection by broomrape in endemic biologie, spécificité et méthodes de lutte. Doctoral Thesis.
regions. The findings reported here emphasize the Univ. Pierre et Marie Curie (Paris VI)
Brun G, Braem L, Thoiron S, Gevaert K, Goormachtig S,
challenges farmers and researchers face for successful Delavault P (2017) Seed germination in parasitic plants:
handling of broomrape parasitism with cultural tech- what insights can we expect from strigolactone research?
niques and scientific endeavours. Firstly, overcoming J Exp Bot. https://doi.org/10.1093/jxb/erx472
crop losses due to broomrape requires in-depth knowl- Buschmann H, Gonsior G, Sauerborn J (2005) Pathogenicity of
branched broomrape (Orobanche ramosa) populations on
edge of the specific modifications that underlie ‘Wika’ tobacco cultivars. Plant Pathol. https://doi.org/10.1111/j.
recessive O. ramosa tolerance and the associated 1365-3059.2005.01211.x
inheritance pattern. With this study, we provide the Cailleteau B, Mornet F, Verrrier JL (2006) A genetically
first molecular evidence of deletion mutations involved determined trait in flue-cured tobacco results in low ger-
mination of Orobanch ramose L. seed. In: CORESTA
in ‘Wika’ tolerance phenotype, and the technical and Congress, Paris, 2006 (Oral presentation)
biological validation of markers that can be used for the Cardoso C, Ruyter-Spira C (2005) Strigolactones and root
improvement of commercial cultivars. And secondly, infestation by plant-parasitic Striga, Orobanche and Phe-
understanding the physiological mechanisms involved lipanche spp. Plant Sci. https://doi.org/10.1016/j.plantsci.
2010.11.007
in O. ramosa resistance in deletion plants will provide Ćavar S, Zwanenburg B, Tarkowski P (2015) Strigolactones:
valuable knowledge for applying the novel findings to occurrence, structure, and biological activity in the rhizo-
other crops that suffer from Orobanche spp. parasitism. sphere. Phytochem Rev 14:691–711
Covarelli L (2002) Studies on the control of broomrape (Oro-
banche ramosa L.) in virginia tobacco (Nicotiana tabacum
L.). Contrib Tob Res. https://doi.org/10.2478/cttr-2013-
0733
References Darvishzadeh R (2016) Genetic variability, structure analysis,
and association mapping of resistance to broomrape
Abbes Z, Kharrat M, Pouvreau JB, Delavault P, Chaibi W, (Orobanche aegyptiaca Pers.) in tobacco. J Agric Sci
Simier P (2010) The dynamics of faba bean (Vicia faba L.) Technol 18:1407–1418
Dı́az-Ruiz R, Torres AM, Satovic Z, Gutierrez MV, Cubero JI,
Román B (2010) Validation of QTLs for Orobanche

123
Euphytica (2020)216:6 Page 17 of 18 6

crenata resistance in faba bean (Vicia faba L.) across Leitch IJ, Hanson L, Lim KY, Kovarik A, Chase MW, Clarkson
environments and generations. Theor Appl Genet JJ et al (2008) The Ups and downs of genome size evolu-
120:909–919 tion in polyploid species of nicotiana (solanaceae). Ann
Dor E, Alperin B, Wininger S, Ben-Dor B, Somvanshi VS, Bot 101:805–814
Koltai H et al (2009) Characterization of a novel tomato Letousey P, De Zélicourt A, Vieira Dos Santos C, Thoiron S,
mutant resistant to the weedy parasites Orobanche and Monteau F, Simier P et al (2007) Molecular analysis of
Phelipanche spp. Euphytica. https://doi.org/10.1007/ resistance mechanisms to Orobanche cumana in sunflower.
s10681-009-0041-2 Plant Pathol. https://doi.org/10.1111/j.1365-3059.2007.
Dor E, Yoneyama K, Wininger S, Kapulnik Y, Yoneyama K, 01575.x
Koltai H et al (2011) Genetics and resistance strigolactone Lim KY, Matyasek M, Kovarik A, Leitch AR (2004) Genome
deficiency confers resistance in tomato line SL-ORT1 to evolution in allotetraploid Nicotiana. Reports of the
the parasitic weeds Phelipanche and Orobanche spp. International Polyploidy Conference, London, UK, 27–30
Phytopathology 101:213–222 April 2003
Edwards KD, Fernandez-Pozo N, Drake-Stowe K, Humphry M, Lins RD, Colquhoun JB, Mallory-Smith CA (2007) Effect of
Evans AD et al (2017) A reference genome for Nicotiana small broomrape (Orobanche minor) on red clover growth
tabacum enables map-based cloning of homeologous loci and dry matter partitioning. Weed Sci 55:517–520
implicated in nitrogen utilization efficiency. BMC Genom. Manschadi AM, Kroschel J, Sauerborn J (1996) Dry matter
https://doi.org/10.1186/s12864-017-3791-6 production and partitioning in the host-parasite-association
Eizenberg H, Hershenhorn D, Eizenberg J, Plakhine Y, Rubin B Vicia faba Orobanche crenata. J Appl Bot 70:224–229
(2003) Resistance to broomrape (Orobanche spp.) in sun- Manschadi AM, Sauerborn J, Stutzel H (2001) Quantitative
flower (Helianthus annuus L.) is temperature dependent. aspects of Orobanche crenata infestation in faba beans as
J Exp Bot 54(395):1305–1311 affected by abiotic factors and parasite soil seedbank.
Fernández-Aparicio M, Westwood JH, Rubiales D (2011) Weed Res 41:311–324
Agronomic, breeding, and biotechnological approaches to Ocaña-Moral S, Gutiérrez N, Torres AM, Madrid E (2017)
parasitic plant management through manipulation of ger- Saturation mapping of regions determining resistance to
mination stimulant levels in agricultural soils. Botany Ascochyta blight and broomrape in faba bean using tran-
89(12):813–826 scriptome-based SNP genotyping. Theor Appl Genet
Fernández-Aparicio M, Kisugi T, Xie X, Rubiales D, Yoneyama 130:2271–2282
K (2014) Low strigolactone root exudation: a novel Parker C (1994) The present state of the Orobanche problem.
mechanism of broomrape (Orobanche and Phelipanche Biology and management of Orobanche. In: Proceedings
spp.) resistance available for faba bean breeding. J Agric of the third international workshop on Orobanche and
Food Chem 62:7063–7071 related Striga research, Amsterdam, Netherlands, 8–12
Fernández-Aparicio M, Reboud X, Gibot-Leclerc S (2016) November 1993, pp 17–26. Royal Tropical Institute,
Broomrape weeds. Underground mechanisms of parasitism Amsterdam, Netherlands
and associated strategies for their control: a review. Front Parker C (2009) Observations on the current status of Oro-
Plant Sci. https://doi.org/10.3389/fpls.2016.00135 banche and striga problems worldwide. Pest Manag Sci
Gibot-Leclerc S, Brault M, Pinochet X, Sallé G (2003) Potential 65:453–459
role of winter rape weeds in the extension of broomrape in Parker C, Riches CR (1993) Parasitic weeds of the world:
Poitou-Charentes. C R Biol 326(7):645–658 biology and control. CAB International, Wallingford, UK
Habimana S, Nduwumuremyi A, Chinama RJD (2014) Man- Pavan S, Schiavulli A, Marcotrigiano AR, Bardaro N, Bracuto V
agement of Orobanche in field crops: a review. J Soil Sci et al (2016) Characterization of low-strigolactone germ-
Plant Nutr. https://doi.org/10.4067/S0718- plasm in Pea (Pisum sativum L.) resistant to Crenate
95162014005000004 Broomrape (Orobanche crenata Forsk.). Mol Plant
Joel DM, Hershenhorn J, Eizenberg H, Aly R, Ejeta G, Rich PJ Microbe Interact 29:743–749
et al (2007) Biology and management of weedy root par- Pérez-de-Luque A, Jorrı́n J, Cubero JI, Rubiales D (2005a)
asites. In: Janick J (ed) Horticultural reviews. Wiley, Orobanche crenata resistance and avoidance in pea (Pisum
Hoboken. https://doi.org/10.1002/9780470168011.ch4 spp.) operate at different developmental stages of the par-
Julio E, Cotucheau J, Volpatti C, Decorps R, Sentenac C, asite. Weed Res 45:379–387
Candresse T, Dorlhac de borne F (2015) A eukaryotic Pérez-de-Luque A, Rubiales D, Cubero JI, Press MC, Scholes J
translation initiation factor 4E (eIF4E) is responsible for et al (2005b) Interaction between Orobanche crenata and
the ‘‘va’’ tobacco recessive resistance to potyviruses. Plant its host legumes: unsuccessful haustorial penetration and
Mol Biol Rep 1:1. https://doi.org/10.1007/s11105-014- necrosis of the developing parasite. Ann Bot 95:935–942
0775-4 Renny-Byfield S, Chester M, Kovarik A, Le Comber SC,
Koltai H, LekKala SP, Bhattacharya C, Mayzlish-Gati E, Res- Grandbastien MA et al (2011) Next generation Sequencing
nick N, Wininger S et al (2010) A tomato strigolactone- reveals genome downsizing in allotetraploid Nicotiana
impaired mutant displays aberrant shoot morphology and tabacum, predominantly through the elimination of pater-
plant interactions. J Exp Bot 61(6):1739–1749 nally derived repetitive DNAs. Mol Biol Evol
Labrousse P, Arnaud MC, Serieys H, Bervillé A, Thalouarn P 28:2843–2854
(2001) Several mechanisms are involved in resistance of Rispail N, Dita MA, González-Verdejo C, Pérez-De-Luque A,
Helianthus to Orobanche cumana Wallr. Ann Bot Castillejo MA, Prats E et al (2007) Plant resistance to
88:859–868 parasitic plants: molecular approaches to an old foe:

123
6 Page 18 of 18 Euphytica (2020)216:6

Research review. New Phytol. https://doi.org/10.1111/j. nitrogen, and potassium deficiencies on strigolactone pro-
1469-8137.2007.01980.x duction. South Afr J Bot 108:15–22
Rubiales D (2003) Parasitic plants, wild relatives and the nature Umehara M, Hanada A, Yoshida S, Akiyama K, Arite T,
of resistance. New Phytol 160:459–461 Takeda-Kamiya N et al (2008) Inhibition of shoot
Rubiales D (2014) Legume breeding for broomrape resistance. branching by new terpenoid plant hormones. Nature
Czech J Genet Plant Breed 50:144–150 455:195–200
Rubiales D (2018) Can we breed for durable resistance to Van Ooijen J (2006) JoinMap 4. Software for the calculation of
broomrapes? Phytopathol Mediterr 57(1):170–185. https:// genetic linkage maps in experimental populations. Kyazma
doi.org/10.14601/Phytopathol_Mediterr-22543 BV, Wageningen
Rubiales D, Rojas-Molina MM, Sillero JC (2016) Characteri- Vogel JT, Walter MH, Giavalisco P, Lytovchenko A, Kohlen W,
zation of resistance mechanisms in Faba Bean (Vicia faba) Charnikhova T et al (2010) SlCCD7 controls strigolactone
against Broomrape Species (Orobanche and Phelipanche biosynthesis, shoot branching and mycorrhiza-induced
spp.). Front Plant Sci 7:1747. https://doi.org/10.3389/fpls. apocarotenoid formation in tomato. Plant J 61:300–311
2016.01747 Waters MT, Gutjahr C, Bennett T, Nelson DC (2017) Strigo-
Samejima H, Sugimoto Y (2018) Recent research progress in lactone signaling and evolution. Annu Rev Plant Biol
combatting root parasitic weeds. Biotechnol Biotechnol 68:291–322
Equip. https://doi.org/10.1080/13102818.2017.1420427 Westwood JH (2000) Characterization of the Orobanche-Ara-
Sierro N, Battey JND, Ouadi S, Bakaher N, Bovet L, Willig A bidopsis system for studying parasite-host interactions.
et al (2014) The tobacco genome sequence and its com- Weed Sci 48:742–748
parison with those of tomato and potato. Nat Commun Xie X, Yoneyama K, Yoneyama K (2010) The strigolactone
5:3833 story. Annu Rev Phytopathol 48:93–117. https://doi.org/
Strickler SR, Bombarely A, Mueller LA (2012) Designing a 10.1146/annurev-phyto-073009-114453
transcriptome next-generation sequencing project for a Yang C, Xu L, Zhang N, Islam F, Song W et al (2017) iTRAQ-
nonmodel plant species. Rev Am J Bot 99:257–266 based proteomics of sunflower cultivars differing in resis-
Tong Z, Xiao B, Jiao F, Fang D, Zeng J, Wu X, Chen X, Yang J, tance to parasitic weed Orobanche cumana. Proteomics.
Li Y (2016) Large-scale development of SSR markers in https://doi.org/10.1002/pmic.201700009
tobacco and construction of a linkage map in flue-cured Yoder JI, Scholes JD (2010) Host plant resistance to parasitic
tobacco. Breed Sci 66:381–390 weeds; recent progress and bottlenecks. Curr Opin Plant
Torres-Vera R, Garcıa JM, Pozo MJ et al (2016) Expression of Biol. https://doi.org/10.1016/j.pbi.2010.04.011
molecular markers associated to defense signalling path- Yoshida S, Shirasu K (2012) Plants that attack plants: molecular
ways and strigolactone biosynthesis during the early elucidation of plant parasitism. Curr Opin Plant Biol.
interaction tomato-Phelipanche ramosa. Physiol Mol Plant https://doi.org/10.1016/j.pbi.2012.07.004
Pathol 94:100–107 Yoshida S, Cui S, Ichihashi Y, Shirasu K (2016) The hausto-
Trabelsi I, Abbes Z, Amri M, Kharrat M (2016) Study of some rium, a specialized invasive organ in parasitic plants. Annu
resistance mechanisms to Orobanche spp. infestation in Rev Plant Biol 67:643–667. https://doi.org/10.1146/
faba bean (Vicia faba L.) breeding lines in Tunisia. Plant annurev-arplant-043015-111702
Prod Sci 19:562–573
Trabelsi I, Yoneyama K, Abbes Z, Amri M, Xie X, Kisugi T et al
Publisher’s Note Springer Nature remains neutral with
(2017) Characterization of strigolactones produced by
regard to jurisdictional claims in published maps and
Orobanche foetida and Orobanche crenata resistant faba
institutional affiliations.
bean (Vicia faba L.) genotypes and effects of phosphorous,

123