Letters in Applied Microbiology 2002, 34, 337–342

Production, purification and characterization of an

extracellular keratinase from Lysobacter NCIMB 9497

J.D. Allpress, G. Mountain and P.C. Gowland

Department of Biological Sciences, Staffordshire University, Stoke-on-Trent, UK

2001/280: received 20 September 2001 and accepted 15 January 2002

J . D . A L L P R E S S , G . M O U N T A I N A N D P . C . G O W L A N D . 2002.
Aims: The aim of this study was to determine the keratinolytic ability of a range of bacteria
and subsequently, to characterize the keratinase showing the greatest biotechnological potential.
Methods and Results: Nine bacteria, reported to produce extracellular proteases, were
screened for production of keratinases. Of these, Lysobacter NCIMB 9497 exhibited the highest
keratinolytic activity. The keratinase from this strain (Mr 148 kDa) was purified and
characterized. Optimum activity occurred at 50
C; no inhibition was demonstrated by
phenylmethylsulphonyl fluoride (PMSF), but inhibition by EDTA was demonstrated.
Conclusions: This study indicates that keratinase is a metalloprotease with a high degree of
keratinolytic activity and stability.
Significance and Impact of the Study: This is the first detailed report of a metalloprotease
with keratinolytic activity. The novel reaction mechanism, degree of keratinolytic activity and
stability indicate considerable biotechnological potential in the leather industry, and in the
processing of poultry waste.

represent a potentially valuable source of protein as animal

INTRODUCTION
feedstock if keratinolysis can be achieved (Shih 1993).
Keratin is an insoluble macromolecule requiring the secre- Keratinolytic enzymes have been studied from a variety
tion of extracellular enzymes for biodegradation to occur. of fungi and, to a lesser extent, bacteria. However, much
Keratin comprises long polypeptide chains, which are current research is centred on the potential use of
resistant to the activity of non-substrate-specific proteases. keratinases of bacterial origin for the industrial treatment
Adjacent chains are linked by disulphide bonds thought of keratin-containing compounds, e.g. serine proteases
responsible for the stability and resistance to degradation of produced by Bacillus licheniformis PWD-1 (Lin et al.
keratin (Safranek and Goos 1982). 1997) and B. licheniformis Carlsberg NCIMB 6816 (Evans
The degradation of keratinous material is important et al. 2000). Such interest results from the broad substrate
medically and agriculturally (Shih 1993; Matsumoto 1996). range of these bacterial enzymes, their rates of activity
Secretion of keratinolytic enzymes is associated with der- towards keratin-containing compounds and their thermo-
matophytic fungi, for which keratin is the major substrate stability.
(Matsumoto 1996). However, the production of such This paper describes screening of bacterial isolates with
enzymes is not exclusive to dermatophytes, since geophilic known proteolytic activity for keratinolysis, and the purifi-
species have demonstrated keratinase production (Kushwaha cation and characterization of an extracellular keratinase
and Nigam 1996). World-wide poultry processing plants produced by Lysobacter NCIMB 9497.
produce millions of tons of feathers as a waste product
annually (Santos et al. 1996), which consists of approxi-
MATERIALS AND METHODS
mately 90% keratin; the keratin is largely responsible for
their high degree of recalcitrance. However, they also Organisms
The strains used in this study, and their sources, are shown
Correspondence to: Dr J.D. Allpress, Department of Biological Sciences, in Table 1.
Staffordshire University, College Road, Stoke-on-Trent, ST4 2DE, UK.

ª 2002 The Society for Applied Microbiology

338 J . D . A L L P R E S S ET AL.

Table 1 Bacterial strains screened for extra-

Strain Source
cellular proteolytic activity
Lysobacter NCIMB 9497 National Collection of Industrial and Marine
Bacteria Ltd (Torrey, Scotland)
Bacillus amyloliquefaciens NCIMB 12077 National Collection of Industrial and Marine
Bacteria Ltd (Torrey, Scotland)
Bacillus brevis NCIMB 8146 National Collection of Industrial and Marine
Bacteria Ltd (Torrey, Scotland)
Bacillus licheniformis NCIMB 10689 National Collection of Industrial and Marine
Bacteria Ltd (Torrey, Scotland)
Bacillus cereus NCTC 10320 National Collection of Type Cultures and
Pathogenic Fungi (Colindale, England)
Bacillus subtilis NCTC 8236 National Collection of Type Cultures and
Pathogenic Fungi (Colindale, England)
Bacillus sphaericus 2632 Collection of Bacillus sphaericus, Institut
Pasteur, Paris, France
Bacillus megaterium H50 600/4 Philip Harris Ltd (Lichfield, England)

25
C, and release of azo dye was determined spectro-

Growth conditions and screening of isolates
photometrically (595 nm) following centrifugation
Cultures were grown overnight on nutrient broth with (16 000 g, 30 min).
rotary shaking (25
C, 120 rev min)1). Growth was deter-
mined spectrophotometrically (600 nm). Cultures were
Production and preparation of keratinase
incubated for 14 days.
Lysobacter. NCIMB 9497 was grown on MMV medium
(2 l) containing u.v.-sterilized powdered keratin (2%, w/v)
Determination of proteolytic activity
for 31 days. Viable counts and extracellular protein
Culture medium from each flask was filter-sterilized concentration (Lowry et al. 1951) were determined. Cells
(Millipore, 0Æ2 lm) and aliquots (100 ll) pipetted into wells were removed from the medium by filter sterilization
(diameter 6 mm) cut into skimmed milk agar plates (15% (Whatman polycap TF, 0Æ2 lm pore size). Keratinase
(v/v) skimmed milk, 1Æ4% (w/v) Difco technical agar concentration was increased by ultrafiltration (Omegacell,
number 3). The plates were incubated at 25
C for 6 h and Filtron, Pall Life Sciences, Portsmouth, UK; 10 kDa
zones of clearance were measured. molecular size cut-off). The filtrate was passed down a
Sephadex G-50 gel-filtration column (Pharmacia Fine
Chemicals, Amersham Pharmacia Biotech UK Ltd, Little
Determination of keratinase activity
Chalfont, Bucks., UK) (750 · 45 mm: height · diameter)
Keratinolytic activity was assayed using the keratin azure and purity was determined by sodium dodecyl sulphate-
method (Hänel et al. 1991). Powdered keratin (ICN Bio- polyacrylamide gel electrophoresis (SDS-PAGE) (Laemmli
medicals Inc., Aurora, USA) was sterilized by u.v. irradiation 1970) using pre-stained SDS-PAGE markers (molecular
(Ultraviolet Products Ltd, Cambridge, UK; Chromatovue weight 34–123 kDa; Sigma).
TM-10) (302 nm, maximum peak intensity 7Æ0 mW cm)2,
2 h) before addition to vitamin-supplemented minimal salts
Enzyme characterization
medium (MMV) (Janssen et al. 1984) (final keratin concen-
tration 2% (w/v)). The flasks were inoculated with 5 ml Temperature. The effect of temperature on keratinase
overnight nutrient broth-grown cultures and incubated for activity was determined by the addition of 20 ll keratinase
14 days at 25
C with shaking (120 rev min)1). (3 mg ml)1 protein) to 1Æ5 ml phosphate buffer (100 mmol
Following incubation, culture samples were filter- 1)1, pH 7Æ5) containing 15 mg powdered keratin, and
sterilized (as above), and aliquots (50 ll) added to incubating at a range of temperatures (30, 50, 60 and
sterile 1Æ5 ml microcentrifuge tubes (Eppendorf UK Ltd, 70
C) for 10 h. Peptide release was determined spectrophoto-
Cambridge, UK) containing approximately 5 mm lengths metrically (280 nm; 1 unit of activity (U) ¼ the amount of
of u.v.-sterilized keratin azure (30 mg) and 950 ll enzyme causing an increase of 1Æ0 A280 unit within
MMV medium. Assays were incubated for 12 days at 1 minute).
ª 2002 The Society for Applied Microbiology, Letters in Applied Microbiology, 34, 337–342
 KERATINASE OF LYSOBACTER 9497 339

Stability. The keratinase was stored at 4
C for a period of enzymes produced by each strain, the ratio of proteolytic
three months. Samples of keratinase were assayed monthly activity to cell density was estimated by dividing the zone
at 25
C using the technique described above. diameter by the optical density of the overnight culture from
which the enzyme was obtained. Bacillus amyloliquefaciens
Synthetic substrates. The substrate specificity was deter- NCIMB 12077 showed the greatest total proteolytic activity
mined by incubation with chromogenic substrates (zone of clearance ¼ 19 mm), while Bacillus subtilis NCTC
(N-CBZ-L-phenylalanine p-nitroanaline, n-acetyl-L-alanine 8236 had the highest activity to final cell density ratio
p-nitroalanine, N-CBZ-L-arginine p-nitroalanine; Sigma). (25 mm absorbance unit)1). Keratinolytic activity of cell-
Substrates were dissolved in dimethyl sulphoxide and added free media from cultures grown on nutrient broth, deter-
to the assay buffer (5 ml Tris-sulphate buffer, 50 mmol 1)1, mined by the keratin azure method, revealed Lysobacter
pH 7Æ8) to a final concentration of 10 mmol 1)1 before NCIMB 9497 to have both the highest overall keratinolytic
equilibration at 25
C in an orbital shaker (120 rev min)1). activity and the highest activity to protein ratio (0Æ778
Assays were initiated by addition of enzyme (250 ll, absorbance units and 7Æ0 · 10)3 absorbance units (lg
3 mg ml)1 protein). Enzyme activity was determined by ml)1))1, respectively) (Table 3).
measuring dye release spectrophotometrically (405 nm).
Production and preparation of keratinase
Inhibitors. Assays were performed in phosphate buffer
(2Æ5 ml; 100 mmol 1)1, pH 7Æ5) containing u.v.-sterilized Lysobacter NCIMB 9497 was harvested shortly after
powdered keratin (25 mg). Buffer was incubated overnight extracellular protein concentration peaked (31 days of
at 4
C in the presence of inhibitor. Phenylmethylsulphonyl incubation). This was 5 days after cell numbers began to
fluoride (PMSF) was dissolved in acetonitrile and added to a decrease (Fig. 1). The keratinase was purified to homogen-
final concentration of 5 mmol 1)1. Ethylenediaminetetra- eity following gel chromatography, as determined by SDS-
acetic acid (EDTA) was added to a concentration of 5 mmol PAGE (Fig. 2). The purification procedure is summarized
1)1. Assays were initiated by the addition of enzyme (250 ll, in Table 4. Following calibration of the gel filtration column
3 mg ml)1 protein). Assays were incubated at 25
C for 10 h. with molecular mass markers, the molecular mass of the
Peptide release was determined spectrophotometrically keratinase was estimated at 130–150 kDa. The molecular
(280 nm). Decrease in activity was compared with controls mass as determined by SDS-PAGE was approximately
containing the same enzyme concentration of either chy- 148 kDa (Fig. 2).
motrypsin or carboxypeptidase, using casein as substrate
under otherwise identical conditions.
Enzyme characterization
The optimum temperature for keratinolytic activity was
RESULTS
50
C (40 U mg)1).
Screening of isolates The enzyme showed no decrease in activity during the
3 month storage period. Activity levels of 12Æ6, 12Æ2, 15Æ0,
All isolates screened for caesinolytic activity showed proteo-
13Æ1 U mg)1 protein were recorded at 0, 1, 2 and 3 months,
lytic activity, as demonstrated by zones of clearing around
respectively.
each well (Table 2). To compare the level of proteolytic

Table 2 Activity of 14 day culture media

Absorbance of Ratio of zone of
towards casein, as determined by zone of
overnight culture Zone of clearance clearance to cell density
clearance in skimmed milk agar (means of
three replicates) Organism (600 nm) (diameter in mm)* (mm absorbance unit)1)

Bacillus subtilis 0Æ438 11 25

Bacillus amyloliquefaciens 1Æ060 19 18
Bacillus licheniformis 0Æ337 6 18
Lysobacter sp. 0Æ500 8 16
Bacillus brevis 1Æ030 13 13
Bacillus megaterium 1Æ106 14 13
Bacillus sphaericus 1Æ216 14 12
Bacillus cereus 1Æ348 14 10

*Total diameter of clearance minus well diameter.

No clearing was observed surrounding wells containing uninoculated nutrient broth.

ª 2002 The Society for Applied Microbiology, Letters in Applied Microbiology, 34, 337–342
340 J . D . A L L P R E S S ET AL.

Table 3 Keratinase activity of cell-free

Protein Increase in Ratio of increase in absorbance
media from cultures grown on keratin as
concentration absorbance at to protein concentration measured by the keratin azure assay
Organism (lg ml)1)* 595 nm (absorbance units [lg ml)1])1)

Bacillus amyloliquefaciens 155 0Æ203 1Æ3 · 10)3

Bacillus brevis 185 0Æ595 3Æ2 · 10)3
Bacillus cereus 80 0Æ315 3Æ9 · 10)3
Bacillus megaterium 115 0Æ158 1Æ4 · 10)3
Bacillus sphaericus 60 0Æ014 2Æ3 · 10)4
Bacillus subtilis 145 0Æ089 6Æ1 · 10)4
Bacillus licheniformis 70 0Æ001 1Æ4 · 10)5
Lysobacter sp. 110 0Æ765 6Æ9 · 10)3

*Protein concentration of cell-free filtrate after 14 days of culture incubation.

After 12 days of incubation of cell-free filtrate with keratin azure, mean of three replicates –
increase in absorbance of sterile control after 12 days of incubation.

Fig. 1 Growth of Lysobacter NCIMB 9497

on powdered keratin. (m) Cell concentration
determined by dilution plate technique; (j)
protein concentration determined by the
method of Lowry et al. 1951

1 2 3 4 5
Endoprotease activity was observed with N-CBZ-L-
phenylalanine p-nitroanaline and N-CBZ-L-arginine
123 kDa p-nitroalanine. Activity was strongest towards N-CBZ-L-
phenylalanine p-nitroanaline. No activity was observed with
89 kDa N-acetyl-L-alanine p-nitroalanine.
67 kDa No decrease in keratinase activity occurred in the presence
of PMSF at 5 mmol 1)1. The same concentration of PMSF
resulted in a 90% decrease in chymotrypsin activity towards
50 kDa casein.
Addition of EDTA (5 mmol 1)1) resulted in a 50%
37·5 kDa reduction in keratinase activity towards keratin, compared
34 kDa with a 73% reduction of carboxypeptidase activity towards
casein.

Fig. 2 SDS-PAGE of keratinase. Samples were extracted during the DISCUSSION

purification procedure and analysed by SDS-PAGE according to the
method of Laemmli 1970. Lane 1: post-gel filtration concentrate; lane Cultures selected for the screening programme were all
2: molecular mass marker proteins; lane 3: post-gel filtration concen- reported to produce extracellular proteases (suppliers
trate; lane 4: medium post-concentration; lane 5: medium pre- catalogues). Results of the initial screening confirmed
concentration these findings (Table 2). Keratinolytic activity was
ª 2002 The Society for Applied Microbiology, Letters in Applied Microbiology, 34, 337–342
 KERATINASE OF LYSOBACTER 9497 341

Table 4 Summary of the purification of

Protein Total Total Specific Degree of
keratinase from Lysobacter 9497
concentration volume protein activity purification
Purification step (mg ml)1) (ml) (mg) (U* mg)1) (fold)

Medium 0Æ255 2060 463 2Æ5 1

Concentrated medium 2Æ700 100 270 3Æ8 2
Post-gel filtration 0Æ625 50 31 14Æ6 6
Post-gel filtration concentrate 3Æ000 10 30 15Æ1 6

*U ¼ amount of enzyme required to cause an increase of 1Æ0 A280 unit within 1 minute (after
Böckle et al. 1995).

detected in all strains assayed with the exception of enzyme–substrate interactions on the surface of keratin
B. licheniformis NCIMB 10689 and B. sphaericus 2632. particles have been reported to limit keratinolysis (Böckle
Lysobacter NCIMB 9497 was investigated further because et al. 1995). Potentially, this may be overcome by the
of the relatively high ratio of keratinolytic activity to cell addition of a secondary metalloproteases.
density which indicated the greatest biotechnological Inhibition of the metalloprotease characterized in this
potential in terms of efficiency of enzyme production study by chelating agents provides a potential method for
(Table 2). temporary inactivation during storage, reducing autolysis
Comparison of keratinolytic activities with those in the associated with proteolytic enzymes. In addition, the
literature is difficult due to the variety of keratin substrates metalloenzyme nature of this keratinase presents a potential
employed by investigators. However, over a similar 12 day method of enzyme immobilization, which has been reported
period using keratin azure as substrate, cell suspensions to increase stability due to a reduced enzyme autolysis (Lin
from Dermatophilus congolensis (Hänel et al. 1991) showed a et al. 1996). The application of immobilization techniques to
mean increase of 0Æ41 absorbance units compared with a the metalloprotease characterized here is therefore worthy of
mean increase of 0Æ778 absorbance units for Lysobacter further study.
NCIMB 9497 cell-free extract.
The concentration of extracellular keratinase produced by
REFERENCES
Lysobacter NCIMB 9497 was maximal after 29 days of
growth on keratin. This rate of keratinase production Böckle, B., Galunsky, B. and Müller, R. (1995) Characterisation of a
compares with a maximum concentration after 21 days in keratinolytic serine proteinase from Streptomyces pactum DSM40530.
Chrysosporium queenslandicum and Graphium penicilloideus Applied and Environmental Microbiology 61, 3705–3710.
(Malviya et al. 1992). Dozie, I.N.S., Okeke, C.N. and Unaeze, N.C. (1994) A thermostable,
alkaline-active, keratinolytic proteinase from Chrysosporium keratino-
Purified keratinase showed no inhibition in the presence
philum. World Journal of Microbiology and Biotechnology 10, 563–567.
of the serine protease inhibitor PMSF. However, inhibition
Dunn, B.M. (1989) Determination of protease mechanism. In Proteo-
was demonstrated in the presence of EDTA, indicating that lytic Enzymes: a Practical Approach ed. Beynon, R.J. and Bond, J.S.
the keratinase is a metalloprotease. The majority of the pp. 57–81. Oxford: Clarendon Press.
keratinases reported in the literature have been shown to be Evans, K.L., Crowder, J. and Miller, E.S. (2000) Subtilisins of Bacillus
serine proteases. A metalloprotease was reported to be spp. hydrolyse keratin and allow growth on feathers. Canadian
produced extracellularly by Streptomyces pactum DSM40530 Journal of Microbiology 46, 1004–1011.
(Böckle et al. 1995). However, a serine protease produced in Hänel, H., Kalisch, J., Keil, M., Marsch, W.C. and Buslau, M. (1991)
conjunction with the metalloprotease was largely responsible Quantification of keratinolytic activity from Dermatophilus congolensis.
for keratinolytic activity in this strain. Metalloproteases Medical Microbiology and Immunology 180, 45–51.
produced as the sole keratinolytic enzyme are limited to the Janssen, D.B., Scheper, A. and Witholt, B. (1984) Biodegradation of
2-chlorethanol and 1,2-dichloroethane by pure bacterial cultures. In
tentative report of Dozie et al. (1994), who found keratin-
Innovations in Biotechnology. Progress in Industrial Microbiology Vol.
olytic activity to be stimulated in the presence of Fe2+ ions
20, ed. Houwink, E.H. and van der Meer, R.R. p. 169. Amsterdam:
and inhibited by 1,10 o-phenanthroline. However, EDTA Elsevier Biomedical Press.
was found not to be inhibitory. The effect of PMSF was not Kushwaha, R.K.S. and Nigam, N. (1996) Keratinase production by
determined. Metal ions have been found to stimulate the some geophilic fungi. National Academy of Science (India) Letters 19,
activity of some serine proteases due to stabilizing effects 107–109.
(Dunn 1989), which cannot be ruled out in this case. Laemmli, U.K. (1970) Cleavage of structural proteins during the
The tentative classification of this keratinase as a metallo- assembly of the head of bacteriophage T4. Nature 227,
protease holds great biotechnological promise. Restricted 680–685.

ª 2002 The Society for Applied Microbiology, Letters in Applied Microbiology, 34, 337–342
342 J . D . A L L P R E S S ET AL.

Lin, X., Lee, C.-G., Casale, E.S. and Shih, J.C.H. (1997) Purification grounds of a gelatin factory, Jabalpur, India. Mycopathologia 120,
and characterization of a keratinase from a feather-degrading Bacillus 1–4.
licheniformis strain. Applied and Environmental Microbiology 58, Matsumoto, T. (1996) Fungal diseases in dermatology. In Principle and
3271–3275. Practice of Clinical Mycology ed. Kibbler, C.C., Mackenzie, D.W.
Lin, X., Shih, J.C.H. and Swaisgood, H.E. (1996) Hydrolysis of feather and Odds, F.C.J. pp. 103–129. New York: J. Wiley and Sons Ltd.
keratin by immobilized keratinase. Applied and Environmental Safranek, W.W. and Goos, R.D. (1982) Degradation of wool by
Microbiology 62, 4273–4275. saprophytic fungi. Canadian Journal of Microbiology 28, 137–140.
Lowry, O.H., Rosenbrough, N.J., Farr, A.L. and Randall, R.K. (1951) Santos, R.M.D.B., Firmino, A.A.P., de Sá, C.M. and Felix, C.R.
Protein measurement with the folin phenol reagent. Journal of (1996) Keratinolytic activity of Aspergillus fumigatus Fresenius.
Biological Chemistry 193, 265–275. Current Microbiology 33, 364–370.
Malviya, H.K., Rajak, R.C. and Hasija, S.K. (1992) Synthesis and Shih, J.C.H. (1993) Recent developments in poultry waste digestion
regulation of extracellular keratinase in three fungi isolated from the and feather utilization – a review. Poultry Science 72, 1617–1620.

ª 2002 The Society for Applied Microbiology, Letters in Applied Microbiology, 34, 337–342