PlantCell Reports

Plant Cell Reports (1995) 15:222-226

@) Springer-Verlag1995

A Streptomyces chitosanase is active in transgenic tobacco Souad El Quakfaoui 1, Claude Potvin 1, Ryszard Brzezinski 2, and Alain Asselin a D6partement de phytologie, Facult6 des Sciences de l'Agriculture et de l'Alimentation, Universit6 Laval, Qu6bec, Canada GIK 7P4 = Groupe de Recherche en Biologic des Actinomyc6tes, D6parternent de Biologic, Facult6 des Sciences, Universit6 de Sherbrooke, Sherbrooke, Qu+bec, Canada JtK 2R1 Received 27 February 1995/Revised version received 18 May 1995 - Communicated by S. Gleddie

Summary. Growth inhibition towards R h i z o p u s nigricans, Fusarium oxysporum f. sp. radicis-lycopersici, Verticillium albo-atrum and Pythium ultimum was observed in vitro using a purified chitosanase from an actinomycete, Streptomyces sp. strain N174. The corresponding gene, with its own signal peptide, was inserted into pBI121.7 shuttle vector to transform tobacco. Transgenic plants were analysed for chitos,'mase activity by a sodium dodecyl sulfate-polyacryl,'unide gel elecuophoresis assay. Two major and one minor active electa'ophoretie forms were detected in transgenic tobacco. Some chitosanases were recovered not only in leaf homogenates but also in leaf intercellul,'u" fluid exuacts. One chitosanase electrophoretic form migrated very closely to the purified Streptomyces mature protein while the others corresponded to molecules of higher molecular mass. The N-terminus sequence was determined for one of the three chitosanase forms. It exhibited a different signal peptide cleavage site when compared to the mature chitosanase from Streptomyces. This is the first report on the expression of an active chitosanase gene with antimicrobial potential in plants.

Abbreviations: aa, amino acid; CIP, calf intestinal phosphatase; CM, cm'boxymethyi; GUS, 13-glucuronidase; IF, intercellulm" fluid; MS, Murashige and Skoog; PAGE, polyacryl~nide gel electrophoresis; PR, pathogenesisrelatcd; PVDF, polyvinylidene difluoride; SP, signal peptide.

Introduction Chitosan, a polymer of fi-l,4-D-gluc0s~unine, represents a structural component in the cell walls of zygomycetes occurring mostly in association with chitin, a linear homopolymer of g-l,4-N-acetyl-D-glucosarnine residues (Araki and ito, 1988). Chitosan is also present in smaller mnounts in cell walls of other cl&sses of fungi (Davis and Eveleigh, 1984). Commercially available chitosan is usually obtained by alkaline deacetylation of shellfish chitin. Chitosan and its derivatives have potenlial industrial, biomedical and agricultm-al zpplications. Diverse

Correspondence to: A. Asselin

microbial chitosanases, which are glycosyl hydrolases degrading chitosan, have been purified and characterized (Monaghan, 1973; Zikakis, 1984; Pelletier and Sygusch, 1990; Masson et at. 1993). Recently, by using a PAGE assay, chitosanases were detected in plant IF extracts and reported as PR proteins (Grenier and Asselin, 1990). Like chitinases and g-l,3glucanases, chitosanases might be iuvolved as defeuse proteins by degrading the cell walls of fungal pathogens (El Ouakfaoui and Asselin, 1992 a, b; Dumas-Gaudot et al. 1992; Cuero and Osuji, 1993; Shm'ma et at. 1993). Several eukaryotic chitinases (Broglie et al. 1991; Neuhaus et al. 1991; Vierheilig et al. 1993), hen lysozyme (Trudel et al. 1995) and 13-1,3-glucanases (Sela-Buuflage et al. 1993) have been expressed constitutively in transgenic plants for conferring increased resistance to various pathogens. Among prokaryotic hydrolases, a chitinase from Serratia marescens (Howie et al. 1992) was also used for this purpose. However, no chitosanase has been expressed in transgenic plm~ts until now. Chitosanases were found widely disn'ibuted in higher and lower plants (El Ouakfaoui and Asselin, 1992 a). Interestingly, some chitosauasc molecular forms were specific to developmental stages and/or organs (El Ouakfaoui and Asselin, 1992 b). Up to now, no mnino acid or gene sequence has been reported for plato chitosanases. Two chitinase/chitosanase isoforms have recently been purified from Citrus (Ossw,qld et al. 1993). To our knowledge, only three microbial chitosanase gene sequences have been determined: one from Bacillus circulans MH-K1 (Ando et al. 1992); one from Streptomyces sp. strain N174 (Masson et al. 1994) and one from Nocardioides sp. N106 (R. Brzezinski, submitted for publication; GSDB accession number L40408). It was overexpressed in Streptomyces lividans and purified as an extracellulm mature endochitosanase of 238 mnino acids (Fink et al. 1991; Boucher et al. 1992). This chitosanase specifically hydrolysed chitosan (without any activity toward chitin or CM-cellulose). In this report, the antifungal potential of the Streptomyces sp. strain N174 chitosanase was investigated and the expression of this aetinomycete gene in transgenic tobacco was analysed.

223

Materials and Methods

Fungal growth inhibition assay. Purified Streptomyces sp. strain N174 chitosanase produced by the recombinant strain Streptomyees lividans [pRL270] (Masson et al. 1993) was tested against Rhizopus nigrieans, Fusarium oxysporum f. sp. radicis-lycopersici (FORL), VerticiUium aIbo-atrum and Pythium ultimum for growth inhibition on 10% (w/v) polyacrylamide slab gels previously immersed in 500 mL Yeast Peptone Dextrose (YPD) broth and autoclaved for 20 rain at 12I~ The sterile gel slabs were then placed in Petri dishes (150 mm diameter) and air dried for 5 rain under sterile air flow. The fungi were inoculated in the center of gel slabs and radial growth was allowed at room temperature for l day (R. nigricans, P. ultimum), 2 days (FORL) and 7 days (V. alboatrum). Purified chitosanase was applied in front of the growing fungi. The plates were kept at room temperature for fungal growth.

Chitosanase construct. The chitosanase gene from Streptomyees sp. strain N174 was previously cloned and sequenced (Fink et al. 1991; Masson et al. 1994). For this study, the clone pRL270 (Masson et al. 1993) was digested with HindIII/XhoI generating a 1607 bp fragment containing the chitosanase gene. This fragment was subcloned in pBluescript SK(-) (Stratagene) digested with the same restriction enzymes, resulting in pSN9. This last clone was used to engineer pSN9C, the chitosanase gene with its own 40aa signal peptide. Clone pSN9 was digested by HindIII and NdeI in order to eliminate the non-coding sequence upstream the chitosanase gene. After end gap filling-in and circularization, a HindIII site was regenerated next to the ATG of the chitosanase gene. Following E. coli DH5aF' (BRL) transformation, one clone identified as pSN9C was selected and confirmed by HindIIIlXhoI digestions.

Plasmid constructs for plant transformation.

The plant transformation vector pBI121.7 originated from conunercial pBI121 (Clontech) in which the GUS gene has been removed (Trudel et al. 1992). The EcoRI/ClaI fragment from pSN9C and the dephosphorylated (CIP) BamHI linearized vector were treated with T4 DNA polymerase and dNTP to blunt the ends, before overnight ligation. A BamHI restriction site was regenerated downstream the chitosanase gene. Clones obtained after E. coli DH5aF' transformation were analysed by Sinai, BamHI/EcoRI and HindIIIlEcoRI digestions to determine insert size and orientation. The plasmid containing the chitosanase gene under the control of the cauliflower mosaic virus (CaMV) 35S promoter was identified as pBI9C. A clone with the chitosanase gene in reverse orientation (pBI9Cr) was a control.

Tobacco transformation. Constructs pBI9C and pBI9Cr were transferred in Agrobacterium tumefaciens LBA4404 harboring Ti plasmid pAL4404 using standard triparental mating procedures (Horsch et al. 1985) with the pRK2013 conjugative plasmid in E. coli HB101. Tobacco (Nicotiana tabacum L. cv. Xanthi-nc) leaf pieces were inoculated with transformed A. tumefilciens using nurse culture plates with MS-104 medium (Horsch et al. 1985). Resulting kanamycin-resistant calli were propagated in vitro and regenerated plantlets were transferred to soil and grown under normal greenhouse conditions (Trudel et al. 1992).

Chitosanase PAGE assay. Leaf homogenates [1:3 (w/v)] prepared in 2.5% (w/v) SDS, 15% (w/v) sucrose, 25 mM sodium phosphate buffer (pH 7.0) and 1% (v/v) g-mercaptoethanol, were heated at 100~ for 3 rain and clarified at 10,000g for 10 rain at 10~ Leaf IF extracts were prepared by performing

vacuum infiltration of leaf pieces in 50 mM sodium phosphate buffer (pH 6.5) and low speed centrifugation (Parent and Asselin, 1984). Protein concentration and detection of chitosanase activity after 15% (w/v) PAGE under denaturing (SDS) conditions were performed as previously described (Grenier and Asselin, 1990).

Chitosanase lysoplate assay. Leaf IF extracts were tested for chitnsanase activity in 1% (w/v) agarose slab gels containing 0.05% (w/v) of chitosan (Fluka Chemical Corp.) as substrate. Chitosan was solubilized in 20% (v/v) acetic acid overnight and the pH was adjusted to 5.0 with 1 N NaOH before it was added to the agarose gel. The gel contained 10 mM sodium acetate buffer (pH 5.0) and 1% (v/v) purified Triton X-100 (Grenier and Asselin, 1990). Gel slabs (30 mL) were poured on 15 x 10 cm glass plates. Wells of 3.5 mm in diameter contained 10 laL samples. Incubation was for 16 h at 37~ in closed boxes. Lysis zones were visualized by transparency against a black background or by Coomassie Blue R-250 staining (Grenier and Asselin, 1990). Protein purification for microsequencing. Chitosanase forms were purified by successive native and denaturing PAGE according to Trudel and Asselin (1994). Leaf IF extracts were prepared from pBI9C R 2 progeny transgenic tobacco. The IF was concentrated (10X) and a first electrophoretic separation was performed using eight preparative native 15% (w/v) polyacrylamide gels (0.75 mm thick) run at pH 4.3. After Coomassie Blue staining, ehitosanase bands were excised in the denaturing buffer (Trudel and Asselin, 1994). Gel slices were washed in distilled water and equilibrated. The eight gel slices were subjected to preparative SDS-PAGE using four 15% (w/v) polyacrylamide (1 iron thick) gels. After Coomassie Blue staining, the three chitosanase bands were excised and each set of bands was stacked on top of one denaturing 15% (w/v) polyacrylamide gel (1.5 man thick). The proteins (2-3 lag each) were then electroblotted on PVDF membrane (Trudel and Asselin, 1994). Protein N-terminus sequencing was undertaken by automated Edman degradation using Applied Biosystenrs Sequencer 473A (Service de S~quence de Peptides de rest da Quebec, Centre Hospitalier de rUniversitg Laval, Ste-Foy, Canada). The purified chitosanase from Streptomyces sp. strain N174 was subjected to the same sequential PAGE steps as a control.

Results

Fungal growth inhibition assay Growth inhibition was observed with 2.5 t.tg of purified Streptomyces sp. strain N174 chitosanase towards Rhizopus nigricans and with 25 gg towm-ds Verticillium albo-atrum, Fusarium oxysporum f. sp. radicis-lycopersici and Pythium ultimum (Fig. 1). No inhibition was observed with the chitosanase buffer ,alone.

Chitosanase analysis in transgenic tobacco Using the PAGE assay, chitosanase activity was already detected in transformed R 0 callus homogenates (data not shown). Subsequently, chitosanase activity was tested in leaves from the R0 tmnsfol'rnants that had been grown in soil for two weeks. Leaves of the R1 and R2 (selfpollinated) tobacco progeny were also tested for chitosanase activity. No unusual phenotype was observed with

224 transgenic tobacco plants expressing chitosanase. Chitosanase activity was detected after SDS-PAGE in leaf homogenates (R 2 progeny) as three bands (a, b, c) in pBI9C transgenic tobacco (Fig. 2). Only one molecular form (a) migrated very closely to the purified mature chitosanase from Streptomyces sp. strain N174, at approximately 29.5 kD (Fig. 2, lane CSN). The other forms (b and c) migrated as proteins of higher molecular mass (approximately 30.5 and 31.5 kD for b and c, respectively) (Fig. 2). The major folxns were (a) and (b) in leaf homogenates while in leaf IF extracts, the (b) form was the major one (Fig. 2). No activity was detected in leaf homogenate or IF extracts of pBI9Cr transgenic tobacco and untransformed tobacco (dam not shown). Chitosanase renaturation was not adversely affected by the use of Bmercaptoethanol. Moreover, this reducing agent was necessary for efficient chitosanase recovery in tissue homogenates but not in IF exu'acts (data not shown). The localization of the chitosanase forms in transgenic tobacco was studied by comparing the amount of chitosanase recovered in leaf homogenates before and after three successive IF extracts (Fig. 2). The electrophoretic (a) form was detected as an intracellular protein in pBI9C tobacco. However, the (b) form, which was the most abundant active chitosanase in IF extracts, was found in approximately the same amount in IF extracts and leaf homogenates. The chitosanase (c) form was a faint activity.

Quantifying chitosanase in transgenic tobacco Chitosanase activity in IF extracts (pBI9C) was quantified by a lysoplate assay similar to that for lysozyme activity in transgenic tobacco (Trudel et al. 1992). A cow,relation was established between the lysis zone diameter (ram) and the amount of purified Streptomyces N174 chitosanase (pg). The amount of chitosanase was found to be equal to 0.00173 X 10(0.132 X lysis zone diameter) according to an exponential curve with an R 2 correlation coefficient value of 0.986. By this measure, the amount of chitosanase was estimated at 10 ng/pL in pBI9C IF extracts. As for other results, at least 25 independent transgenic tobacco were considered in this analysis. Plants varied slightly in their chitosanase contents (up to 2-3X).

Chitosanase N-terminus sequences The three chitosanase forms were purified for microsequencing by sequential native and denaturing PAGE. The purified Streptomyces sp. strain N174 mature chitosanase was also subjected to the stone electrophoretic steps. As previously reported for purified Streptomyces chitosanase (Boucher et al. 1992), the N-terminus sequence (AGAGL) was obtained. No N-terminus sequence could be obtained for chitosanase (a) and (b) forms. In fact, they were found to be contaminated with a protein having the Nterminus sequence (GDIVVY) identical to a basic 29.7-kD

Fig. 1. Fungal growth inhibition by Streptomyces NI74 chitosanase. The purified Streptomyces N174 chitosanase was tested for growth inhibition against: Rhizopus nigricans (upper left), Fusarium oxysporum f. sp. radieis-lycopersici (upper right), Verticillium albo-atrum (lower left) and Pythium uhimum (lower right). Various amounts of enzyme were used: (1) 100 pg, (2) 50 lag, (3) 25 pg, (4) 2.5 pg, (5) 0.25 ~tg and (6) 0.025 pg. Sterile 50% (v/v) glycerol (c) was used as conffol. Assays were performed on 10% (w/v) polyacrylamide gel slabs as described in Materials and Methods.

225 Fig. 2. Electrophoretic analysis of chitosanase activity from transgenic tobacco. Leaf homogenates (20 p.g protein) before (HB) and after (HA) three successive intercellular fluid extracts (0.2 lag protein) (IF1, IF2, IF3) from pBI9C transgenic tobacco were subjected to SDS-PAGE containing 0.01% (w/v) glycol chitosan as substrate Chitosanase activity was revealed by Calcofluor White M2R staining. CSN corresponds to l0 ng of purified Streptomyces N174 chitosanase. Known amounts (1, 10, 100 and 1000 rig) of purified Streptomyces chitosanase (CSN) were used to compare the chitosanasc activity in transgenic tobacco. The three active chitosanase forms (a, b, c) in pBI9C transgenic tobacco are indicated on left. The arrow head on right indicates the molecular mass of the purified mature chitosanase t~:om Streptomyces N174. The white fluorescent bands found in homogenate ext,'acts correspond to proteins present in large amounts after Coomassie Blue R-250 staining (not shown). The prestained molecular mass markers (Bio-Rad) are: lysozyme (18.5 kD), soybean trypsin inhibitor (27.5 kD), carbonic anhydrase (32.5 kD), ovalbumin (49.5 kD), bovine serum albumin (80 kD) and phosphorylase 13 (106.5 kD). class III tobacco chitinase (Lawton et al. 1992). The presence of a contaminating chitinase activity associated with the two chitosanase forms was further confirmed in PAGE chitinase assays (Trudel et al. 1989) (data not shown). However, this chitinase did not exhibit chitosanase activity according to PAGE activity assays (Grenier and Asselin 1990). A chitosanase N-terminus sequence (STAVKAGAGL) was obtained for the (c) folTn (Fig. 3). Thus, this transgenic tobacco chitosanase minor form (31.5 kD) was processed five amino acids upstream of the signal peptide cleavage site for the Strepwmyces sp. strain N174 mature enzyme (29.5 kD) (Fig. 3). This yielded a protein of higher molecul~ mass (794 D) and of altered overall charge (one basic lysine among the five additional amino acids). Analysis of transgenic chitos~a,~e forms after native PAGE confirmed that they did not migrate exactly as the Streptomyces mature chitosanase (data not shown).

Discussion In this study, we describe the stable transformation of tobacco plants with an antifungal Streptomyces sp. strain N174 chitosanase. The chitosanase is expressed and partially secreted using the actinomycete signal peptide. As with other hydrolases acting on microbial cell walls, chitosanases may possess an antimicrobial activity. In this context, the antimicrobial activity of purified Streptomyces sp. strain N174 chitosanase was evaluated in vitro. The growth of some pathogenic fungi was inhibited by the Streptomyces chitosanase. Moreover, the growth inhibition of Pythium uItimum was rather suprising as this oomycete fungus was reported to contain mainly cellulose and small amounts of chitin in its cell wall (Ch6rif et al. 1992). This inhibition suggests that chitosan or chitosan-like components could possibly exist in association with some P. ultimum cell wall component or that the chitosanase could affect cellulose directly. However, Streptomyces chitosanase was reported not to

degrade chitin, CM-cellulose, Avicell or laminarin (Boucher et al. 1992). As with other fungal cell wall hydrolases (Broglie et al. 1991; Sela-Buurlage et al. 1993), relatively large amounts of purified protein (2.5-25 I.tg) were required for growth inhibition. These results, along with recent reports on chitosanase induction upon fungal and symbiotic interactions (Sharma et al. 1993; DumasGaudot et al. Gaudot et al. 1992), suggest that chitosanase may be an additional enzyme involved in plant defense. The fact that only the chitos,'mase (a) form migrated very closely to the secreted mature chitosanase from Streptomyces N174 while the other forms migrated a~ proteins of higher molecular masses suggests that

401 41 31

A

S

'1'

Q

A

S

T

A

V

K

A

G

A

G

1

A

V

K

A

G

A

O

L

CSN S. N174

3130 31

Fig.

A

$

3.

T

Q

A

S

T

N-terminus

sequence

45

comparison

pBIgC (r

form

of the

Streptomyces sp. strain N174 chitosanase and the transgenic chitosanase (c) form. Approximately 3 lag of the chitosanase (c) form from pBIgC transgenic tobacco was purified by successive native and denaturing PAGE from R 2 progeny of self-pollinated tobacco, blotted to PVDF membrane and sequenced for its N-terminus (see Materials and Methods). The purified chitosanase from Streptomyces sp. strain N174 (CSN S. N174) was subjected to the same steps. Sequences of both chitosanases were aligned according to the reported amino acid sequence of Streptomyces Nl74 chitosanase (Masson et al. 1994; Boucher et al. 1992). Underlined sequences indicate the N-terminus sequences determined by microsequencing and arrows indicate the cleavage site: after aa 40 for chitosanase (CSN) S. N174 and after aa35 for chitosanase (CSN) pBI9C (c) form.

226 chitosanase could have undergone post-translational modifications. Two possibilities can be considered: distinct signal peptide cleavage and/or glycosylation. N-terminus sequencing of the chitosanase (c) form revealed that the signal peptide sequence was processed in a different m ~ n e r from the cleavage site in Streptomyces. This cleavage site obeys the (-3, -D-rule as a prokaryotic signal peptide (von Heijne, 1986) while the observed processing site for the Streptomyces sp. strain N174 chitosanase expressed in Streptomyces lividans did not (Boucher et al. 1992, Masson et al. , 1994). The five additional amino acids (794 D) could not fully explain the estimated difference in molecular mass (2000 D) observed after SDS-PAGE between the chitosanase (c) form and the mature purified Streptomyces chitosanase. Thus, glycosylation may be another putative modification. Indeed, two potential N-glycosylation (AsnXaa-Ser/Thr) sites were found in the chitosanase sequence (DNA Strider TM 1.2 analysis). A number of bacterial proteins have been previously reported to be glycosylated when produced in tobacco (Pen et al. 1992; Lund and Dunsmnir, 1992). The small difference in molecular mass after SDS-PAGE between the chitosanase (a) form and the purified Streptomyces mature chitosanase could con'espond to 794 D (molecular mass of the five additional mnino acids). We suggest that the three chitosanase forms were cleaved at the same site and the difference observed between their molecular masses may be due to differential glycosylation. This study has shown the expression of an active actinomycete chitosanase in t.ransgenic tobacco without affecting normal plant growth and development. Moreover, the Streptomyces chitosanase showed an antifungal potential in vitro.

Acknowledgments. This work was supported by a grant from the Natural Sciences and Engineering Research Council of Canada (NSERC) and a fellowship (S.E.O.) from the Progralmne de Bourses de La Franeophonie. We thank Jean Grenier and Jean Trudel for their collaboration.

References

Ando A, Noguchi K, Yanagi M, Shinoyama H, Hirata H, Yabuki M, Fujii T (1992) J Gen Appl Microbiol 38:135-144 Aa'aki B, Ito E (1975) Eur J Biochem 55:71-78 Boucher I, Dupuy A, Vidat P, Neugebauer WA, Brzezinski R (1992) Appl Mierobiol Biotechnol 38:188-193 Broglie K, Chet I, Holliday M, Cressman R, Biddle P, Knowlton C, Mauvais CJ, Broglie R (1991) Science 254: 1194-1197 Ch6rif M, Benhamou N, B~langer RR (1992) Can J Microbiol 39:213-222 Cuero RG, Osuji GO (1993) In: Muzzarelli RAA (ed) Chitin Enzymology. European Chitin Society, Ancona, pp 277282 Davis B, Eveleigh DE (1984) In: Zikakis JP (ed) Chitin, Chitosan and Related Enzymes, Academic Press, pp 161179 Dumas-Gaudot E, Grenier J, Furlan V, Asselin A (1992) Plant Sci 84:17-24 E10uakfaoui S, Asselin A (1992 a) Phytochemistry 31: 15131518

E10uakfaoui S, Asselin A (1992 b) Plant Sci 85:33-41 Fink D, Boucher I, Denis F, Brzezinski R (1991) Biotechnol Lett 13:845-850 Grenier J, Asselin A (1990) Mol Plant-Microbe Interact 3: 401-407 Horsch R, Fry J, Hoffman NL, Eichholtz D, Rogers SG, Fraley RT (1985) Science 227:1229-1231 Howie W, Newbigin E, Joe L, Penzes E, Suslow T, Dunsmuir P (1992) Abstract, Sixth Int Symp Mol Plant-Microb Inter, Seattle. Lawton K, Ward E, Payne G, Meyer M, Ryals J (1992) Plant Mol Biol 19:735-743 Lund P, Dunsmuir P (1992) Plant Mol Biol 18:47-53 Masson JY, Denis F, Brzezinski R (1994) Gene 140:1133-107 Masson JY, Li T, Boucher I, Beaulieu C, Brzezinski R (1993) In: Muzzarelli RAA (ed) Chitin Enzymology. European Chitin Society, Ancona, pp 423-430 Monaghan RL, Eveleigh DE, Tewari RP, Reese ET (1973) Nature 245:78-80 Neuhaus JM, Ahl-Goy P, Hinz U, Flores U, Meins F (1991) Plant Mol Biol 16:141-151 Osswald WF, Shapiro JP, McDonald RE, Niedz RP, Mayer RT (1993) Experientia 49:888-892 Parent JG, Asselin A (1984) Can J Bot 62:564-569 Pelletier A, Sygusch J (1990) Appl Environ Microlfiol 56: 844-848 Pen J, Molendijk L, Quax WJ, Sijmons PC, van Ooyen AJJ, van den Elzen PJM, Rietveld K, Hoekema A (1992) Biotechnol 10:292-296 Sela-Buurlage MB, Ponstein AS, Vloemans SA, Melchers LS, van den Elzen PJM, Cornelissen BJC (1993) Plant Physiol 101:857-863 Sharma P, B0rja D, Stougaard P, L6nneborg A (1993) Physiol Mol Plant Pathol 43:57-67 Trudel J, Asselin A (1994) Anal Biochem 221:214-216 Trudel J, Audy P, Asselin A (1989) Mol Plant-Microbe Interact 2:315-324 Trudel J, Potvin C, Asselin A (1992) Plant Sci 87:55-67 Trudel J, Potvin C, Asselin A (1995) Plant Sci 106:55-62 Vierheilig M, Alt M, Neuhaus JM, Boller T, Wiemken A (1993) Mol Plant-Microbe Interac 6:261-263 yon Heijne G (1986) Nucleic Acids Res 14:4683-4990 Zikakis JP (1984) Chitin, chitosan and related enzymes. Part IlI, Enzymology and Genetic Enzyme Engit~eering. Zikakis JP (ed) Academic Press, Inc.