The chitinolytic activity of Penicillium janthinellum P9: purification, partial characterization and potential applications


Fenice Dipartimento di Agrobiologia e Agrochimica, University of Tuscia, Via S. C. De Lellis, 01100 Viterbo, Italy (e-mail:


Aims: To purify and characterize the chitinolytic activity of Penicillium janthinellum P9 and to evaluate possible uses of the purified enzymes in the control of fungal growth and spore germination.

Methods and Results: The chitinolytic activity of P. janthinellum P9 was associated to two β-N-acetyl-hexosaminidases (CHI1 and CHI2) that were purified by preparative isoelectric focusing and preparative electrophoresis and partially characterized. Treatment of test fungi with purified enzyme solutions caused reduced spore germination, reduction of hyphal length and mycelial damage. The combined action of the two enzymes and a systemic fungicide completely inactivated pests and food-spoiling moulds such as Fusarium solanii, P. canescens and Cladosporium cladosporioides. Treatment with the two enzymes increased germination of freeze-dried fungal spores.

Conclusions:The chitinolytic activity of P. janthinellum P9 is associated with two extracellular β-N-acetyl-hexosaminidases that can cause damage to the cell walls of other fungi.

Significance and Impact of the Study: This appears to be the first report on the characterization of extracellular chitinolytic enzymes produced by a Penicillium strain. The results of this study might have some impact in the applied research field.


Species of Penicillium are versatile organisms and have long been known as producers of a wide variety of extracellular enzymes, some of which are of industrial interest (Hamlyn et al. 1987; Petruccioli et al. 1988; Godfrey and West 1996); in addition, some of them are considered food-grade organisms and are currently used as sources of enzymes for the food and feed industry (Godfrey and West 1996; Wigley 1996).

Although Penicillium species are often involved in the hydrolysis of molecules such as cellulose and xylans (Hamlyn et al. 1987), little is known on their ability to degrade chitin while most information refers to the degradation of endogenous chitin during fungal cell wall development (Yamamoto et al. 1985; Rodriguez et al. 1994). Due to the many applications of chitin-hydrolysing enzymes (Patil et al. 2000), a strain of Penicillium characterized by high levels of extracellular chitinolytic activity might be of great interest at industrial level; in this context, among fungi Trichoderma harzianum is currently of major interest (Godfrey and West 1996).

A preliminary screening (unpublished results) of 66 strains of 31 species of Penicillium for their ability to degrade chitin led to the surprising result that only one strain, P. janthinellum P9, had appreciable activity. In this paper, we report on the purification and partial characterization of its chitinolytic activity and on some potential applications of the purified enzymes such as the inhibition of fungal growth, when used alone or in combination with a fungicide, and the enhancement of percentage germination of freeze-dried fungal spores.



Chitin (from crab shells), β-D-N-acetylglucosamine (GlcNAc), β-D-N-N′-diacetylchitobiose (chitobiose), β-D-N-N′-N′′-triacetylchitotriose (chitotriose), p-nitrophenyl-β-D-N-acetylglucosaminide, p-nitrophenyl-β-D-N-N′-diacetylchitobiose, p-nitrophenyl-β-D-N-N′-N′′-triacetylchitotriose and glycol chitosan were from Sigma Chemical Co. (St Louis, MO, USA). Chitosan (low, medium and high molecular weight), cellulose acetate and CM-cellulose were from Fluka BioChemika (Buchs, Switzerland). Malt Extract Agar (MEA) and Potato Dextrose Agar (PDA) were from Oxoid (Unipath Ltd, Basingstoke, Hampshire, UK); Yeast Nitrogen Base (YNB) was from Difco Laboratories (Detroit, MI, USA).

Colloidal chitin was prepared as reported by Hankin and Anagnostakis (1975); chitin from cell walls of Fusarium solanii and Sclerotium rolfii were prepared as reported by Gupta et al. (1995); glycol chitin was prepared as reported by Molano et al. (1979).

The fungicide used was a commercial preparation of methyl 1-(butylcarbamoyl) benzimidazol-2-ylcarbamate (MBC), containing 50% of the active principle, from Dupont de Nemours and Co. (Wilmington, DE, USA).

All other chemicals were of analytical grade.


P. janthinellum P9 was from the culture collection of the Dipartimento di Agrobiologia Agrochimica, University of Tuscia, Viterbo, Italy. The strains of Cladosporium cladosporioides, Fusarium moniliforme, F. solanii, Eremicella nidulans, Mucor plumbeus, Paecilomyces variotti, Penicillium citrinum, P. corylophilum, P. camembertii, P. canescens, P. granulatum and S. rolfii were from the culture collection of the Nestlè Research Center, Nestlè Ltd, Lausanne, Switzerland. The cultures were maintained on MEA at 4–6°C and subcultured every month.

Enzyme production and purification

The chitinolytic enzymes were obtained from P. janthinellum P9, cultivated as reported by Fenice et al. (1998).

After centrifugation (5000 g, 10 min) and filtration (GFC membrane, 20 mm, Whatman, UK) of the cultural broth to remove the mycelium, the crude enzyme solution was brought to 80% saturation with ammonium sulphate and left overnight at 4°C. The precipitate formed was collected by centrifugation (8000 g, 15 min), dissolved in 15 ml of 5 mmol l−1 citrate-phosphate buffer pH 5·0 and dialysed overnight against several volumes of the same buffer. The final enzyme solution, which contained ca 90% of the initial chitinolytic activity, was concentrated by lyophilization. The lyophilized material had a specific activity of 7·65 U mg−1 of protein.

Step 1, preparative isoelectric focusing (Rotofor).

After resuspension in distilled water, the lyophilized enzyme obtained as above was applied to a Rotofor apparatus (Bio-Rad Laboratories, Hercules, CA, USA) according to the manufacturer’s directions using a Rotolytes mixture (range 4·5–6·1, Bio-Rad Laboratories) to form a pH gradient between 4·5 and 5·7. Active fractions were collected, pooled and concentrated by centrifugation (6000 g, 10 min) with a Centricon 10000 membrane system (Amicon, USA). Protein purity was tested by sodium dodecyl sulphate polyacrylamide gel electrophoresis (SDS-PAGE).

Step 2, preparative electrophoresis.

Electrophoresis was carried out under denaturing conditions in presence of sodium dodecyl sulphate (SDS) as reported by Laemmli (1970), using the Prep Cell Model 491 equipment (Bio-Rad Laboratories, Hercules, CA, USA). Acrylamide concentration of the upper (stacking) and lower (separating) gels were 4 and 12%, respectively.

The active pool from step 1 (containing two proteins) was diluted in SDS electrophoresis buffer and loaded onto the gel. After running (6 h at 40 mA, 250–300 V), the fractions that contained proteins (detected spectrophotometrically at 280 nm) were collected, renaturated (citrate–phosphate buffer 50 mmol l−1, pH 5·5) and tested for activity. Purity was tested by SDS-PAGE.

All purification procedures were carried out at 4°C.

Enzyme characterization

The effect of pH on the enzyme activity was determined by varying the pH of the reaction mixture using 50 mmol l−1 glycine/HCl buffer (pH 1·0–3·0), 50 mmol l−1 citrate–phosphate buffer (pH 3·5–7·0) and 50 mmol l−1 phosphate buffer (pH 7·0–8·0). The denaturing effects of pH were investigated by incubating the enzyme solution for 1 h at 25°C in the range of pH from 2·0 to 6·0. The effect of temperature on the enzyme activity was determined at pH 4·0 in the range of temperatures from 5 to 80°C. The enzyme half-life was carried out, incubating the enzyme solution at optimal temperature and pH for 120 min: samples were taken every 10 min to determine the residual enzyme activity by the standard enzyme assay.

The effects of metal ions such as Ca2+, Cu2+, Fe2+, Hg2+, Mg2+, Mn2+, Ni2+, Zn2+ (as CaCl2, CuSO4, FeSO4, HgCl2, MgSO4, MnCl2, NiCl2, ZnCl2) and sodium azide, iodoacetic acid and ethylene-diamino-tetracetic acid (EDTA) on the enzyme activity were determined, adding to the reaction mixture the above compounds to a final concentration of 10 mmol l−1; the relative activities were determined using colloidal chitin as the reaction substrate (see the standard enzyme assay). Inhibition by 10 mmol l−1 of D-glucosamine, N-acetyl-β-D-glucosamine, chitobiose and chitotriose was also tested under the same conditions.


SDS-PAGE was used to check the purity of protein and to determine the molecular mass of the purified enzymes under denaturing conditions according to Laemmli (1970). Stacking and separating gel containing 5 and 12·5% polyacrylamide, respectively, were used. After electrophoresis, gels were stained with Coomassie Brilliant Blue R-250 (Pharmacia, Sweden) or by the periodic acid-Schiff reagent for glycoproteins (Zacharius et al. 1969). Molecular markers (low range) were from Bio-Rad Laboratories.

Isoelectric focusing

Isoelectric points of the purified enzymes were determined using a Multiphor II cell (Pharmacia, Sweden) and a precasted polyacrylamide gel Pagplate (Pharmacia, Sweden) with ampholine pH range 4·0–6·5. Broad range (3·5–10) IEF gels protein calibration kits were from Pharmacia (Sweden); staining for protein was as described for SDS-PAGE.

Protein estimation

Protein concentration was estimated spectrophotometrically at 595 nm using bovine serum albumin as the standard protein, by the Bio-Rad Laboratories procedure (Bradford 1976).

Substrate specificity and mechanism of action

Enzyme specific activity was evaluated on the following substrates (0·5% in citrate-phosphate buffer 50 mmol l−1, pH 5·0): chitin from crab shells, chitin from fungal cell walls (S. rolfii and F. solanii), glycol chitin, cellulose, carboxymethyl cellulose and chitosan (high, medium and low molecular weight). The activities, expressed as release of reducing sugars after incubation at 50°C for 10 min, were compared with that obtained on colloidal chitin (see enzyme assay).

Chitinolytic activities were also evaluated as the p-nitrophenol released from the following synthetic substrates: p-nitrophenyl-β-D-N-galactosaminide, p-nitrophenyl-β-D-N-acetylglucosaminide, p-nitrophenyl-β-D-N-N′-diacetylchitobiose, p-nitrophenyl-β-D-N-N′-N′′-triacetylchitotriose as reported by Valadares and Peberdy (1993). Relative activities were calculated by taking the activity recorded on the p-nitrophenyl-β-D-N-acetylglucosaminide as 100%. Apparent Km and Vmax values for this substrate were calculated by non-linear regression analysis utilizing the Enzfitter program (version 1·05, Elsevier).

The mechanism of action of the purified enzyme was determined by studying the release of hydrolysis products from chitin and synthetic chito-oligosaccharides. The products of the enzymatic reaction were analysed by HPLC as reported by Fenice et al. (1998).

Inhibition of fungal spore germination

Inhibition of spore germination was carried out by incubating (25°C, 24 h) fungal spores (ca 5 × 105 spores ml−1) in a medium containing YNB and glucose (5%) with a mixture of concentrated enzyme solutions added to give a final enzyme concentration of 10 U ml−1. To determine the inhibition of spore germination, samples were taken after 12 and 24 h of incubation and observed under the light microscope: the germinating spores were counted and the length of the germinated hyphae measured. The test fungi used were: P. citrinum, P. corylophilum, P. granulatum, P. camembertii, P. canescens, C. cladosporioides, M. plumbeus, F. moniliforme, F. solanii, E. nidulans and P. variotti.

The combined effects of the enzymes from P. janthinellum P9 and the fungicide MBC were studied on spore suspensions (ca 5 × 105 spores ml−1) of either P. canescens, F. solanii or C. cladosporioides, to which the antifungal agent and the enzyme solutions were added at a range of concentrations as reported in Table 6. After each treatment, the spore ability to germinate (% of germinated spores) was observed under the light microscope on samples taken after 12 and 24 h of incubation at 25°C. Moreover, after 24 h of incubation samples were plated on MEA to observe the survival after each treatment. After a further 24, 48 and 72 h of incubation at 25°C, colonies were counted as cfu ml−1.

Enhancement of germination of freeze-dried spores

Spores of the test moulds (ca 109 spores ml−1) were freeze-dried using skim milk (10%) as cryoprotectant. The lyophilized material was then resuspended in sterile water solutions, having final chitinase activities of 0, 0·01, 0·05 and 0·5 U ml−1. After 3 h of incubation at room temperature, samples were taken and plated on MEA; the germinating ability was then evaluated as cfu ml−1.

Standard enzyme assay

The chitinolytic activity was determined as reported previously (Fenice et al. 1998) by measuring the amount of reducing sugars liberated from colloidal chitin by the enzyme activity; N-acetyl-D-glucosamine was used for the standard curve. Under the assay conditions, one unit (U) of enzyme activity was defined as the amount of enzyme which released 1 μmol of N-D-acetylglucosamine ml−1 min−1.

Statistical analysis

One-way analysis of variance (ANOVA) and pairwise multiple comparison procedures (Tukey test) were carried out using the statistical software SigmaStat, version 2·0 (Jandel Corp., San Rafael, CA, USA).


Partial purification and properties of the chitinolytic activity of P. janthinellum P9

After the two-step procedure, two distinct enzymes (CHI1 and CHI2) were purified (5·8- and 7·0-fold, respectively) with a total yield of 28·3% (Table 1) and characterized (Table 2).

Table 1.   Purification of the chitinolytic activity of Penicillium janthinellum P9 Thumbnail image of
Table 2.   Physico-chemical and kinetic properties of the chitinolytic enzymes (CHI1 and CHI2) of Penicilliumjanthinellum P9 Thumbnail image of

SDS-PAGE showed that CHI1 and CHI2 had molecular masses of 45 and 35·4 kDa, respectively. After staining with the periodic acid–Schiff reagent, both enzymes appeared to be glycoproteins. Isoelectric points were 4·6 and 5·3 (thus indicating the possible presence of two isoenzymes) for CHI1 and 4·9 for CHI2. The optimal pH values and temperatures for activity of CHI1 and CHI2 were 5·5 and 4·5, and 55 and 50°C, respectively, with half-lives of 84 and 46 min.

The effect of various compounds on the enzyme activities was tested using colloidal chitin as the reaction substrate. Hg2+ caused strong inhibition of both enzymes (relative activities were 21 and 7% for CHI1 and CHI2, respectively). Mg2+ and Ni2+ had clear activating effects on CHI1 (128 and 125% of relative activity, respectively). Strong reduction of activity, in particular of CHI1, was recorded in presence of D-glucosamine and N-acetyl-D-glucosamine (42 and 50% of relative activity, respectively), while chitobiose and chitotriose caused complete inhibition of both enzymes.

Substrate specificity and mechanism of action

The specific activities of CHI1 and CHI2 were tested on several β-1,4 polysaccharides. Both enzymes showed high specificity for colloidal chitin and chitin from crab shells (100 and ca 90%, respectively). CHI1 and CHI2 differed, however, in their activities on glycol chitin which were 65 and 20%, respectively, of those on colloidal chitin and on fungal cell walls on which, in particular, CHI2 did not show any detectable activity. Both enzymes showed the same little activity on all the chitosans tested (20 and 15% of relative activity for CHI1 and CHI2, respectively). Finally, no activity was detected on cellulose.

p-Nitrophenyl-N-acetyl-β-D-glucosaminide, p-nitrophenyl-N-acetyl-β-D-galactosaminide, p-nitrophenyl-N,N′-diacetyl-β-D-chitobiose, p-nitrophenyl-N,N′,N′′-triacetyl-β-D-chitotriose were also tested as reaction substrates. The activities of both enzymes were highest on p-nitrophenyl-N-acetyl-β-D-glucosaminide (see Table 2 for Km and Vmax values) that were therefore taken as 100%. The activities of both CHI1 and CHI2 decreased with the increasing length of the chito-oligosaccharide chain being 83 and 92% and 60 and 81% on p-nitrophenyl-N,N′-diacetyl-β-D-chitobiose and on p-nitrophenyl-N,N′,N′′-triacetyl-β-D-chitotriose, respectively. Both enzymes were active also on p-nitrophenyl-N-acetyl-β-D-galactosaminide (73 and 61%, respectively). This behaviour indicated an activity characteristic of β-N-acetyl-D-hexosaminidases similar to those of the exo-chitinase 1 and 2 from T. harzianum (Drarborg et al. 1995) which, in fact, are now classified as hexosaminidases, EC

HPLC analysis of the products obtained from the enzyme/substrate reaction confirmed this indication. Both enzymes were able to hydrolyse colloidal chitin and several chito-oligosaccharides by detaching single units of N-acetyl-β-D-glucosamine (results not shown).

Action of the enzymes from P. janthinellum P9 on fungal spore germination and hyphal elongation

The effects of the treatment with the chitinolytic enzymes from P. janthinellum P9 on spore germination and hyphal elongation of several test fungi are shown in Table 3: experiments were carried out using a mixture of purified CHI1 and CHI2.

Table 3.   Effect of the chitinolytic enzymes (10 U ml–1) of Penicillium Janthinellum P9 on spore and hyphal elongation of test fungi* Thumbnail image of

Some inhibition of spore germination was recorded with P. granulatum, P. camembertii and C. cladosporioides while M. plumbeus was completely inhibited. In the case of P. canescens and F. solanii, on the contrary, the treatment with the enzyme solution resulted in increased spore germination (8 and 38%, respectively, after 24 h of incubation).

As for fungal growth, the large majority of the organisms tested showed inhibition of hyphal elongation. In the case of C. cladosporioides, for instance, the inhibition was 60%. In general, the enzymes of P. janthinellum affected all the species tested but did not result in the complete disruption of their cell walls; clear hyphal damages such as vacuolization, swelling and lysis were, however, evident (Fig. 1).

Figure 1.

 Effect of the treatment with 10 U ml−1 of chitinolytic enzymes from P. janthinellum P9 on the mycelial development of P. corylophilum. (a), untreated mycelium (magnification 400×); (b), treated mycelium showing shorter hyphae with vacuolization (arrows) (magnification 400×); (c) and (d), treated mycelium showing cell lysis and protoplasm extrusion (arrows) (magnification 1000×)

Combined action of the enzymes from P. janthinellum P9 and MBC

Preliminary experiments (unpublished results) showed that complete inhibition of P. canescens and F. solanii was obtained using 40 and 20 p.p.m. of the fungicide used alone, respectively, while the germination of C. cladosporioides was completely inhibited with 10 p.p.m.

Treatments with various combinations of the fungicide and the enzymes caused different levels of inhibition of germination (Table 4) but generally higher than those obtained using MCB alone. F. solanii was almost unaffected (germination of 83·3% after 24 h of incubation) while P. canescens and C. cladosporioides were more sensitive (only 32·7 and 5·3% of germination, respectively).

Table 4.   Combined effects of the fungicide MBC and the chitinolytic enzymes (CHI) of P. janthinellum P9 on spore germination of test fungi* Thumbnail image of

Also, the same samples of the experiment of Table 4 (24 h of incubation) were plated to evaluate the survival capacity of the three test fungi after each treatment: survival values, determined as cfu ml−1, are shown in Table 5.

Table 5.   Combined effects of the fungicide MBC and the chitinolytic enzymes (CHI) of Penicillium janthinellum P9 on test fungi survival* Thumbnail image of

The role of the chitinolytic enzymes in promoting the action of the fungicide is evident: when the fungicide was used alone, for instance, C. cladosporioides was completely inhibited at the concentration of 10 p.p.m. but, in combination with the enzymes (1 U ml−1), the same level of inhibition was achieved with 5 p.p.m. of fungicide. Moreover, while the growth of both P. canescens and F. solanii was only partially reduced by the fungicide used alone at the concentration of 10 p.p.m. (56·2 and 40·6%, respectively) its combination, in the same concentration, with the enzymes caused an almost complete fungal depletion.

Enhancement of freeze-dried spore germination

Freeze-dried conidia of F. solanii, P.canescens and C. cladosporioides were treated with various amounts (0·01, 0·05 and 0·5 U ml−1) of the enzymes from P. janthinellum P9. In comparison with untreated conidia, the germination of which was 31·7, 14·3 and 24·7%, respectively, treatment with chitinase resulted in generally increased germination. In particular, with only 0·05 U ml−1 of enzyme the germination ability of P. canescens, F. solanii and C. cladosporioides increased of 49·3, 58·7 and 63·4%, respectively.


Chitinolytic enzymes might be used in the bioconversion of chitin-rich materials, i.e. the wastes of the shellfish industry (Chen and Chang 1994), in the production of chito-oligosaccharides and N-acetyl-β-D-glucosamine for pharmaceutical or chemical purposes and for the food and feed industry (Sakai et al. 1990) and in biological pest control (Di Pietro et al. 1993; Lorito et al. 1994; Lorito et al. 1996). Their use, either alone or in combination with other active compounds (Lorito et al. 1994, 1996) or treatments (Fenice et al. 1999), has also been suggested to reduce some problems connected with traditional procedures that are normally expensive or cause environmental problems (Chet et al. 1993).

In this study, we have purified and partially characterized the chitinolytic activity of P. janthinellum P9 and used its enzymes (CHI1 and CHI2, two β-N-acetyl-D-hexosaminidases) in applications involving the germination and/or the development of filamentous fungi.

To this end, test fungi known as pests (F. solanii) and food-spoiling organisms (P. canescens and C. cladosporioides) were treated with the chitinolytic enzymes from P. janthinellum P9 used either alone or in combination with MBC. Growth inhibition was normally observed that, however, rather than spore germination, involved the elongation of the germinating hyphae. The damage to hyphae by the action of the exogenous chitinolytic enzymes was due probably to alterations in the physiological mechanism of hyphal elongation imbalancing the equilibrium between chitin synthesis and chitin degradation that normally results from the co-ordinated actions of endogenous chitin synthases and chitinases (Gooday 1990; Cohen 1993). With the exception of M. plumbeus, however, the chitinolytic enzymes of P. janthinellum were not sufficiently effective to inhibit the fungi completely. The chitinolytic system of P. janthinellum P9 appeared to be simple and not supported by other enzymes, e.g. endochitinases, that could lead to an aggressive role for the fungus. Penicillium species are generally saprophytic organisms that do not play any particular ecological role in mycoparasitism and, with exceptions such as P. purpurogenum (Moss 1987; Larena and Melgarejo 1993), do not have role in the biocontrol of other fungi.

Also, it has been reported that in P. oxalicum the production of β-N-acetyl-D-hexosaminidases during the autolytic process is connected with cell-wall development (Rodriguez et al. 1994). In our case, the enzymes were produced extracellularly and, apparently, no autolysis was involved. Thus, the physiological role of P. janthinellum P9 β-N-acetyl-D-hexosaminidases is possibly only at a nutritional level.

Enzyme concentrations higher than 10 U ml−1 were not tested: their cost would not be compatible with eventual applications. Thus, 1·0 and 0·2 U ml−1 were used in combination with MBC in order to reduce the amounts of this potent fungicide reaching, however, the same effects of complete fungal inhibition. In recent years, in fact, chemical pesticides and food preservatives have been strongly criticized for their undesirable and toxic effects on the environment and human health (Chet et al. 1993; Lorito et al. 1994).

The use of the enzymes from P. janthinellum allowed a 50% reduction of the amount of fungicide necessary to inhibit the growth of important pests and food-spoiling moulds completely, such as F. solanii and C. cladosporioides. Total inhibition of P. canescens was obtained using only one-quarter of the amount of fungicide that would have been necessary without the fungal chitinase. This result is particularly interesting in view of eventual applications in agriculture, such as the prevention and/or the treatment of pest diseases and the conservation of fruits, areas where new strategies are required in order to reduce or even eliminate the chemicals at present needed (Cook and Granados 1991; Lorito et al. 1994).

Finally, the chitinolytic enzymes produced by the strain P9 of P. janthinellum were used, in a preliminary experiment, to improve the germination of lyophilized spores. Preservation of fungi by lyophilization is a common procedure and methods and machinery have greatly been improved in recent years to ensure stability and a long storage period of sporulating microfungi; nevertheless, survival after treatment is still low (often less than 30%) and, often, with germination problems (Smith and Kolkowski 1996). The use of chitinolytic enzymes might be a possible strategy.


The authors are very grateful to Prof. G.W. Gooday, Department of Molecular and Cell Biology, University of Aberdeen, UK for critical reading of the manuscript. R.D.B. kindly thanks Nestec Ltd, Lausanne, Switzerland for financial support of her PhD scholarship.


  1. Present address: R & D, Barilla Alimentare S.p.A. Via Mantova 166, Parma-43100, Italy.