- Email: [email protected]
Production and purification of a chitinase from Ewingella americana, a recently described pathogen of the mushroom, Agaricus bisporus
FEMS Microbiology Letters 157 (1997) 189^194
Production and puri¢cation of a chitinase from Ewingella americana, a recently described pathogen of the mushroom, Agaricus bisporus Peter W. Inglis *, John F. Peberdy Department of Life Science, University of Nottingham, Nottingham NG7 2RD, UK
Received 16 August 1997 ; accepted 24 August 1997
Abstract
The chitinolytic properties of Ewingella americana, a recently described pathogen of the mushroom, Agaricus bisporus, are reported. E. americana was shown to produce chitinolytic activity in the absence of chitin and in the presence of glucose and Nacetylglucosamine, indicating constitutive synthesis by these strains. A single 33-kDa protein with chitinolytic activity was purified to homogeneity from culture filtrates, by hydrophobic interaction chromatography using a phenyl-group substituted matrix. This enzyme, by virtue of differential activity against chromogenic chitooligosaccharides and against dye-labelled soluble carboxymethylated chitin (CM-chitin-RBV), was demonstrated to be an endochitinase. Our data suggest this 33-kDa chitinase appeared to be the only chitinolytic enzyme produced by E. americana, strains of which do not grow using chitin as a carbon source. The significance of these findings in the context of mushroom disease is discussed. Keywords : Ewingella americana
; Chitinase, EC 3.2.1.14 ; Hydrophobic interaction chromatography; Mushroom disease
1. Introduction
Chitin is an unbranched, aggregated polymer of L1,4-linked N-acetylglucosamine residues. The complete degradation of chitin chains to their component N-acetylglucosamine units is usually achieved by a complex of enzymes acting to initially break the chains into short, random chitooligosaccharides.
* Corresponding author. Present address: Area de Controle Bioloègico (ACB), CENARGEN/EMBRAPA, SAIN Parque Rural ^ W5 Norte (final), CP 02372, Bras|èlia, D.F., CEP 70849-970, Brazil. Tel.: +55 (61) 3408575; Fax: +55 (61) 3403573; E-mail: [email protected]
Separate enzymes then act to break these oligomers down into their component monomers. This chitinolytic system has been found in organisms that themselves contain chitin, production of which is at least partially related to growth; and in other organisms, such as bacteria, plants and animals, in which the enzymes probably have nutritional, defensive or other ecological roles [1]. Chitin plays a vital role in the cell walls of many fungi [2] and has been shown to be important in the expansion and development of rigidity of the mushroom stipe [3]. Agaricus bisporus is susceptible to a browning disorder of the stipe called internal stipe necrosis [4,5]. Recently, this disease has been shown to be associated with infection by the enteric bacterium Ewingel-
0378-1097 / 97 / $17.00 ß 1997 Federation of European Microbiological Societies. Published by Elsevier Science B.V. PII S 0 3 7 8 - 1 0 9 7 ( 9 7 ) 0 0 4 7 5 - 8
FEMSLE 7903 20-11-97
190
P.W. Inglis, J.F. Peberdy / FEMS Microbiology Letters 157 (1997) 189^194
which was found to possess chitinolytic activity in screening tests [6]. The chitinolytic nature of this bacterium is therefore of considerable interest as a possible virulence factor in mushroom disease. The aims of this work were to investigate the production of chitinase by a mushroom-derived strain of E. americana and to report some of the properties of the puri¢ed enzyme. la americana
2. Materials and methods
2.1. Strains and fermentation
The mushroom-derived strain of E. americana, PI98 ([5]; deposited in the UK National Collection of Plant Pathogenic Bacteria as NCPPB 3905), was used for enzyme puri¢cation and characterisation studies. This strain has been shown to produce clearing zones on LB agar supplemented with remazol brilliant violet-stained carboxymethylated chitin (CM-chitin-RBV; [6]). In separate tests, this medium was also supplemented with 0.5 g l31 glucose or Nacetylglucosamine, in order to examine their e¡ect on chitinase production by E. americana. To provide material for puri¢cation studies, bacteria were grown in a culture volume of 1.5 l in a de¢ned broth (l31 ): lactose 2 g, mannitol 10 g, K2 HPO4 7 g, KH2 PO4 2 g, ammonium sulfate 0.5 g, 0.1 ml vitamin cocktail (l31 : biotin 2 mg; pyridoxine 0.5 g; aneurine 0.5 g; nicotinic acid 1 g; ribo£avine 1 g); and incubated at 28³C for 72 h at 200 rpm in shake £asks. Cells were removed by centrifugation at 17 000Ug for 30 min, after which the culture supernatant was passed through a sterile 0.45-Wm ¢lter and stored at 320³C until needed. To test for inducibility of chitinolytic enzymes, a 100-ml batch of the above medium was also prepared containing 5 g l31 glycol chitin. The inoculated £ask containing this medium was incubated as above and subsequently assayed for chitinase. A growth screening test was also set up whereby 5 g l31 glycol chitin solution was added to Ayres minimal medium [7]. Following streaking of bacteria on this medium, plates were incubated for 7 days at 28³C, after which, assessment was made as to the growth of E. americana with chitin as sole source of carbon.
2.2. Hydrophobic interaction chromatography (HIC)
Culture ¢ltrate (1.5 l) was adjusted to 0.5 M ammonium sulfate and pH 7.0, and applied using the manufacturer's maximum recommended £ow rates, to a 15-ml column of phenyl sepharose CL-6B (Pharmacia), while monitoring optical absorbance at 280 nm. The void from the column was collected and the column washed with 0.1 M ammonium sulfate, pH 7.0. The column was then washed with 1% (v/v) isopropanol in distilled water and the single protein peak eluted by these conditions collected. This peak, which was found to possess chitinase activity, was concentrated by polyethylene glycol dialysis (PEG-8000; Sigma). Solutions were assayed for chitinase activity; and for protein content using a dye binding assay kit (Bio-Rad [8]). Proteins were analysed by SDS-PAGE in 12.5% gels using the discontinuous system of Laemmli [9], where it was found necessary to include 30% (w/v) urea along with the reducing gel loading bu¡er to preserve solubility of the chitinolytic fraction. Gel images were captured using an Eagleeye II system (Stratagene), following Coomassie blue staining and were printed without further digital enhancement. 2.3. Chitinase assay
CM-chitin-RBV [10] (5.0 mg ml31 sterile stock solution) was used as a substrate in chitinase assays to de¢ne activity under a range of varying conditions. This assay is thought to be speci¢c for chitinases and is una¡ected by contaminating N-acetyl-LD-glucosaminidase, L-glucosidase, L-galactosidase and N-acetyl-muramidase (Lysozyme) activities [11]. Assays were performed in sterile 1.5-ml reaction tubes. The standard reaction mixture contained 400 Wl bu¡er (0.2 M phosphate, pH 6.5), 200 Wl CMchitin-RBV solution and 200 Wl enzyme solution. All solutions were kept on ice prior to mixing and incubation at 37³C for an appropriate time, usually 10 min. Reactions were then returned to ice, and 200 Wl 1 M HCl was added. The samples were then incubated on ice for a further 15 min to e¡ect complete acid precipitation of undigested CM-chitin, then centrifuged and absorbance read at 550 nm, following blanking against a control reaction where acid was
FEMSLE 7903 20-11-97
P.W. Inglis, J.F. Peberdy / FEMS Microbiology Letters 157 (1997) 189^194
added simultaneously with the other assay components. 2.4. Activity of chitinase against chromogenic soluble chitin derivatives
The soluble chitin derivatives p-nitrophenyl-N-acetyl L-D-glucosaminide [pNP-GlcNAc], p-nitrophenylL-D-N,NP-diacetylchitobiose [pNP-(GlcNAc)2 ] and pnitrophenyl-L-D-N,NP,NP-triacetylchitotriose [pNP(GlcNAc)3 ] (all Sigma) were used in a chromogenic assay for E. americana chitinase [12], where activity against these substrates was used to infer the mode of action of the enzyme [13,14]. A 125 WM solution of each derivative was prepared in 50 mM phosphate bu¡er pH 6.5. Substrate (900 Wl) was dispensed in an Eppendorf tube, and preincubated at 37³C for 5 min, after which 100 ng enzyme was added. Tubes were incubated for 20 min, after which reactions were stopped by the addition of 100 Wl 1 M NaOH. The released p-nitrophenol was estimated against a previously constructed standard curve (in the range 0.1^ 50 Wmol) by measuring absorbance at 405 nm. All determinations were performed in triplicate. 2.5. Determination of temperature and pH optima for chitinase activity
The optimal temperature for E. americana PI98 chitinase activity was determined at pH 6.5 using the standard chitinase assay, but was preincubated at the required temperature for 10 min prior to adding 200 Wl of crude cell-free ¢ltrate obtained from a 72-h E. americana culture in LB broth. Activity was measured at 25^90³C at increments of 5³C. The effect of pH on E. americana chitinase was determined by varying the composition of the standard assay bu¡er between pH values 4.0 and 10.5 (citrate-phosphate bu¡er, 0.2 M, pH 4.0^6.5; phosphate bu¡er, 0.2 M pH 6.5^8.5; glycine-NaOH, 0.2 M, pH 8.5^ 10.5) in increments of 0.5 pH units. All determinations were performed in triplicate. 2.6. E¡ect of metal ions, EDTA, SDS and urea on chitinase activity
The in£uence of several metal ions on the activity of chitinase was investigated. Metal salts at 20 mM
191
were added to the standard assay mix at pH 6.5 and at 37³C. Metal ions tested were: K , Mg2 , Ca2 , Na , Mn2 , Zn2 , Fe2 , Hg2, Co2 and Cu2 . The possible inhibitory e¡ect of EDTA, SDS and urea (all Sigma) on chitinase activity was investigated by addition of the above chemicals at both 10 mM and 20 mM to the chitinase assay mix. 3. Result and discussion
3.1. Chitinase production by E. americana
All strains of E. americana isolated from mushrooms showing symptoms of internal stipe necrosis and the type strain of E. americana ATCC 33852 were found to produce chitinolytic activity against CM-chitin-RBV when incorporated into LB agar. These strains also produced identically sized zones of degradation when this medium was supplemented with 0.5 g l31 glucose or N-acetylglucosamine (data not shown). Similarly, chitinolytic activity accumulated in the de¢ned fermentation broth which contained no form of chitin, indicating that chitinase production is, unusually among chitinolytic bacteria, probably constitutive in E. americana. Serratia marcescens strain QMB1466, for example, was shown to produce ¢ve di¡erent chitinolytic enzymes on induction by chitin [15]. No E. americana strain in this study was found to grow on minimal media containing glycol chitin as the sole source of carbon. This situation has been previously reported in clinically derived isolates [16]. 3.2. Chitinase puri¢cation
Assays conducted during hydrophobic interaction chromatography indicated that all chitinase activity in the applied culture ¢ltrate was bound under the loading conditions and retained during the washing step with 0.1 M ammonium sulfate. Washing of the bound column with 1% isopropanol produced a single, symmetrical absorbance peak (data not shown), which was found to possess chitinolytic activity. In terms of total activity against CM-chitin-RBV, the recovered fraction possessed 31% of the activity of the culture ¢ltrate. SDS-PAGE analysis of this chitinolytic fraction revealed the presence of a single
FEMSLE 7903 20-11-97
192
P.W. Inglis, J.F. Peberdy / FEMS Microbiology Letters 157 (1997) 189^194
Fig. 1. Chitinase, puri¢ed by hydrophobic interaction chromatography from E. americana strain PI98 (arrowed), resolved on a 12% SDS-PAGE gel and stained with Coomassie brilliant blue. Molecular masses are indicated by wide-range prestained markers (New England Biolabs).
protein band with an estimated molecular mass of 33 kDa (Fig. 1). Washing the column with a 2^30% (v/ v) linear gradient of isopropanol released no additional proteins possessing chitinolytic activity. 3.3. Enzymatic properties of E. americana chitinase
Tests of the puri¢ed E. americana chitinase using chromogenic derivatives of chitooligosaccharides Table 1 E¡ect of denaturants/inhibitors on the activity of puri¢ed E. americana PI98 chitinase against CM-chitin-RBV Inhibitor Relative activitya None 100 þ 0.4 Urea, 20 mM 95 þ 0.9 Urea, 10 mM 96 þ 1.4 SDS, 20 mM 53 þ 1.0 SDS, 10 mM 55 þ 1.2 EDTA, 20 mM 84 þ 1.2 EDTA, 10 mM 88 þ 2.3 a Results are the mean of three determinations with S.E.M. indicated. Activities calculated as a percentage of the maximum which was taken as 100%.
showed it to have closely similar activity against the trimeric and tetrameric derivatives (47.7 þ 6.4 Wmol pNP Wg31 protein h31 and 49.2þ 4.7 Wmol pNP Wg31 protein h31 respectively), while activity against the dimeric derivative was negligible (0.7 þ 0.04 Wmol pNP Wg31 protein h31 ). According to Haran et al. [17], this classi¢es this E. americana enzyme as an endochitinase (EC 3.2.1.14). A similar pattern of activity was obtained using crude E. americana culture ¢ltrate (data not shown). This ¢nding probably explains the failure of E. americana to grow on chitin as a carbon source, since these strains do not produce N-acetylglucosaminidase (or chitobiosidase) activity and are unable, therefore, to further break down or utilise the products of chitin digestion by endochitinase. This gives extra weight to the hypothesis that the endochitinase of E. americana can function as a colonisation factor in A. bisporus. Previous work has indicated that the chitin of the expanding mushroom stipe is not well crystallised and is particularly susceptible to the action of endochitinases [18,19]. The vulnerability of nascent chitin was also noted by Molano et al. [20] who studied wheat germ chitinase. This e¡ect is probably related to the accessibility of the chitin chains prior to micro¢bril aggregation [2]. The optimal temperature for the activity of E. americana PI98 chitinase was found to be 50³C and retained 80% of relative activity when preincubated Table 2 E¡ect of metal ions on the activity of E. americana PI98 chitinase against CM-chitin-RBV Metal Relative activitya None 100 þ 0.9 Na 98 þ 1.8 K 90 þ 1.0 Mg2 97 þ 2.0 Ca2 92 þ 2.0 Zn2 47 þ 1.5 Fe2 31 þ 1.9 Hg2 25 þ 1.5 Mn2 4 þ 0.6 Co2 2 þ 0.7 Cu2 ND a Results are the mean of triplicate tests, with S.E.M. indicated. Activities calculated as a percentage of the control treatment which was taken as 100%. ND, not determined due to interference with the assay.
FEMSLE 7903 20-11-97
P.W. Inglis, J.F. Peberdy / FEMS Microbiology Letters 157 (1997) 189^194
at this temperature for 1 h (data not shown). The enzyme was fully stable under the test conditions up to 40³C, but rapidly lost activity at temperatures above 50³C. The enzyme was virtually inactivated by preincubation at 70³C. The optimal pH for chitinase activity was found to be 6.5. Activity was also signi¢cant at pH values below this, with 50% of activity being retained at pH 4.0. At alkaline pHs, however, activity was strongly inhibited, dropping to below 20% relative activity at pH 8.5. This may re£ect signi¢cantly on the catalytic mode of action, where acidic amino acids are often signi¢cant in the active sites of enzymes mediating the acid hydrolysis of glycosidic bonds, including chitinases [21,22]. Urea, at the concentrations used, did not signi¢cantly a¡ect chitinase activity and EDTA caused only a small reduction in activity (Table 1). EDTA resistance indicates that metal ions are probably not required by the E. americana chitinase for activity. The only chitinolytic enzyme to have been demonstrated to have a requirement for a metallic co-factor is that of Phascolomyces articulosus [23]. SDS treatment caused a reduction in activity of less than 50%. This indicates that E. americana chitinase is not fully denatured by SDS alone. Certain metal ions had a drastic e¡ect on chitinase activity (Table 2). Cobalt and manganese, at 20 mM, almost abolished chitinase activity. Other divalent metal ions, with the exception of calcium and magnesium, also inhibited activity, but not to the extent of the above. Copper ions were found to a¡ect the colour of the RBV stain used in the chitinase assay and so could not be used in further tests. The particular sensitivity of the E. americana chitinase to cobalt and manganese could also be related to the importance of aspartic and glutamic acid residues in chitinases, where these amino acids have been shown to bind certain divalent cations, thereby causing inhibition [24]. Further experiments are required to fully characterise the chitinase of E. americana and to con¢rm its role in pathogenesis of internal stipe necrosis disease of A. bisporus. In particular, cloning and mutagenesis work is important in order to experimentally investigate this. Furthermore, the factors in£uencing the timing and regulation of chitinase synthesis by E. americana during colonisation of its fungal host and the location of this enzyme at the host-pathogen interface warrant further detailed investigation.
193
Acknowledgments
The authors gratefully acknowledge the ¢nancial support of the BBSRC (formally SERC) and Middlebrook Mushrooms for provision of a PhD studentship for P.W.I. We would also like to thank Dr. Cleèria Valadares-Inglis and Dr. Eddie Deane for valuable discussions during the course of this work. References [1] Flach, J., Pilet, P.E. and Jolleés, P. (1992) What's new in chitinase research? Experientia 48, 701^716. [2] Cabib, E. (1987) The synthesis and degradation of chitin. Adv. Enzymol. 59, 101. [3] Craig, G.D., Wood, D.A. and Gull, K. (1981) Chitin synthase in the stipe of Agaricus bisporus. FEMS Microbiol. Lett. 10, 43^48. [4] Richardson, P.N. (1993) Stipe necrosis of cultivated mushrooms (Agaricus bisporus) associated with a £uorescent pseudomonad. Plant Pathol. 42, 927^929. [5] Inglis, P.W. (1995) Internal Stipe Necrosis of Agaricus bisporus ^ Etiology and Molecular Genetic Studies. PhD Thesis, University of Nottingham. [6] Inglis, P.W., Burden, J.L. and Peberdy, J.F. (1996) Evidence for the association of the enteric bacterium Ewingella americana with internal stipe necrosis of Agaricus bisporus. Microbiology 142, 3253^3260. [7] Ayers, S.H., Rupp, P. and Johnson, W.T. (1919) A study of the alkali-forming bacteria in milk. U.S. Dept. Agric. Bull. 782. [8] Bradford, M.M. (1976) A rapid and sensitive method for the quanti¢cation of microgram quantities of protein using the principle of protein-dye binding. Anal. Biochem. 72, 248^254. [9] Laemmli, U.K. (1970) Cleavage of structural proteins during the assembly of bacteriophage T4. Nature 227, 680^683. [10] Wirth, S.J. and Wolf, G.A. (1990) Dye-labelled substrates for the assay and detection of chitinase and lysozyme activity. J. Microbiol. Methods 12, 197^205. [11] Saborowski, W., Buchholz, F., Vetter, R.A.H., Wirth, S.J. and Wolf, G.A. (1993) A soluble, dye-labelled chitin derivative adapted for the assay of krill chitinase. Comp. Biochem. Physiol. 105b, 673^678. [12] Roberts, W.K. and Selitrenniko¡, C.P. (1988) Plant and bacterial chitinases di¡er in antifungal activity. J. Gen. Microbiol. 134, 169^176. [13] Lorito, M., Harman, G.E., Hayes, C.K., Broardway, R.M., Tronsmo, A., Woo, S.L. and Di Pietro, A. (1993) Chitinolytic enzymes produced by Trichoderma harzianum : antifungal activity of puri¢ed endochitinase and chitobiosidase. Phytopathology 83, 302^307. [14] Chernin, L., Ismailov, Z., Haran, S. and Chet, I. (1995) Chi-
FEMSLE 7903 20-11-97
P.W. Inglis, J.F. Peberdy / FEMS Microbiology Letters 157 (1997) 189^194
194
Enterobacter agglomerans
tinolytic
antagonistic
to
fungal
plant pathogens. Appl. Environ. Microbiol. 61, 1720^726. [15] Fuchs, R.L., McPherson, S.A. and Drahos, D.J. (1986) Clon-
Serratia marcescens
ing of a
gene encoding chitinase. Appl.
P.A.D.,
Farmer
III,
J.J.,
Grimont,
F.,
Asbury,
Ewingella americana Enterobacteriaceae isolated from
M.A., Brenner, D.J. and Deval, C. (1983) gen. nov., sp. nov., a new
clinical specimens. Ann. Microbiol. (Paris) 134a, 39^52. [17] Haran, S., Schickler, H., Oppenheim, A. and Chet, I. (1995) New components of the chitinolytic system of
harzianum. [18] Mol,
P.C.
Trichoderma
Wessels,
J.G.H.
(1990)
Di¡erences
in
wall
structure between substrate hyphae and hyphae of fruit-body stipes in
Agaricus bisporus.
preformed chitin. J. Biol. Chem. 254., 4901^4907. [21] Henrissat, B. (1990) Weak sequence homologies among chiti-
Uchida, M. and Tanaka, H. (1993) Identi¢cation of glutamic acid 204 and aspartic acid 200 in chitinase A1 of
Bacillus
circulans
activity.
WL-12
as
[19] Mol, P.C., Vermeulen, C.A. and Wessels, J.G.H. (1990) Dif-
Agaricus bisporus
may be
based on a unique wall structure. Mycol. Res. 94, 480^488.
essential
residues
for
chitinase
M.S.
(1992)
J. Biol. Chem. 268, 18567^18572. R.
and
Manocha,
Cytosolic
and membrane-bound chitinases of two mucoraceous fungi : a comparative study. Can. J. Microbiol. 38, 331^338. [24] Milewski,
Mycol. Res. 94, 472^479.
fuse extension of hyphae in stipes of
Anal. 3, 523^526. [22] Watanabe, T., Kobori, K., Miyashita, K., Fujii, T., Sakai, H.,
[23] Balasubramanian,
Mycol. Res. 99, 441^446. and
An endochitinase from wheat germ. Activity on nascent and
nases detected by clustering analysis. Protein Sequences Data
Environ. Microbiol. 51, 504^509. [16] Grimont,
[20] Molano, J., Polacheck, U., Duran, A. and Cabib, E. (1979)
S.,
O'Donnell,
R.W.
and
Gooday,
G.W.
Chemical modi¢cation studies of the active centre of
albicans
chitinase and its inhibition by allosamidin. J. Gen.
Microbiol. 138, 2545^2550.
FEMSLE 7903 20-11-97
(1992)
Candida
Production and puri¢cation of a chitinase from Ewingella americana, a recently described pathogen of the mushroom, Agaricus bisporus Peter W. Inglis *, John F. Peberdy Department of Life Science, University of Nottingham, Nottingham NG7 2RD, UK
Received 16 August 1997 ; accepted 24 August 1997
Abstract
The chitinolytic properties of Ewingella americana, a recently described pathogen of the mushroom, Agaricus bisporus, are reported. E. americana was shown to produce chitinolytic activity in the absence of chitin and in the presence of glucose and Nacetylglucosamine, indicating constitutive synthesis by these strains. A single 33-kDa protein with chitinolytic activity was purified to homogeneity from culture filtrates, by hydrophobic interaction chromatography using a phenyl-group substituted matrix. This enzyme, by virtue of differential activity against chromogenic chitooligosaccharides and against dye-labelled soluble carboxymethylated chitin (CM-chitin-RBV), was demonstrated to be an endochitinase. Our data suggest this 33-kDa chitinase appeared to be the only chitinolytic enzyme produced by E. americana, strains of which do not grow using chitin as a carbon source. The significance of these findings in the context of mushroom disease is discussed. Keywords : Ewingella americana
; Chitinase, EC 3.2.1.14 ; Hydrophobic interaction chromatography; Mushroom disease
1. Introduction
Chitin is an unbranched, aggregated polymer of L1,4-linked N-acetylglucosamine residues. The complete degradation of chitin chains to their component N-acetylglucosamine units is usually achieved by a complex of enzymes acting to initially break the chains into short, random chitooligosaccharides.
* Corresponding author. Present address: Area de Controle Bioloègico (ACB), CENARGEN/EMBRAPA, SAIN Parque Rural ^ W5 Norte (final), CP 02372, Bras|èlia, D.F., CEP 70849-970, Brazil. Tel.: +55 (61) 3408575; Fax: +55 (61) 3403573; E-mail: [email protected]
Separate enzymes then act to break these oligomers down into their component monomers. This chitinolytic system has been found in organisms that themselves contain chitin, production of which is at least partially related to growth; and in other organisms, such as bacteria, plants and animals, in which the enzymes probably have nutritional, defensive or other ecological roles [1]. Chitin plays a vital role in the cell walls of many fungi [2] and has been shown to be important in the expansion and development of rigidity of the mushroom stipe [3]. Agaricus bisporus is susceptible to a browning disorder of the stipe called internal stipe necrosis [4,5]. Recently, this disease has been shown to be associated with infection by the enteric bacterium Ewingel-
0378-1097 / 97 / $17.00 ß 1997 Federation of European Microbiological Societies. Published by Elsevier Science B.V. PII S 0 3 7 8 - 1 0 9 7 ( 9 7 ) 0 0 4 7 5 - 8
FEMSLE 7903 20-11-97
190
P.W. Inglis, J.F. Peberdy / FEMS Microbiology Letters 157 (1997) 189^194
which was found to possess chitinolytic activity in screening tests [6]. The chitinolytic nature of this bacterium is therefore of considerable interest as a possible virulence factor in mushroom disease. The aims of this work were to investigate the production of chitinase by a mushroom-derived strain of E. americana and to report some of the properties of the puri¢ed enzyme. la americana
2. Materials and methods
2.1. Strains and fermentation
The mushroom-derived strain of E. americana, PI98 ([5]; deposited in the UK National Collection of Plant Pathogenic Bacteria as NCPPB 3905), was used for enzyme puri¢cation and characterisation studies. This strain has been shown to produce clearing zones on LB agar supplemented with remazol brilliant violet-stained carboxymethylated chitin (CM-chitin-RBV; [6]). In separate tests, this medium was also supplemented with 0.5 g l31 glucose or Nacetylglucosamine, in order to examine their e¡ect on chitinase production by E. americana. To provide material for puri¢cation studies, bacteria were grown in a culture volume of 1.5 l in a de¢ned broth (l31 ): lactose 2 g, mannitol 10 g, K2 HPO4 7 g, KH2 PO4 2 g, ammonium sulfate 0.5 g, 0.1 ml vitamin cocktail (l31 : biotin 2 mg; pyridoxine 0.5 g; aneurine 0.5 g; nicotinic acid 1 g; ribo£avine 1 g); and incubated at 28³C for 72 h at 200 rpm in shake £asks. Cells were removed by centrifugation at 17 000Ug for 30 min, after which the culture supernatant was passed through a sterile 0.45-Wm ¢lter and stored at 320³C until needed. To test for inducibility of chitinolytic enzymes, a 100-ml batch of the above medium was also prepared containing 5 g l31 glycol chitin. The inoculated £ask containing this medium was incubated as above and subsequently assayed for chitinase. A growth screening test was also set up whereby 5 g l31 glycol chitin solution was added to Ayres minimal medium [7]. Following streaking of bacteria on this medium, plates were incubated for 7 days at 28³C, after which, assessment was made as to the growth of E. americana with chitin as sole source of carbon.
2.2. Hydrophobic interaction chromatography (HIC)
Culture ¢ltrate (1.5 l) was adjusted to 0.5 M ammonium sulfate and pH 7.0, and applied using the manufacturer's maximum recommended £ow rates, to a 15-ml column of phenyl sepharose CL-6B (Pharmacia), while monitoring optical absorbance at 280 nm. The void from the column was collected and the column washed with 0.1 M ammonium sulfate, pH 7.0. The column was then washed with 1% (v/v) isopropanol in distilled water and the single protein peak eluted by these conditions collected. This peak, which was found to possess chitinase activity, was concentrated by polyethylene glycol dialysis (PEG-8000; Sigma). Solutions were assayed for chitinase activity; and for protein content using a dye binding assay kit (Bio-Rad [8]). Proteins were analysed by SDS-PAGE in 12.5% gels using the discontinuous system of Laemmli [9], where it was found necessary to include 30% (w/v) urea along with the reducing gel loading bu¡er to preserve solubility of the chitinolytic fraction. Gel images were captured using an Eagleeye II system (Stratagene), following Coomassie blue staining and were printed without further digital enhancement. 2.3. Chitinase assay
CM-chitin-RBV [10] (5.0 mg ml31 sterile stock solution) was used as a substrate in chitinase assays to de¢ne activity under a range of varying conditions. This assay is thought to be speci¢c for chitinases and is una¡ected by contaminating N-acetyl-LD-glucosaminidase, L-glucosidase, L-galactosidase and N-acetyl-muramidase (Lysozyme) activities [11]. Assays were performed in sterile 1.5-ml reaction tubes. The standard reaction mixture contained 400 Wl bu¡er (0.2 M phosphate, pH 6.5), 200 Wl CMchitin-RBV solution and 200 Wl enzyme solution. All solutions were kept on ice prior to mixing and incubation at 37³C for an appropriate time, usually 10 min. Reactions were then returned to ice, and 200 Wl 1 M HCl was added. The samples were then incubated on ice for a further 15 min to e¡ect complete acid precipitation of undigested CM-chitin, then centrifuged and absorbance read at 550 nm, following blanking against a control reaction where acid was
FEMSLE 7903 20-11-97
P.W. Inglis, J.F. Peberdy / FEMS Microbiology Letters 157 (1997) 189^194
added simultaneously with the other assay components. 2.4. Activity of chitinase against chromogenic soluble chitin derivatives
The soluble chitin derivatives p-nitrophenyl-N-acetyl L-D-glucosaminide [pNP-GlcNAc], p-nitrophenylL-D-N,NP-diacetylchitobiose [pNP-(GlcNAc)2 ] and pnitrophenyl-L-D-N,NP,NP-triacetylchitotriose [pNP(GlcNAc)3 ] (all Sigma) were used in a chromogenic assay for E. americana chitinase [12], where activity against these substrates was used to infer the mode of action of the enzyme [13,14]. A 125 WM solution of each derivative was prepared in 50 mM phosphate bu¡er pH 6.5. Substrate (900 Wl) was dispensed in an Eppendorf tube, and preincubated at 37³C for 5 min, after which 100 ng enzyme was added. Tubes were incubated for 20 min, after which reactions were stopped by the addition of 100 Wl 1 M NaOH. The released p-nitrophenol was estimated against a previously constructed standard curve (in the range 0.1^ 50 Wmol) by measuring absorbance at 405 nm. All determinations were performed in triplicate. 2.5. Determination of temperature and pH optima for chitinase activity
The optimal temperature for E. americana PI98 chitinase activity was determined at pH 6.5 using the standard chitinase assay, but was preincubated at the required temperature for 10 min prior to adding 200 Wl of crude cell-free ¢ltrate obtained from a 72-h E. americana culture in LB broth. Activity was measured at 25^90³C at increments of 5³C. The effect of pH on E. americana chitinase was determined by varying the composition of the standard assay bu¡er between pH values 4.0 and 10.5 (citrate-phosphate bu¡er, 0.2 M, pH 4.0^6.5; phosphate bu¡er, 0.2 M pH 6.5^8.5; glycine-NaOH, 0.2 M, pH 8.5^ 10.5) in increments of 0.5 pH units. All determinations were performed in triplicate. 2.6. E¡ect of metal ions, EDTA, SDS and urea on chitinase activity
The in£uence of several metal ions on the activity of chitinase was investigated. Metal salts at 20 mM
191
were added to the standard assay mix at pH 6.5 and at 37³C. Metal ions tested were: K , Mg2 , Ca2 , Na , Mn2 , Zn2 , Fe2 , Hg2, Co2 and Cu2 . The possible inhibitory e¡ect of EDTA, SDS and urea (all Sigma) on chitinase activity was investigated by addition of the above chemicals at both 10 mM and 20 mM to the chitinase assay mix. 3. Result and discussion
3.1. Chitinase production by E. americana
All strains of E. americana isolated from mushrooms showing symptoms of internal stipe necrosis and the type strain of E. americana ATCC 33852 were found to produce chitinolytic activity against CM-chitin-RBV when incorporated into LB agar. These strains also produced identically sized zones of degradation when this medium was supplemented with 0.5 g l31 glucose or N-acetylglucosamine (data not shown). Similarly, chitinolytic activity accumulated in the de¢ned fermentation broth which contained no form of chitin, indicating that chitinase production is, unusually among chitinolytic bacteria, probably constitutive in E. americana. Serratia marcescens strain QMB1466, for example, was shown to produce ¢ve di¡erent chitinolytic enzymes on induction by chitin [15]. No E. americana strain in this study was found to grow on minimal media containing glycol chitin as the sole source of carbon. This situation has been previously reported in clinically derived isolates [16]. 3.2. Chitinase puri¢cation
Assays conducted during hydrophobic interaction chromatography indicated that all chitinase activity in the applied culture ¢ltrate was bound under the loading conditions and retained during the washing step with 0.1 M ammonium sulfate. Washing of the bound column with 1% isopropanol produced a single, symmetrical absorbance peak (data not shown), which was found to possess chitinolytic activity. In terms of total activity against CM-chitin-RBV, the recovered fraction possessed 31% of the activity of the culture ¢ltrate. SDS-PAGE analysis of this chitinolytic fraction revealed the presence of a single
FEMSLE 7903 20-11-97
192
P.W. Inglis, J.F. Peberdy / FEMS Microbiology Letters 157 (1997) 189^194
Fig. 1. Chitinase, puri¢ed by hydrophobic interaction chromatography from E. americana strain PI98 (arrowed), resolved on a 12% SDS-PAGE gel and stained with Coomassie brilliant blue. Molecular masses are indicated by wide-range prestained markers (New England Biolabs).
protein band with an estimated molecular mass of 33 kDa (Fig. 1). Washing the column with a 2^30% (v/ v) linear gradient of isopropanol released no additional proteins possessing chitinolytic activity. 3.3. Enzymatic properties of E. americana chitinase
Tests of the puri¢ed E. americana chitinase using chromogenic derivatives of chitooligosaccharides Table 1 E¡ect of denaturants/inhibitors on the activity of puri¢ed E. americana PI98 chitinase against CM-chitin-RBV Inhibitor Relative activitya None 100 þ 0.4 Urea, 20 mM 95 þ 0.9 Urea, 10 mM 96 þ 1.4 SDS, 20 mM 53 þ 1.0 SDS, 10 mM 55 þ 1.2 EDTA, 20 mM 84 þ 1.2 EDTA, 10 mM 88 þ 2.3 a Results are the mean of three determinations with S.E.M. indicated. Activities calculated as a percentage of the maximum which was taken as 100%.
showed it to have closely similar activity against the trimeric and tetrameric derivatives (47.7 þ 6.4 Wmol pNP Wg31 protein h31 and 49.2þ 4.7 Wmol pNP Wg31 protein h31 respectively), while activity against the dimeric derivative was negligible (0.7 þ 0.04 Wmol pNP Wg31 protein h31 ). According to Haran et al. [17], this classi¢es this E. americana enzyme as an endochitinase (EC 3.2.1.14). A similar pattern of activity was obtained using crude E. americana culture ¢ltrate (data not shown). This ¢nding probably explains the failure of E. americana to grow on chitin as a carbon source, since these strains do not produce N-acetylglucosaminidase (or chitobiosidase) activity and are unable, therefore, to further break down or utilise the products of chitin digestion by endochitinase. This gives extra weight to the hypothesis that the endochitinase of E. americana can function as a colonisation factor in A. bisporus. Previous work has indicated that the chitin of the expanding mushroom stipe is not well crystallised and is particularly susceptible to the action of endochitinases [18,19]. The vulnerability of nascent chitin was also noted by Molano et al. [20] who studied wheat germ chitinase. This e¡ect is probably related to the accessibility of the chitin chains prior to micro¢bril aggregation [2]. The optimal temperature for the activity of E. americana PI98 chitinase was found to be 50³C and retained 80% of relative activity when preincubated Table 2 E¡ect of metal ions on the activity of E. americana PI98 chitinase against CM-chitin-RBV Metal Relative activitya None 100 þ 0.9 Na 98 þ 1.8 K 90 þ 1.0 Mg2 97 þ 2.0 Ca2 92 þ 2.0 Zn2 47 þ 1.5 Fe2 31 þ 1.9 Hg2 25 þ 1.5 Mn2 4 þ 0.6 Co2 2 þ 0.7 Cu2 ND a Results are the mean of triplicate tests, with S.E.M. indicated. Activities calculated as a percentage of the control treatment which was taken as 100%. ND, not determined due to interference with the assay.
FEMSLE 7903 20-11-97
P.W. Inglis, J.F. Peberdy / FEMS Microbiology Letters 157 (1997) 189^194
at this temperature for 1 h (data not shown). The enzyme was fully stable under the test conditions up to 40³C, but rapidly lost activity at temperatures above 50³C. The enzyme was virtually inactivated by preincubation at 70³C. The optimal pH for chitinase activity was found to be 6.5. Activity was also signi¢cant at pH values below this, with 50% of activity being retained at pH 4.0. At alkaline pHs, however, activity was strongly inhibited, dropping to below 20% relative activity at pH 8.5. This may re£ect signi¢cantly on the catalytic mode of action, where acidic amino acids are often signi¢cant in the active sites of enzymes mediating the acid hydrolysis of glycosidic bonds, including chitinases [21,22]. Urea, at the concentrations used, did not signi¢cantly a¡ect chitinase activity and EDTA caused only a small reduction in activity (Table 1). EDTA resistance indicates that metal ions are probably not required by the E. americana chitinase for activity. The only chitinolytic enzyme to have been demonstrated to have a requirement for a metallic co-factor is that of Phascolomyces articulosus [23]. SDS treatment caused a reduction in activity of less than 50%. This indicates that E. americana chitinase is not fully denatured by SDS alone. Certain metal ions had a drastic e¡ect on chitinase activity (Table 2). Cobalt and manganese, at 20 mM, almost abolished chitinase activity. Other divalent metal ions, with the exception of calcium and magnesium, also inhibited activity, but not to the extent of the above. Copper ions were found to a¡ect the colour of the RBV stain used in the chitinase assay and so could not be used in further tests. The particular sensitivity of the E. americana chitinase to cobalt and manganese could also be related to the importance of aspartic and glutamic acid residues in chitinases, where these amino acids have been shown to bind certain divalent cations, thereby causing inhibition [24]. Further experiments are required to fully characterise the chitinase of E. americana and to con¢rm its role in pathogenesis of internal stipe necrosis disease of A. bisporus. In particular, cloning and mutagenesis work is important in order to experimentally investigate this. Furthermore, the factors in£uencing the timing and regulation of chitinase synthesis by E. americana during colonisation of its fungal host and the location of this enzyme at the host-pathogen interface warrant further detailed investigation.
193
Acknowledgments
The authors gratefully acknowledge the ¢nancial support of the BBSRC (formally SERC) and Middlebrook Mushrooms for provision of a PhD studentship for P.W.I. We would also like to thank Dr. Cleèria Valadares-Inglis and Dr. Eddie Deane for valuable discussions during the course of this work. References [1] Flach, J., Pilet, P.E. and Jolleés, P. (1992) What's new in chitinase research? Experientia 48, 701^716. [2] Cabib, E. (1987) The synthesis and degradation of chitin. Adv. Enzymol. 59, 101. [3] Craig, G.D., Wood, D.A. and Gull, K. (1981) Chitin synthase in the stipe of Agaricus bisporus. FEMS Microbiol. Lett. 10, 43^48. [4] Richardson, P.N. (1993) Stipe necrosis of cultivated mushrooms (Agaricus bisporus) associated with a £uorescent pseudomonad. Plant Pathol. 42, 927^929. [5] Inglis, P.W. (1995) Internal Stipe Necrosis of Agaricus bisporus ^ Etiology and Molecular Genetic Studies. PhD Thesis, University of Nottingham. [6] Inglis, P.W., Burden, J.L. and Peberdy, J.F. (1996) Evidence for the association of the enteric bacterium Ewingella americana with internal stipe necrosis of Agaricus bisporus. Microbiology 142, 3253^3260. [7] Ayers, S.H., Rupp, P. and Johnson, W.T. (1919) A study of the alkali-forming bacteria in milk. U.S. Dept. Agric. Bull. 782. [8] Bradford, M.M. (1976) A rapid and sensitive method for the quanti¢cation of microgram quantities of protein using the principle of protein-dye binding. Anal. Biochem. 72, 248^254. [9] Laemmli, U.K. (1970) Cleavage of structural proteins during the assembly of bacteriophage T4. Nature 227, 680^683. [10] Wirth, S.J. and Wolf, G.A. (1990) Dye-labelled substrates for the assay and detection of chitinase and lysozyme activity. J. Microbiol. Methods 12, 197^205. [11] Saborowski, W., Buchholz, F., Vetter, R.A.H., Wirth, S.J. and Wolf, G.A. (1993) A soluble, dye-labelled chitin derivative adapted for the assay of krill chitinase. Comp. Biochem. Physiol. 105b, 673^678. [12] Roberts, W.K. and Selitrenniko¡, C.P. (1988) Plant and bacterial chitinases di¡er in antifungal activity. J. Gen. Microbiol. 134, 169^176. [13] Lorito, M., Harman, G.E., Hayes, C.K., Broardway, R.M., Tronsmo, A., Woo, S.L. and Di Pietro, A. (1993) Chitinolytic enzymes produced by Trichoderma harzianum : antifungal activity of puri¢ed endochitinase and chitobiosidase. Phytopathology 83, 302^307. [14] Chernin, L., Ismailov, Z., Haran, S. and Chet, I. (1995) Chi-
FEMSLE 7903 20-11-97
P.W. Inglis, J.F. Peberdy / FEMS Microbiology Letters 157 (1997) 189^194
194
Enterobacter agglomerans
tinolytic
antagonistic
to
fungal
plant pathogens. Appl. Environ. Microbiol. 61, 1720^726. [15] Fuchs, R.L., McPherson, S.A. and Drahos, D.J. (1986) Clon-
Serratia marcescens
ing of a
gene encoding chitinase. Appl.
P.A.D.,
Farmer
III,
J.J.,
Grimont,
F.,
Asbury,
Ewingella americana Enterobacteriaceae isolated from
M.A., Brenner, D.J. and Deval, C. (1983) gen. nov., sp. nov., a new
clinical specimens. Ann. Microbiol. (Paris) 134a, 39^52. [17] Haran, S., Schickler, H., Oppenheim, A. and Chet, I. (1995) New components of the chitinolytic system of
harzianum. [18] Mol,
P.C.
Trichoderma
Wessels,
J.G.H.
(1990)
Di¡erences
in
wall
structure between substrate hyphae and hyphae of fruit-body stipes in
Agaricus bisporus.
preformed chitin. J. Biol. Chem. 254., 4901^4907. [21] Henrissat, B. (1990) Weak sequence homologies among chiti-
Uchida, M. and Tanaka, H. (1993) Identi¢cation of glutamic acid 204 and aspartic acid 200 in chitinase A1 of
Bacillus
circulans
activity.
WL-12
as
[19] Mol, P.C., Vermeulen, C.A. and Wessels, J.G.H. (1990) Dif-
Agaricus bisporus
may be
based on a unique wall structure. Mycol. Res. 94, 480^488.
essential
residues
for
chitinase
M.S.
(1992)
J. Biol. Chem. 268, 18567^18572. R.
and
Manocha,
Cytosolic
and membrane-bound chitinases of two mucoraceous fungi : a comparative study. Can. J. Microbiol. 38, 331^338. [24] Milewski,
Mycol. Res. 94, 472^479.
fuse extension of hyphae in stipes of
Anal. 3, 523^526. [22] Watanabe, T., Kobori, K., Miyashita, K., Fujii, T., Sakai, H.,
[23] Balasubramanian,
Mycol. Res. 99, 441^446. and
An endochitinase from wheat germ. Activity on nascent and
nases detected by clustering analysis. Protein Sequences Data
Environ. Microbiol. 51, 504^509. [16] Grimont,
[20] Molano, J., Polacheck, U., Duran, A. and Cabib, E. (1979)
S.,
O'Donnell,
R.W.
and
Gooday,
G.W.
Chemical modi¢cation studies of the active centre of
albicans
chitinase and its inhibition by allosamidin. J. Gen.
Microbiol. 138, 2545^2550.
FEMSLE 7903 20-11-97
(1992)
Candida