WO1997032973A1

WO1997032973A1 - Combinations of fungal cell wall degrading enzyme and fungal cell membrane affecting compound

Info

Publication number: WO1997032973A1
Application number: PCT/US1997/003344
Authority: WO
Inventors: Gary E. Harman; Matteo Lorito; Antonio Di Pietro; Christopher K. Hayes; Felice Scala; Christian P. Kubicek
Original assignee: Cornell Research Foundation, Inc.
Priority date: 1996-03-05
Filing date: 1997-03-05
Publication date: 1997-09-12
Also published as: EP0886677A1; US6512166B1; AU2065597A; JP2000507222A

Abstract

A system for inhibiting the germination or growth of a fungus comprises (a) fungal cell wall degrading chitinolytic or glucanolytic enzyme and (b) antifungal cell membrane affecting compound. Exemplified antifungal fungal cell membrane affecting compounds include flusilazole, miconazole, osmotin, gramicidin, valinomycin, phospholipase B, and trichorzianines. The system components (a) and (b) may be supplemented with polyene macrolide antibiotic, antifungal epithiodiketopiperazine antibiotic (e.g., gliotoxin), fungal cell wall biosynthesis inhibitor (e.g., L-sorbose) and/or detergent. Embodiments include method of contacting a plan which expresses cell wall degrading enzyme with antifungal fungal cell membrane affecting compound.

Description

COMBINATIONS OF FUNGAL CELL WALL DEGRADING ENZYME AND FUNGAL CELL MEMBRANE AFFECTING COMPOUND

This invention was made at least in part with Government support under U.S. -Israel Binational Agricultural Research and Development Fund (BARD) grant number US-1723-89. The Government has certain rights in the invention. Cross-Reference to Related Application This is a continuation-in-part of Serial No. 08/499,164, filed July 7, 1995, which is continuation of serial no. 08/249,927, now U.S. Patent No. 5,433,947, which is a continuation of serial no. 07/990,609, now U.S. Patent No. 5,326,561. This is also a continuation-in-part of Serial No. 60/007567, filed November 27, 1995. Technical Field

This invention is directed at antifungal synergistic combinations of fungal cell wall degrading enzyme and fungal cell membrane affecting fungicide and use thereof for topical or internal application in agriculture or medicine to inhibit germination or growth of fungi . Background of the Invention

The primary methods of controlling disease-causing fungi on crop plants and on animals, including humans, comprise treatment with synthetic chemical pesticides. However, the exposure of man and habitats to increasing amounts of pesticides has come under criticism, resulting in a search for environmentally safer methods including the use of synergistic combinations of fungicides to reduce the amounts of application.

Poulose, A.J., in Koeller, W. , ed. , Target Sites of Fungicide Action, CRC Press, Boca Raton, Florida, 1992, at pages 313-314 reviews the disclosures of a number of authors directed to synergistic interaction of different lytic enzymes produced by a variety of microorganisms with a small number of antifungal compounds including amphotericin B, benomyl, polyoxin B, kitazin P and nikkomycin. Summary of the Invention

It is an object of this invention to expand the range of synergistic combinations of fungicide/enzyme.

In one embodiment, the invention is directed to a system for inhibiting the germination or growth of a fungus, said system comprising (a) fungal cell wall degrading chitinolytic or glucanolytic enzyme; (b) antifungal fungal cell membrane affecting compound which is not expressed by the same organism as the fungal cell wall degrading enzyme in nature, in an amount to provide a concentration where it provides about 4 to 95% inhibition of spore germination when used without (a) ; the weight ratio of (a) to (b) being 0.005:1 to 500,000:1.

Preferably the fungal cell wall degrading chitinolytic or glucanolytic enzyme is present in an amount to provide a concentration where said enzyme provides 2 to 50% inhibition of spore germination when it is used without antifungal fungal cell membrane affecting compound and the antifungal fungal cell membrane affecting compound is present in an amount where it provides 10 to 70% inhibition of spore germination when it is used without fungal cell wall degrading chitinolytic or glucanolytic enzyme and the total of the percentage inhibitions provided by the fungal cell wall chitinolytic or glucanolytic enzyme and the antifungal fungal cell membrane affecting compound when each is used without the other is less than 100%.

Very preferably, the fungal cell wall degrading chitinolytic or glucanolytic enzyme is present in an amount to provide a concentration where said enzyme provides 5 to 20% inhibition of spore germination when it is used without antifungal fungal cell membrane affecting compound and the antifungal fungal cell membrane affecting compound is present in an amount to provide a concentration where said compound provides 15 to 60% inhibition of spore germination when it is used without fungal cell wall degrading enzyme. The term "system" is used because the fungal wall degrading chitinolytic or glucanolytic enzyme and antifungal fungal cell membrane affecting compound can be applied as part of the same composition or can be applied concurrently as separate compositions or can be applied separately at different times. Preferably, the two kinds of antifungal components of the system are applied in the same composition or concurrently as separate compositions or the antifungal fungal cell membrane affecting compound is applied up to 8 hours after the cell wall degrading chitinolytic or glucanolytic enzyme.

The term "inhibit" is used herein to mean reduce the growth and/or development of fungi compared to where inhibiting agent is not present.

The term "fungal cell wall degrading chitinolytic or glucanolytic enzyme" is used herein to mean chitinolytic or glucanolytic enzyme that effects lysis of fungal cell walls. The term "antifungal fungal cell membrane affecting compound" is used herein to mean sterol synthesis inhibiting fungicide, antifungal peptide antibiotic, zeamatin and proteins that are serologically related to zeamatin and antifungal lipid lytic enzymes. The term "fungal cell membrane" means plas alemma and membranes surrounding secretory vesicles, vacuoles, mitochondria, endoplasmic reticulum, and nuclei.

The limitation "which is not expressed by the same organism as the fungal cell wall degrading enzyme in nature" is to exclude combinations which occur in nature.

The concentration where fungal cell wall degrading enzyme individually provides a specified percentage fungal inhibition or where antifungal fungal cell membrane affecting compound individually provides a specific percentage fungal inhibition can be determined as follows: Assays are performed under sterile conditions. Equal volumes of spore suspensions, 3X potato dextrose broth, and the test solution or suspension in 5 mM Tris-HCl (pH 6.0) or 5 mM potassium phosphate buffer (pH 6.7) are mixed. The control is the same as the test solution except for the control the antifungal agent is omitted. The assay mixtures (total volume 45 or 30 μl) are incubated on flat-bottomed ELISA plates, each containing 96 wells, with 2,000 to 3,000 spores per well, at 25oC. After 22 to 30 hours, the plates are placed under an inverted microscope. The percentage of conidia germinating is determined as the percentage of the first 100 spores randomly found in a well. In addition, the lengths of 20 germ tubes are measured and averaged. All experiments are performed twice, with three replicates for each treatment. The inhibition values obtained in the two experiments are combined and averaged, and standard deviations are calculated from the 6 data points. The values obtained for the control are taken as 0% inhibition and all other values are divided by the values obtained for the control and multiplied by 100% to obtain percent inhibition.

Determination of concentration corresponding to a particular percent inhibition is carried out by subjecting dose-response curves to regression analysis by using a binomial regression of the third order, with R² ranging between 0.95 and 0.99.

The system of the invention can optionally contain as additional components, for example, antifungal polyene macrolide antibiotic, antifungal epithiodiketopiperazine antibiotic, fungal cell wall biosynthesis inhibitor (e.g., chitin synthetase inhibitor or β- 1, 3-glucan synthetase inhibitor) and/or detergent, in an inhibition improving amount. The term "inhibition improving amount" is used to mean an amount causing a greater % fungal inhibition than if the additional component (s) is/are omitted.

In another embodiment, the invention is directed to a method of inhibiting the germination or growth of a fungus and comprises contacting such fungus or a locus to be protected from such fungus with an antifungal effective amount of combination of fungal cell wall degrading chitinolytic or glucanolytic enzyme in a concentration where said enzyme individually provides 2 to 50% inhibition of spore germination and antifungal fungal cell membrane affecting compound which is not chitinolytic or glucanolytic enzyme and which is not expressed by the same organism as the fungal cell wall degrading enzyme in nature in a concentration where said compound individually provides about 4 to 95%, for example, 10 to 70%, inhibition of spore germination, the total of the percentage inhibitions individually provided by the fungal cell wall degrading chitinolytic or glucanolytic enzyme and the antifungal fungal cell membrane affecting compound being less than 100%.

The term "locus to be protected from such fungus" includes seeds, roots, stems, leaves, flowers and fruits to be protected and to the soil surrounding seeds and roots to be protected, as well as animal or human tissues or organs to be protected.

In another embodiment, the invention is directed to a method of protecting from a fungus, a plant which expresses fungal cell wall degrading chitinolytic or glucanolytic enzyme at a level of about 0.05 to 5% of total cellular protein, said method comprising contacting said plant with an antifungal effective amount of an antifungal fungal cell membrane affecting compound at a concentration where it individually provides about 4 to 95%, for example, 10 to 70% inhibition of spore germination. In still another embodiment, the invention is directed to a transgenic plant protected against pathogenic fungi which is a plant susceptible to fungal attack which has been transformed to contain gene which expresses fungal cell wall degrading chitinolytic or glucanolytic enzyme at a level of about 0.05 to 5% of total cellular protein and also which has been transformed to contain gene which expresses protein antifungal cell membrane affecting compound or which has been infected with transgenic endomorphic microorganism producing said protein antifungal fungal cell membrane affecting compound typically in the xylem, in an amount to provide a concentration of said compound where it individually provides about 4 to 95% inhibition of spore germination.

The term "antifungal effective amount" is used herein to mean an amount effective to inhibit the germination or growth of a fungus.

Brief Description of the Drawings

Fig. 1 is a set of bar graphs depicting % inhibition at various flusilazole concentrations, in the presence of different enzymes and in the absence of enzyme, showing results of Example II.

Fig. 2 is a set of bar graphs depicting % inhibition at various miconazole concentrations, in the presence of different enzymes and in the absence of enzyme, showing results of Example II.

Fig. 3 is a set of graphs of osmotin concentration versus percent inhibition of spore germination of B . cinerea with concentration of endochitinase from G . virens strain 41 varied and shows results of Example III.

Fig. 4 is a set of graphs of osmotin concentration versus percent inhibition of spore germination of B . cinerea with concentration of glucosidase from T. harzianum strain PI varied and shows results of Example IV.

Fig. 5 is a set of graphs of osmotin concentration versus percent inhibition of spore germination of F. oxysporum with concentration of endochitinase from T. harzianum strain PI varied and shows results of Example V. Fig. 6 is a set of graphs of osmotin concentration versus percent inhibition of spore germination of F. oxysporum with concentration of glucosidase from T. harzianum strain PI varied and shows results of Example VI .

Fig. 7 is a set of graphs of gramicidin concentration versus percent inhibition of spore germination of B . cinerea with concentration of endochitinase from G . virens strain 41 varied and shows results of Example VII.

Fig. 8 is a set of graphs of gramicidin concentration versus percent inhibition of spore germination of F. oxysporum with concentration of glucosidase from T. harzianum strain PI varied and shows results of Example VIII.

Fig. 9 is a set of graphs of valinomycin concentration versus percent inhibition of spore germination of B . cinerea with shows results of Example IX. Fig. 10 is a set of graphs of valinomycin concentration versus percent inhibition of spore germination of F. oxysporum with concentration of glucosidase from T. harzianum strain PI varied and shows results of Example X.

Fig. 11 is a set of graphs of phospholipase B concentration versus percent inhibition of spore germination of B . cinerea with concentration of endochitinase from T. harzianum strain PI varied and shows results of Example XI .

Fig. 12 is a set of graphs of phospholipase B concentration versus percent inhibition of spore germination of B . cinerea with concentration of chitobiosidase from S . albidoflavus NRRL B-16746 varied and shows results of Example XII.

Fig. 13 is a set of graphs of trichorzianine Al concentration versus percent inhibition of F. oxysporum with concentration of endochitinase from G. virens strain 41 varied and shows results of Example XIII. Fig. 14 is a set of graphs of trichorzianine Al concentration versus percent inhibition of F. oxysporum with concentration of nagase from Tri choderma harzianum strain PI varied and shows results of Example XIV.

Fig. 15 is a set of graphs of duration of assay versus percent relative time of application of antifungal fungal cell membrane affecting compound osmotin I and fungal cell wall degrading enzyme endochitinase from Nicotiana tabacum cv. Havana 425 and shows results of one experiment of Example XV. Fig. 16 is a set of graphs of time of application versus percent inhibition obtained at the end of 24 hours where each of the graphs represents a variation in relative time of application of antifungal fungal cell membrane affecting compound osmotin I and fungal cell wall degrading enzyme endochitinase from Nicotiana tabacum cv. Havana 425 and shows results of a second experiment of Example XV.

Fig. 17 is a set of graphs of osmotin concentration versus percent inhibition of spore germination of B . cinerea indicating the effect of application of L-sorbose and shows results of Example XVIII. Detailed Description

The fungal cell wall degrading chitinolytic and glucanolytic enzymes for use in the embodiments of the invention herein- include, for example, chitinolytic enzymes and β- 1 , 3-glucanolytic enzymes for degrading cell walls of fungi where the cell walls contain, as a major structural component, chitin and β- 1, 3-glucans . These enzymes are found in fungi, bacteria and higher plants. They can be in natural form, i.e., not separated from the source, e.g., by utilizing source microorganisms in the system herein, or they may be in partially purified form, i.e, purified compared to natural form but with other protein present or they may be in biologically pure form or may be expressed by transgenic plant. Fungal cell wall degrading enzymes are readily obtained in biologically pure form from source fungal microorganisms by culturing the source microorganism, concentrating the culture filtrate, fractionating by gel filtration chromatography, concentrating and further purifying by chromatofocusing followed, if necessary, by isoelectrofocusing in a Rotofor cell (BioRad, Richmond, CA) . Fungal cell wall degrading enzymes are readily obtained in biologically pure form from bacteria and higher plants by processing comprising culturing, precipitating with NH₄S0₄, dissolving and purifying by chromatography and/or isoelectric focusing.

The fungal cell wall degrading chitinolytic enzymes cleave chitin, and include, for example, antifungal endochitinases, chitin 1, 4-/S-chitobiosidases and β-N- acetylglucosaminidases. These can be obtained from fungi, for example, from the genera Trichoderma , Gliocladium, Lycoperdon and Calvatia ; from bacteria, e.g., from the genera Strep tomyces, Vibrio, Serratia and Bacillus; and from higher plants, e.g., Nicotiana , Cucumis and Phae solus .

The endochitinases are enzymes that randomly cleave chitin. Endochitinase activity is readily measured by determining optical density at 510 nm as reduction of turbidity of a 1% suspension of moist purified colloidal chitin in 100 mM sodium acetate buffer, pH 5.5, or in 50 mM KHP0₄ buffer, pH 6.7, after 24 hours of incubation at 30°C. For calculation of specific activity, one unit is defined as the amount of enzyme required to obtain a 5% turbidity reduction.

A very preferred endochitinase is coded for by gene of the genome of and is isolated and derived from Trichoderma harzianum strain PI having accession No. ATCC 74058. The protein has a molecular weight of 36 kDa (as determined by sodium dodecyl sulfate poiyacrylamide gel electrophoresis after the protein was prepared under reducing conditions, on direct comparison to migration of a 36 kDa protein) and an isoelectric point of 5.3 ± 0.2 as determined based on its elution profile from a chromatofocusing column, and a molecular weight of 40 kDa (as determined by sodium dodecyl sulfate poiyacrylamide gel electrophoresis after the protein was prepared under reducing conditions, from a regression based on the log of molecular weight of standard proteins) and an isoelectric point of 3.9 as determined by isoelectric focusing electrophoresis from a regression of distance versus the isoelectric point of standard proteins. The specific activity of the purified endochitinase was determined to be 0.86 units/μg protein with the turbidity reducing assay and 2.2 nkatal/μg protein when nitrophenyl-(S-D-N,N' ,N"- triacetylchitotriose was used as a substrate. This enzyme and its production and purification to homogeneity are described in Harman et al U.S. Patent No. 5,173,419, and also in S.N. 07/919,784, filed July 27, 1992.

Another endochitinase is coded for by gene of the genome of and is isolated and derived from Gliocladium virens strain 41 having accession No. ATCC 20906 and has a molecular weight of 41 kDa (as determined by sodium dodecyl sulfate poiyacrylamide gel electrophoresis after the protein was prepared under reducing conditions, from a regression based on the log of molecular weight of standard proteins) and an isoelectric point of 7.8 as determined by isoelectric focusing from a regression of distance versus the isoelectric point of standard proteins. The procedures used for molecular weight determination and isoelectric point determination are the same as those described in detail in S.N. 07/919,784. The enzyme is active in citric acid/K₃(PO„) buffer over a pH range of 3.5 to 7.0 and shows a 90-100% activity between pH 4.0 and 6.0 and shows maximum activity at pH 4.5. The optimum temperature for endochitinase activity at pH 5.5 is between 30 and 37oc, and activity drops off sharply at temperatures above 40oc. This enzyme and its production and purification to homogeneity are described in DiPietro, A., et al, Phytopathology 83, No. 3, 308-313 (1993) . Furthermore, its purification to homogeneity is described in detail in Reference Example 1 hereinafter. The enzyme was purified to an activity 105-fold that of its activity in the culture filtrate.

Two endochitinases are coded for by gene of the genome of and are isolated and derived from Ni cotiana tabacum cv. Havana 425 and these respectively have molecular weights of 32 kDa and 34 kDa. These endochitinases and their production and purification and obtaining of cDNA clone for endochitinase from Nicotiana tabacum cv. Havana 425 and transformation of plant to contain gene from Nicotiana tabacum cv. Havana 425 expressing endochitinase activity are described in Shinshi, H., et al, Proc. Natl. Acad. Sci. USA, 84, 89-93 (1/87) and Neuhaus, J.-M., et al, Plant Molecular Biology 16, 141-151 (1991) .

The chitin 1,4-/3-chitobiosidases cleave dimeric units from chitin from one end. Chitin 1,4-β-chitobiosidases are sometimes referred to for convenience hereinafter as chitobiosidases. Chitobiosidase activity is readily determined by measuring the release of p-nitrophenol from p- nitrophenyl-_/S-D-N,N' -diacetylchitobiose, e.g., by the following procedure. A substrate solution is formed by dissolving 3 mg of substrate in 10 ml 50 mM KHP0₄ buffer, pH 6.7. Fifty μl of substrate solution is added to a well in a microtiter plate (Corning) . Thirty μl of test solution is added, and incubation is carried out at 50oC for 15 minutes. Then the reaction is stopped by the addition of 50 μl of 0.4 M Na₂C0₃, and the optical density is read at 410 nm. An activity of one nanokatal (nkatal) corresponds to the release of 1 nmol nitrophenol per second. A chitobiosidase is coded for by gene of the genome of and is isolated and derived from Trichoderma harzianum strain PI having accession No. ATCC 74058 and in its most prevalent form has a molecular weight of 36 kDa (as determined by sodium dodecyl sulfate poiyacrylamide gel electrophoresis after the protein was prepared under reducing conditions, on direct comparison to migration of a 36 kDa protein) , and an isoelectric point of 4.4 ± 0.2 as determined based on its elution profile from a chromatofocusing column and a molecular weight of 40 kDa (as determined by sodium dodecyl sulfate poiyacrylamide gel electrophoresis after the protein was prepared under reducing conditions, from a regression based on the log of the molecular weight of standard proteins) , and an isoelectric point of 3.9 as determined by isoelectric focusing electrophoresis from a regression of distance versus isoelectric point of standard proteins. Conditions for molecular weight determination and isoelectric point determination are described in detail in S.N. 07/919,784. It has an optimum pH for activity of about 3 to 7. This chitobiosidase and its production and purification are described in Harman et al U.S. Patent No. 5,173,419 where it is referred to as a chitobiase, and also in S.N. 07/919,784, filed July 27, 1992, where it is referred to as a chitobiase and also as a chitobiosidase. The enzyme obtained in S.N. 07/919,784 has a specific activity of 127 nkatal/mg protein and is purified to greater than a 200-fold increase in specific activity compared to its activity in the culture filtrate. S.N. 07/919,784 refers to the presence also of a minor band at 36 kDa. It has since been discovered that the chitobiosidase from Trichoderma harzianum strain PI (ATCC 7<±058) gives three closely spaced protein bands with molecular weights of 40 kDa (staining most intensely) , 38 kDa (faintest stain) and 35 kDa (intermediate intensity stain) , as determined by sodium dodecyl sulfate poiyacrylamide gel electrophoresis after the protein was prepared under reducing conditions, from a regression based on the log of the molecular weight of standard proteins, and that the three bands represent different levels of N-glycosylation of the same protein. Another chitobiosidase is coded for by gene of the genome of and is isolated and derived from Gliocladiu virens strain 41 having accession No. ATCC 20906 and has a molecular weight of 38 kDa (as determined by sodium dodecyl sulfate poiyacrylamide gel electrophoresis after the protein was prepared under reducing conditions, from a regression based on the log of molecular weight of standard proteins) , and an isoelectric point of 4.95 (as determined by isoelectric focusing electrophoresis from a regression of distance versus isoelectric point of standard protein) . The proteins used in the determination of molecular mass were 6 standard proteins obtained from Bio-Rad Laboratories, Hercules, CA, and these proteins and their molecular weights in kDa are respectively hen egg white lysozyme, 14.4; soybean trypsin inhibitor, 21.5; bovine carbonic anhydrase, 31; hen egg white ovalbumin, 45; bovine serum albumin, 66.2; and rabbit muscle phosphorylase b, 97.4. The proteins used in the determination of isoelectric point were 12 standard proteins obtained from Sigma Chemical Company and are respectively amyloglucosidase, 3.6; methyl red dye, 3.8; soybean trypsin inhibitor, 4.6; β-lactoglobulin, 5.1; bovine carbonic anhydrase B, 5.9; human carbonic anhydrase B, 6.6; horse myoglobin cyanocytic band, 6.8; horse myoglobin basic band, 7.2; L-lactic dehydrogenase from rabbit muscle acidic band, 8.3; L-lactic dehydrogenase from rabbit muscle middle band, 8.4; L-lactic dehydrogenase from rabbit muscle basic band, 8.6; and trypsinogen, 9.3. For the linear regressions, r² values ranged from 0.94 to 0.99. This enzyme and its production and purification to homogeneity are described in Reference Example 2 hereinafter.

Two chitibiosidases are coded for by gene of the genome of and are isolated and derived from Streptomyces albidoflavus having accession no. NRRL B-16746. These respectively have molecular weights of 27 kDa and 34 kDa and have isoelectric points less than 3.0. The chitobiosidase activity was isolated as follows: The bacteria were grown on slants of trypticase soy agar (BBL, Cockeysville, MD) . Growth was transferred to a liquid medium (50 mM Tris, pH 9.0, 0.012% magnesium sulfate, 0.1% glucose, 0.1% calcium chloride, 0.05% manganese sulfate, 0.025% ferrous sulfate, 0.00125% zinc sulfate, 0.5% crab shell chitin (Sigma Chemical Co., St. Louis) ) . The biomass was removed from the broth by centrifugation and filtration. The remaining liquid was brought to 95% saturation with ammonium sulfate, and the precipitate was collected by centrifugation at 6000 xg for 30 min. at 4°C. The pellet was resuspended in dH₂0, dialyzed against ice-cold dH₂0 to remove salt and centrifuged at 6000 xg for 10 min. at 4°c to remove insoluble particles. The culturing and purification up to this point is described in Broadway, R. M. , et al, Lett. Appl. Microbiol. 20, 271-276 (1995) . Isolation of chitobiosidase activity was obtained by isoelectric focusing separation as follows: The resulting liquid was applied in approximately equal amounts to compartments of a Rotofor Isoelectric Focusing apparatus (Bio- Rad) . The first three fractions contain the chitobiosidase activity.

The 0-N-acetylglucosaminidases cleave monomeric units from chitin from one end. 3-N-Acetylglucosaminidases may be referred to for convenience hereinafter as glucosaminidases or as nagases. Glucosaminidase activity is readily determined by measuring the release of p-nitrophenol from p-nitrophenyl-β-D- N-acetylglucosaminide, e.g., by the same procedure as described above for assaying for chitobiosidase activity except for the substitution of substrate. An activity of one nanokatal (nkatal) corresponds to the release of 1 nmol nitrophenol per second. Glucosaminidase activity is present in culture filtrates from Trichoderma harzianum strain PI having accession No. ATCC 74058 and from Gliocladium virens strain 41 having accession No. ATCC 20906.

A nagase coded for by gene of the genome of and isolated and derived from Tri choderma harzianum strain Pi having accession No. ATCC 74058 has a molecular weight of 72 kDa (as determined by sodium dodecyl sulfate poiyacrylamide gel electrophoresis after the protein was prepared under reducing conditions, from a regression based on the log of molecular weight of standard proteins) , and an isoelectric point of 4.6 (as determined by isoelectric focusing electrophoresis from a regression of distance versus isoelectric point of standard proteins) . It has good activity over a pH range of 4 to 7 and optimal activity between pH 5.0 and 5.5, as determined in a 50 M citric acid/potassium phosphate buffer mixture at pH levels ranging from 3.0 to 9.0. It has good activity over a temperature range of 25 to 85°C with optimal activity at 60 to 70oC, as determined in 50 mM potassium phosphate buffer pH 6.7 at temperatures between 20oc and lOOoc. It is quite resistant to heat inactivation, retaining about 70, 25 and 10% of activity after 15 minutes at 80, 90 and lOOoC, respectively. This enzyme and its production and purification are described in Harman et al S.N. 08/049,390.

The fungal cell wall degrading glucanolytic enzymes include, for example, antifungal glucan 1, 3- -glucosidases. The glucan 1, 3-jβ-glucosidases cleave 1, 3-?-glucans. The sources for these enzymes are typically the same as the sources for chitinolytic enzymes and are preferably microorganisms from the genera Trichoderma and Gliocladium. Glucan 1, 3- -glucosidase activity is readily determined by measuring the amount of reducing sugar release from laminarin in a standard assay containing 250 μl of enzyme solution and 250 μl of a 0.1% solution of laminarin in 50 mM potassium phosphate buffer, pH 6.7, wherein incubation is carried out at 30oC for 1 hour whereupon 250 μl of a copper reagent (prepared by dissolving 28 g Na₂P0₄ and 40 g potassium sodium tatrate in 700 ml deionized water, adding 100 ml of IN NaOH, then adding 80 ml of a 10% (w/v) solution of CuS0₄«5H₂0 with stirring, then adding 180 g Na₂S0₄, when all the ingredients have dissolved, bringing to 1 L with deionized water, then allowing to stand for 2 days, then decanting and filtering) is added, and the admixture is covered with foil and heated for 20 minutes in a steam bath, whereupon, after cooling, 250 μl of arsenomolybdate reagent (prepared by dissolving 25 g of (NH₄) ₆Mo₇0₂₄*4H₂0 in 450 ml deionized water, adding 21 ml concentrated H₂S0₄ with mixing, then adding a solution containing 3 g Na₂HAs0₄»7H₂0 in 25 ml distilled water and mixing, incubating at 37oC for 2 days and storing in a brown bottle until used) is added with mixing, followed by adding of 5 ml deionized water, and reading color in a spectrophotometer at 510 nm, and wherein appropriate controls without either enzyme or substrate may be run simultaneously; the quantity of reducing sugar is calculated from glucose standards included in the assay. An activity of one nkatal corresponds to the release of 1 nmol glucose equivalent per second. Glucan 1,3- -glucosidase activity is present in culture filtrates from Tri choderma harzianum strain PI having accession No. ATCC 74058 and from Gliocladium virens strain 41 having accession No. ATCC 20906.

A glucan 1, 3-/3-glucosidase is coded for by gene of the genome of and is isolated and derived from Trichoderma harzianum strain PI having accession No. ATCC 74058 and has a molecular weight of 78 kDa (as determined by sodium dodecyl sulfate poiyacrylamide gel electrophoresis after the protein was prepared under reducing conditions, from a regression based on the log of molecular weight of standard proteins) and an isoelectric point of 6.2 as determined by isoelectric focusing electrophoresis from a regression of distance versus the isoelectric point of standard proteins. The procedures for molecular weight determination and for isoelectric point determination are the same as those described in S.N. 07/919,784. The enzyme has activity against jβ-1,3 glucan laminarin between pH 4 and 7, with the strongest activity between 4.5 and 5.5. It releases glucose from laminarin at the same rate as reducing groups, which indicates that it is an exoglucanase cleaving monomeric glucose from the laminarin molecule. The enzyme is obtained and purified as generally described above with the medium for culturing of the microorganism being SMCS medium (the same medium used for production of endochitinase from G. virens as described in Reference Example 1 hereinafter) . After the chromatofocusing step, several peaks with glucan 1, 3- -glucosidase activity are detected and fractions from major activity peaks are pooled, dialyzed, concentrated and applied to the Rotofor cell to obtain an electrophoretically pure exo-glucanase. The production and purification of the enzyme are described in detail in Reference Example 3 hereinafter. The enzyme was purified to a specific activity about 35-fold that of its activity in the culture filtrate.

Purified cell wall degrading enzyme has been found to inhibit the germination or growth of a fungus at a concentration in solution, for example, of 50 ppm to 1000 ppm.

As indicated above, the antifungal fungal cell membrane affecting compounds of the embodiments of the invention are selected from the group consisting of sterol synthesis inhibiting fungicides, antifungal peptide antibiotics, zeamatin and proteins that are serologically related to zeamatin, and antifungal lipid lytic enzymes.

The sterol synthesis inhibiting fungicides include dimethylation synthesis step inhibitors which are pyridines and pyrimidines and azoles including imidazoles and triazoles. Pyridines and pyrimidines are useful for agricultural purposes and include, for example, triarimol, fenarimol, nuarimol, buthiobate and pyrifenox. Imidazoles useful for agricultural purposes include, for example, imazalil, prochloraz, and triflumidol. Imidazoles useful for medicinal purposes include, for example, miconazole, isoconazole, econazole, clotrimazole, bifonazole, butoconazole, ketoconazole, tioconazole, oxiconazole, fenticonazole, sulconazole and o oconazole. Triazoles useful for agricultural purposes include, for example, triadimefon, triadimenol, bitertanol, diclobutrazole, propiconazole, penconazole, diniconazole, flutriafol, flusilazole, hexaconazole, tebuconazole, myclobutanil, cyproconazole, furconazole and CGA 169374. Triazoles useful for medicinal purposes include, for example, vibunazole, terconazole, itraconazole, fluconazole and ICI 195-739.

The antifungal peptide antibiotics include, for example, valinomycin, gramicidin and peptaibols including trichorzianines, trichotoxins, alamethicins, paracelsins, trichobrachin, and zervamicins. As indicated in Ghisalberti, E. L, et al, Soil Biol. Biochem., Vol. 23, No. 11, 1011-1020 (1991) , the peptaibols are peptides containing aminojlsobutyric acid and a C-terminal aminoalcohol with the term "peptaibol" being a coined word from the underlined portions of the aforestated definition. Valinomycin and gramicidin are available from Sigma Chemical Company. Langs, D.A., Science, Vol. 241, 188-191 (July 1988) describes gramicidin and refers to it as forming ion channels in lipid membranes. Ghisalberti, E. L. , et al, Soil Biol. Biochem., Vol. 23, No. 11, 1011-1020 in a review article on antifungal antibiotics produced by Trichoderma spp. describes antifungal peptaibols alamethicin 1, alamethicin 2, paracelsins a-d, trichobrachin, trichotoxin a40, trichotoxin a50, trichorzianine A IIIc, and trichorzianine B IIIc. Ghisalberti et al describes these as interacting with phospholipid membranes and inducing membrane permeability. Brewer, D., et al, Canadian J. Microbiol. 33, 619-625 (1987) describes alamethicins produced by Trichoderma spp. and the isolation of two of them, namely alamethicin 3 and alamethicin 6 and mentions and describes zervamicins produced by E ericellopsis spp. Argondelis, A. D., J. Antibiot. 27, 321-328 (1974) describes zervamicins I and II. Schirmbock, M., et al, Applied and Environmental Microbiology, Vol. 60, No. 12, 4364-4370 (12/94) describes trichorzanines Al and Bl from T. harzianum rifae (ATCC 36042) . Schirmboch et al describes these as forming voltage-gated ion channels in black lipid membranes. We turn now to zeamatin and proteins that are serologically related to zeamatin, that is cross react with antizeamatin antibody under reducing conditions. Zeamatin and its purification are described in Roberts, W. K. , et al, Journal of General Microbiology, 136, 1771-1778 (1990) . Roberts et al indicates that zeamatin has a molecular mass of 22 kDa and suggests its antifungal properties are the result of forming transmembrane pores in fungal membranes . The - If

proteins that are serologically related to zeamatin include, for example, osmotin, thaumatin, PR-R, PR-S, NP24 and 22 kDa proteins having similar N-terminal amino acid sequence to zeamatin isolated from sorghum, oats and wheat. These are found in plants in response to stress, e.g., salt stress, and cause cell membrane permeabilization. Singh, N.K., et al, Plant Physiol . , 15, 529-536 (1987) describes the recovery of osmotin from Nicotiana tabacum var Wisconsin 38 and indicates it has a molecular mass of 26 kDa and an isoelectric point greater than 8.2 and occurs in two forms, an aqueous soluble form (osmotin I) and a detergent soluble form (osmotin II) . Thaumatin is described in Edens, L., et al, Gene 18, 1-12 (1982) . PR-R and PR-S, i.e., pathogenesis-related protein R and pathogenesis-related protein S, are characterized in Kauffman, S., et al, Plant Mol. Biol 14, 381-390 (1990) . PR-R is also described in Cornelissen, B. J. C, et al, Nature (London) 321, 531-532 (1986) . NP 24 is described in King, Plant Mol. Biol. 10, 401-411 (1988) . Proteins of approximately 22 kDa molecular mass (i.e., similar to zeamatin) and having similar N-terminal amino acid sequence to zeamatin, thaumatin, PR-R and osmotin I and which cross reacted with antizeamatin antiserum are described in Vigers, A. J., et al, Molecular Plant-Microbe Interactions, Vol. 4, No. 4, 315-323 (1991) which suggests the name permatins to describe "this family of membrane-permeabilizing antifungal proteins" and proposes the names sormatin (for the protein isolated from sorghum) , avematin (for the protein isolated from oats) and trimatin (for the protein isolated from wheat) . Vigers, A. J., et al, Plant Science, 83, 155-161 (1992) describes the serological relation of PR-S and osmotin to zeamatin.

The antifungal lipid lytic enzymes include phospholipases and lipases. Phospholipases include phospholipase A (present in honey bee venom) and phospholipase B (available from Sigma Chemical Company. Lipases include Type I lipase (from wheat germ) , Type I-A lipase (insoluble enzyme from wheat germ attached to 4% beaded agarose) , Type II lipase (from porcine pancreas) , lipase from human pancreas, Type VI-S lipase from porcine pancreas, Type VII lipase (from Candida cylindracea) , Type VII-A lipase (insoluble enzyme from Candida cylindracea) , Type XI lipase (from Rhizopus arrhizus) , Type XII lipase (from Chromobacter ium vicosum) and Type XII lipase (from Pseudomonas spp.) ; all these specifically mentioned lipases are available from Sigma Chemical Company.

We turn now to optional adjuvant components of the systems herein. The optional antifungal polyene macrolide antibiotic adjuvants are described in Martin, J.-F., Am. Rev. Microbiol. 31:13-38 (1977) which describes them as having a lactone ring of 26-38 atoms, a polyene chromophore consisting of a series of 4-7 alternating double bonds that form part of the macrolide ring and usually one aminosugar moiety. Polyene macrolide antibiotics include the following which are listed in Martin: acosin, amphotericin B, aureofungin, aytactin, candicidin, candihexin A, candihexin B, candihexin E, candihexin F, chainin, dermostatin, DJ-400 B^ DJ-400 B₂, etuscomycin, eurocidin A, eurocidin B, filipin, flavofungin, fungichromin, hamycin, heptafungin A, levorcin, mycoheptin, nystatin, perimycin, pimaricin, rimocidin, tetrin A, tetrin B and trichomycin. These, when included, are included in the systems herein in a fungus inhibition improving amount. This can be the conventional antifungal amount (dosage) .

The antifungal epithiodiketopiperizine antibiotics include, for example, gliotoxin, gliovirin, chaetomin and sporidesmin. Jones, R. W. , et al, Journal of General Microbiology, 134, 2067-2075 (1988) states that these are characterized as low-M_r, non polar molecules with bridged polysulfide region which confers activity and suggests that the primary mechanism of action of gliotoxin involves selective binding to cytoplasmic membrane thiol groups. Gliotoxin is available from Sigma Chemical Company. These, when included, are included in a fungus inhibition improving amount. For gliotoxin, this can range, for example, from 1 ng/ml to 5,000 ng/ml. The optional fungal cell wall biosynthesis inhibitor adjuvants include chitin synthetase inhibitors and 0-1,3 glucan synthetase inhibitors .

The chitin synthetase inhibitors include, for example, polyoxins A, B, D, E, F, G, H, J, K, L, M, N and 0; kitazin P and nikkomycin Z. The isolations and characterization of polyoxin A and polyoxin B are described in Isano, K. , et al, Biol. Chem. 29, 848 (1965) . The isolations and characterizations of polyoxins D, E, F, G, H, J, K and L are described in Isono, K. , et al, Agr. Biol. Chem. 30, 817 (1966) and 32, 792 (1968) . The isolation and characterization of polyoxin M are described in Isono, K. , et al, Tetrahedron Letters, 1970, 425. The isolation of polyoxins N and 0 are described in Japanese Kokai 72/23,596 (Chemical Abstracts 78:41566t (1973) . Polyoxin B is available from Kaken Chemical Co., Ltd. Kitazin P is available from Kumiai Chemical Industry Co., Ltd. Antifungal usage of Polyoxin B and Kitazin P is mentioned in Watanabe, R. , et al, Agric. Biol. Chem. 52(4) , 895-901 (1988) . Nikkomycin Z is available from Calbiochem and is mentioned in Roberts, W. K., et al, Journal of General Microbiology, 136, 1771-1778 (1990) . The chitin synthetase inhibitors, when included, are included in the systems herein in a fungus inhibition improving amount which for these agents is a chitin synthesis inhibiting amount . An assay for chitin synthetase activity is described in Cabib, E., et al, Chitin Synthase from Saccharomyces cerevesiae, pages 643-649, in Methods of Enzymology, Vol. 138, Ginsburg, V., editor, Academic Press, New York, 1987. Minimum inhibitory concentrations can be determined by including the inhibitor in the assay mixture of the assay for chitin synthetase activity. Minimum inhibitory concentration of polyoxin B against B. cinerea disclosed in Watanabe et al is 12.5 μg/ml. Minimum inhibitory concentrations of kitazin P against B . cinerea disclosed in Watanabe et al is 500 μg/ml. A β- 1, 3-glucan synthetase inhibitor is L-sorbose. The mechanism of action of L-sorbose is discussed in Mishra, N. C, et al, Proc. Nat. Acad. Sci. USA, Vol. 69, No. 2, pp. 313- 317, 2/72. The β- 1, 3-glucan synthetase inhibitors, when included, are included in the systems herein in a fungus inhibition improving amount. L-Sorbose may be included in compositions for systems herein in an amount ranging from 1 to 10%, for example 1% to 3%.

The optional detergent adjuvant component of the systems herein include, for example, non ionic detergents, e.g., sorbitan esters, polyoxyethylene fatty alkyl ethers, polyoxyethylene nonylphenol ethers, dialkyl sulfosuccinates, ethoxylated and propoxylated mono- or diglycerides, acetylated mono- or diglycerides, lactylated mono- or diglycerides, sugar esters, polysorbates and polyglycerol esters. The sorbitan esters include, for example, polyoxyethylene sorbitan monolaurate (Tween 21) , polyoxyalkylene sorbitan monoleate (Tween 20) , polyoxyalkylene sorbitan monooleate (Tween 81) and polyoxyalkylene sorbitan monopalmatate (Tween 40) . An example of a polyoxyethylene fatty alkyl ether is polyoxyethylene lauryl ether which is sold under the tradename Emulgen 120. An example of a polyoxyethylene nonylphenol ether is Emulgen 909. An example of a dialkyl sulfosuccinate is dioctyl sulfo succinate (Pelex OTP) . The optional adjuvant detergent component can also be an anionic detergent, e.g., sodium lauryl sulfate, or a cationic detergent, e.g., trimethyl palmityl ammonium sulfate. The detergents, when included, are included in the systems herein in a fungus inhibition improving amount and this amount depends on the detergent included and can be as low, for example, as a concentration of 0.001% or as high, for example, as a concentration of 1%. Use of detergents in combination with antifungal agent is described in Watanabe, R. , et al, Agric. Biol. Chem. 52 (4), 895-901 (1988) and minimum concentrations at which certain detergents inhibited mycelial growth of P. oryzae are described therein.

As indicated above, the weight ratio of fungal cell wall degrading enzyme to antifungal fungal cell membrane affecting compound is 0.005:1 to 500,000:1, in many cases 2:1 to 500,000:1. Preferred ratios are set forth in Table 1 below wherein "endochit." stands for endochitinase, and "chitobios' stands for chitobiosidase.

Table 1

Combination Preferred ratio

T. harzianum endochit : flusilazole 5,000 :1 to 85, 000 : :1

T. harzianum endochit :miconazole 100 :1 to 1,500 :1

T. harzianum endochit :osmotin 1 :1 to 150: :1

T. harzianum endochit :gramicidin 1.5; :1 to 100: :1 T. harzianum endochit:valinomycin 10: :1 to 350: :1

T. harzianum endochit : phospholipase B 1:30 to 500:1

T. harzianum endochit : Trichorzianine Al 100:1 to 10, 000 : :1 T. harzianum endochit:

Trichorzianine Bl 50:1 to 12, .000 : :1

G. virens endochit : flusilazole 0,000:1 to 475, .000: :1

G. virens endochit :miconazole 250:1 to 4, 000: :1

G. virens endochit :osmotin 1:1 to 100: :1 G. virens endochit :gramicidin 1:7 to 50: :1

G. virens endochit :valinonycin 5:1 to 850: :1

G. virens endochit :phospholipase B 1:100 to 100: :1

G. virens endochit :Trichorzianine Al 10:1 to 10, ,000: :1

G. virens endochit :Trichorzianine Bl 20:1 to 1, ,500 :1

N. tabacum endochit : flusilazole 1,000:1 to 400, ,000 :1

N. tabacum endochit:miconazole 600:1 to 4, ,000: :1

N. tabacum endochit:osmotin 25:1 to 350 :1

N. tabacum endochit:gramicidin 1:50 to 500 :1 N. tabacum endochit :valinonycin 50:1 to 2, ,ooo :1

N. tabacum endochit :phospholipase B 1:1 to 2 ,000 :1

N. tabacum endochit :

Trichorzianine Al 300:1 to 35,000:1

N. tabacum endochit : Trichorzianine Bl 500:1 to 40,000:1

T. harzianum chitobios :flusilazole 2,500:1 to 350,000:1

T. harzianum chitobios :miconazole 400:1 to 3,500:1

T. harzianum chitobios :osmotin 20:1 to 200:1 T. harzianum chitobios:gramicidin 1.5:1 to 200:1

T. harzianum chitobios :valinonycin 100:1 to 2,500:1

T. harzianum chitobios : phospholipase B 5:1 to 3,000:1

T. harzianum chitobios : Trichorzianine Al 150:1 to 15,000:1

T. harzianum chitobios :

Trichorzianine Bl 300:1 to 15,000:1

G. virens chitobios: flusilazole 1,500:1 to 300,000:1 G. virens chitobios:miconazole 250:1 to 3,000:1 Table 1 (Continued)

G. virens chitobios :osmotin 10 :1 to 150:1 G. virens chitobios:gramicidin 1:1 to 500:1 G. virens chitobios :valinonycin 150:1 to 4,000:1

G. virens chitobios :phospholipase B 10:1 to 3,000:1 G. virens chitobios :

Trichorzianine Al 450:1 to 20,000:1 G. virens chitobios : Trichorzianine Bl 350 :1 to 19, 000 :1

S. albidoflavus chitobios : flusilazole 12, 000: :1 to 250,000: :1

S. albidoflavus chitobios: miconazole 300: :1 to 3,000: :1

S. albidoflavus chitobios :osmotin 1::1 to 30::1

S. albidoflavus chitobios :gramicidin 1 :1 to 100: :1

S. albidoflavus chitobios .-valinonycin 5: :1 to 1, 000::1

S. albidoflavus chitobios : phospholipase B 1: :50 to 500: :1

S. albidoflavus chitobios:

Trichorzianine Al 80:1 to 8,000:1 S. albidoflavus chitobios: Trichorzianine Bl 30:1 to 10, 000:1

T. harzianum nagase :flusilazole 7, ,000:1 to 280,000: :1 T. harzianum nagase :miconazole 100:1 to 20, 000: :1 T. harzianum nagase :osmotin 1:1 to 250: :1 T. harzianum nagase :gramicidin 1:1 to 500: :1 T. harzianum nagase :valinonycin 120:1 to 2, 000: :1 T. harzianum nagase :phospholipase B 1:1 to 1,500: :1 T. harzianum nagase :Trichorzianine Al 1:150 to 5,000: :1 T. harzianum nagase :Trichorzianine Bl 250: :1 to 8, 000: :1 T. harzianum glucosidase : flusilazole 7, ,500:1 to 450,000: :1 T. harzianum glucosidase: miconazole 75:1 to 45, 000: :1 T. harzianum glucosidase :osmotin 5:1 to 220: :1 T. harzianum glucosidase:granicidin 1:1 to 300: :1 T. harzianum glucosidase:valinomycin 7 :1 to 1,500: :1 T. harzianum glucosidase : phospholipase B 10:1 to 500 :1 T. harzianum glucosidase :

Trichorzianine Al 70:1 to 9,000: :1 T. harzianum glucosidase:

Trichorzianine : Bl 100:1 to 10,000 :1

;r ranqes for T. harzianum <sndochit.inase :

,000:1 to 85,000:! L and for T.harzianum endochitinase: miconazole are 100:1 to 650:1.

Compositions for use in the systems herein are readily formulated by admixing the fungal cell wall degrading enzymes and the antifungal fungal cell membrane affecting compounds with non-toxic carriers appropriate for the particular use for a composition, e.g., agriculturally acceptable carriers for agricultural uses and pharmaceutically acceptable carriers for medicinal uses. They may be formulated as liquids (solutions or suspensions) or as solids. Since the fungal cell wall degrading enzymes need free water for activity, water must be present at the time of function. This can be accomplished, for example, by applying the fungal cell wall degrading enzymes as aqueous solutions or by formulating the system components as dry powders and applying the powders with the enzyme becoming active once water becomes available, e.g., from rain. Water is a preferred vehicle where components are soluble in it. Osmotin, phospholipase B, trichorzianine Al and trichorzianine Bl are readily dissolved in water. Organic solvents can also be used and may be required in some cases if a solution is desired. Guidance for forming solutions follows. Gramicidin dissolves in ethanol. Miconazole dissolves in 50% (v:v) ethanol. Flusilazole and valinomycin dissolve in acetone. Gliotoxin dissolves in methanol and ethanol. Suspension can also be employed.

We turn now to formulations for application of the optional adjuvant components. Most polyene macrolide antibiotics have poor water solubility and are therefore normally formulated as dispersions or suspensions -for application or applied as a powder. The chitin synthetase inhibitors may also be applied as a powder. Polyoxin B may be dissolved in aqueous ethanol. Kitazin P may be dissolved in a small amount of dimethylformanide and then diluted with water. L-sorbose may be dissolved in water.

We turn now to the method of the invention of inhibiting the germination or growth of a fungus which comprises contacting such fungus or a locus to be protected from such fungus with an antifungal effective amount of a combination of a fungal cell wall degrading chitinolytic or glucanolytic enzyme in a concentration where said enzyme individually provides 2 to 50% inhibition of spore germination and antifungal fungal cell membrane affecting compound which is not chitinolytic or glucanolytic enzyme and which is not expressed by the same organism as the fungal cell wall degrading enzyme in nature in a concentration where said compound individually provides about 4 to 95% inhibition of spore germination, the total of the percentage inhibitions individually provided by the fungal cell wall degrading chitinolytic or glucanolytic enzyme and the antifungal fungal cell membrane affecting compound being less than 100%. The fungal cell wall degrading enzyme and the antifungal fungal cell membrane affecting compound are the same as those discussed in the description of the system of the invention. The effective concentrations are the same as those discussed in the description of the system of the invention. The antifungal effective amount is an amount which inhibits the germination or growth of the fungus that is treated in the method.

For medicinal purposes (i.e., human and veterinary therapy) all the active components can be administered in the same way as the antifungal fungal cell membrane affecting compound is applied when used as the only active ingredient, e.g., topically applied to the skin of a human or non-human animal. Administration can also be, at least in some instances, via parenteral injection, e.g., intraperitoneally; this administration route is particularly useful where the immune system has been compromised since immune-deficient humans and individuals will inactivate enzymatic proteins more slowly than normal individuals.

For agricultural purposes, application can be, for example, to the seed, foliage, roots or fruit of a plant to be protected, or to the soil surrounding said plant, or to the fungus thereon which is to be inhibited. Normally, application is topical. However, other administration strategies can be used. The system and method described above in the detailed description section contemplate application of the fungal cell wall degrading enzyme and the antifungal fungal cell membrane application as part of the same composition or concurrently as part of separate compositions or separately at different times. When the fungal cell wall degrading enzyme and antifungal fungal cell membrane affecting compounds are applied separately at different times, the inhibition obtained is the same as when the two kinds of agents are applied in the same composition or in different compositions but concurrently, when the antifungal fungal cell membrane affecting compound is applied even as much as 8 hours after the fungal cell wall degrading enzyme. On the other hand, application of the antifungal cell membrane affecting compound before the cell wall degrading enzyme results in reduction in the percentage inhibition obtained compared to when the two kinds of agents are applied as part of the same composition or concurrently in separate compositions until the cell wall degrading enzyme and the antifungal fungal cell membrane affecting compound are both in contact with the fungus for at least about 16 hours. The presence of fungal cell wall degrading enzyme is necessary for about 4 to 8 hours for the highest level of synergistic antifungal effect to be obtained.

The system and method described above in the detailed description section inhibit the germination or growth of fungal species from genera including Fusarium, Gliocladium,

Rhizoctonia , Tri choderma , Uncinula , Us tilago, Erysiphe, Botrytis , Saccharomyces , Sclerotium and Al ternaria . The specific examples hereinafter show synergis for said system and method herein, in every instance where the antifungal fungal cell membrane affecting compound is used in a concentration less than that where it is substantially entirely effective alone, in application to Botrytis cinerea , and in application to Fusarium oxysporum, which were selected in the work supporting this invention as model test fungi. Botrytis cinerea is a fungus which is pathogenic to fruits including grapes, raspberries, and apples and to beans and other crops. Fusarium oxysporum is a fungus which is pathogenic, for example, to tomatoes, melons, garden peas, cowpeas, beans, soybeans, alfalfa, flax, cotton, carnations, and tobacco.

The synergistic interaction provided by the system and method described above in the detailed description section allows reduction of the quantity of antifungal fungal cell membrane affecting compound that is required for use for a particular inhibition of fungi as much 100- to 1000-fold and this reduction allows usage of fungicides which are otherwise too highly toxic or produce unacceptable side effects at fungicidal or fungistatic dosages, allows usage at dosages less than those which produce side effects and should reduce or at least delay the occurrence of natural resistance to important chemical fungicides.

We turn now to the embodiment herein directed to a method of protecting from a fungus a plant which expresses fungal cell wall degrading chitinolytic or glucanolytic enzyme at a level of 0.05 to 5% of total cellular protein, said method comprising contacting said plant with an antifungal effective amount of an antifungal fungal cell membrane affecting compound at a concentration where it individually provides about 4 to 95% inhibition of spore germination. The plant is a plant which is susceptible to the fungus being protected against which is transformed to contain gene which expresses fungal cell wall degrading chitinolytic or glucanolytic enzyme in the stated amount or which has been infected with transgenic endomorphic microorganisms producing said fungal cell wall degrading chitinolytic or glucanolytic enzyme, typically in the xylem, to produce enzyme internally in the plant- in the stated amount. The fungi protected against can be, for example, from the genera of pathogenic fungi described above. Genes coding for fungal cell wall degrading chitinolytic or glucanolytic enzymes can be isolated from microorganisms or other organisms producing them. For example, the characterization and isolation of the gene coding for the aforedescribed endochitinase from Trichoderma harzianum strain PI is described in Harman et al U.S. patent No. 5,378,821. Such gene can be inserted into the genome of a plant to be protected, for example, by Agrobacterium- mediated transformation, by biolistic transformation or by other methods known to those skilled in the art. Methods for use for transformation of plants to contain genes are described in Brogue, K. , et al . , Science 254,1194-1197 (1991) ; and in Neuhaus, J.-M., et al . , Plant Molec . Biol. 16,141-151 (1991) ; and in Norelli, J. L., et al . , J. Amer. Soc. Hort. Sci. 118,311-316 (1993) taken with Norelli, J. L., et al . , Euphytica 77,123-128 (1993) ; these articles are incorporated herein by reference. The antifungal fungal cell membrane affecting compounds are those described above and the application of antifungal cell membrane affecting compound to the plant can be carried out as described above . Example XX hereinafter is directed to transforming the above described endochitinase encoding gene from Trichoderma harzianum strain PI into tobacco plants using Agrobacterium to obtain the expression of active enzyme in different parts of the plant in an amount of 1-3% of the total cellular protein and application of antifungal fungal cell membrane affecting compound thereto so that it becomes systemic to act synergistically with the expressed endochitinase in providing fungal inhibition.

We turn now to the embodiment herein directed to a transgenic plant protected against pathogenic fungi which is a plant susceptible to fungal attack which has been transformed to contain gene which expresses fungal cell wall degrading chitinolytic or glucanolytic enzyme at a level of about 0.05 to 5% of total cellular protein and also which also has been transformed to contain gene which expresses protein antifungal cell membrane affecting compound or which has been infected with transgenic endomorphic microorganism producing said protein antifungal fungal cell membrane affecting compound typically in the xylem, in an amount to provide a concentration of said compound where it individually provides about 4 to 95% inhibition of spore germination. The fungi protected against can be, for example, from the genera of pathogenic fungi described above. The transformation to contain gene which expresses fungal cell wall degrading chitinolytic or glucanolytic enzyme in the named amounts is described above. Genes coding for protein antifungal fungal cell membrane affecting compound are described in Kumar, V. , Plant Molec. Biol. 18,621-622 (1992) and in Watanabe, Y. , et al . , FEMS Microbiology Letters 124,29-34 (1994) . Such genes can be inserted into the genome of a plant, for example, as described in the paragraph directly above . Exemplary of this embodiment is a crop plant, e.g., a tobacco plant, transformed to contain gene from Nicotiana tabacum coding for osmotin and which has been transformed to contain the above described endochitinase encoding gene from Trichoderma harzianum strain PI so the expressed osmotin and expressed endochitinase interact synergistically in providing fungal inhibition.

The following reference examples set forth preparation of some enzymes used in the working examples which illustrate the invention.

Reference Example 1 Endochitinase from Gliocladium virens strain 41 having accession No. ATCC 20906 as described above was prepared as set forth below.

Synthetic medium was made up containing 680 mg KH₂P0₄, 870 mg K₂HP0₄, 200 mg KCl, 1 g NH₄N0₃, 200 mg CaCl₂, 200 mg MgS0₄«7H₂0, 2 mg FeS0₄, 2 mg ZnS0₄, 2 mg MnS0₄, 42 g moist purified colloidal chitin (prepared as described in Vessey, J.C, et al, Trans. Br. Mycol . Soc. 60:710-713, 1973), 5 g sucrose, in 1 L distilled water, final pH 6.0.

100 ml of the synthetic medium was placed in a 250 ml Erlenmeyer flask.

The flask was inoculated with conidia grown by inoculation of potato dextrose agar (conidia of Gliocladium virens ATCC 20906) to provide IO⁷ conidia ml^"1 medium and the admixture was incubated at 25oc for 5 or 7 days on a rotary shaker at 200 rpm. The culture filtrate was harvested by centrifugation at 8000 x g for 10 minutes and removal of residual particulates by filtration through a glass fiber filter.

The purified endochitinase was isolated from the culture filtrate as described below with all steps being carried out at 4oc except for concentration which was carried out at room temperature.

The filtered culture filtrate was transferred into dialysis tubing (6,000 to 8,000 Da cutoff) and concentrated 30-40-fold by placing the tubing in solid polyethylene glycol (35,000 MW; Fluka Chemika-Biochemicha, Buchs, Switzerland) . The concentrate was dialyzed overnight against 50 mM potassium phosphate buffer pH 6.7 (5 L buffer L^"1 culture filtrate) and applied to a gel filtration column (5x60 cm) containing Sephacryl S-300 HR (Pharmacia LKB Biotechnology, Uppsala, Sweden) equilibrated with 50 mM potassium phosphate buffer pH 6.7 containing 200 mM NaCl. The material from 1 L of culture medium was chromatographed separately in two samples on Sephacryl S-300 HR. Fractions, approximately 8 ml each, were eluted with 1500 ml of 50 mM potassium phosphate buffer containing 200 mM NaCl. A first peak between fractions 70 and 120 contained high levels of chitobiosidase and N-acetyl-β-D- glucosaminidase activity. A second peak with endochitinase, jS-1, 3-glucanase, and chitobiosidase activity was detected in fractions 120 to 140. Fractions 140 to 160 contained endochitinase activity; proteins in this region were apparently not separated on the basis of molecular weight, but adsorbed to the gel matrix since they eluted at or greater than the total column volume. The fractions 140 to 160 from the first sample and similar fractions from the other, showing only endochitinase activity, were pooled. The pooled fractions (160 ml) were transferred into dialysis tubing

(6,000 to 8,000 Da cut-off) and concentrated 30- to 40-fold by placing the tubing in solid polyethylene glycol (35,000 MW; Fluka Chemika-Bioche ika, Buchs, Switzerland) and dialyzed overnight against a 20-fold volume of 25 mM ethanolamine-HCl buffer pH 8.7. The sample (about 25 ml) was then applied to a chromatofocusing column (1x30 cm) packed with PBE94 (Pharmacia LKB) , and equilibrated with the same buffer used for dialysis. The column was eluted at a flow rate of 50 ml h^"1 with Polybuffer 96 (Pharmacia LKB) diluted 1:10 and - 31 -

adjusted to pH 7.0 with HCI according to the manufacturer's direction. A sharp peak at pH 8.0 containing endochitinase activity was detected in the eluted fractions. The peak fractions were pooled and the pooled fractions (about 40 ml) 5 were dialyzed first against a 20-fold volume of IM NaCl and then against a 40-fold volume of distilled water to remove Polybuffer, and concentrated to a volume of 2 ml in a collodion bag system (10,000 Daj UH 100/1, Schleicher & Schuell Inc., Keene, NH) . The sample (2 ml) , i.e., the 0 concentrated fractions, was applied to compartments 15 and 16 (pH 8.0-8.5) of a Rotofor isoelectric focusing cell (Bio-Rad, Richmond, CA) loaded with 35 ml distilled water containing 2% of Biolyte Ampholytes pH 3-10 (Bio-Rad) , run at 12 W constant power, at a temperature of 40oc, after one hour of prefocusing 5 run, and the run was continued for 5 hours. The fractions (each about 2 ml) were collected and assayed for endochitinase activity. The peak fractions contained homogeneous endochitinase as shown by the presence of a single protein band upon SDS-PAGE and upon Native PAGE. A single fluorescent

20 activity band was observed following overlay of the native gel with the methylumbelliferyl substrate. This activity corresponded to the position of the single protein band detected with coomassie blue and silver stain.

The results of each purification step are summarized in

25 Table 2 below. The endochitinase was purified 105-fold with a recovery of 8%. The quantity of endochitinase produced in the original culture filtrate was calculated to be at least 10 mg L^"1.

30 Table 2

Step Total Enzyme Specific Punfi- Yield

-. _c protein activity activity cation {%) O (mg) (Units) (U mg'l (fold)

40 Crude filtrate 1065 0 10,400 9 7 1 0 100 Dialysis 9,066 47 2 4 β 87

Sephacryl S-300 HR 2,849 371 0 38 2 27 Chromato ocusing 1,984 620 3 63 9 19 Rotofor Cell 0 8 815 101B 5 105 0

45

The peak (active) fractions were pooled, dialyzed against IM NaCl and then against distilled water as described above, and concentrated to dryness in a Speedvac apparatus. The enzyme was stored at -20oC and reconstituted in an appropriate volume of sterilized distilled water for use.

Reference Example 2 Chitobiosidase from Gliocladium virens strain 41 having accession No. ATCC 20906 as described above was prepared as described below. Conidia of the Fungus G. virens strain 41 (ATCC 20906) were stored on silica gel at -20oc and used to inoculate potato dextrose agar (PDA: Difco Laboratories, Detroit, MI) plates. For enzyme production the fungus was grown in 250 ml Erlenmeyer flasks containing 100 ml of a synthetic medium (SMCS) with colloidal chitin and sucrose as carbon sources. SMCS contained 680 mg KH₂P0₄, 870 mg K₂HP0₄, 200 mg KCl, 1 g NH₄N0₃, 200 mg CaCl₂, 200 mg MgS0₄*7H₂0, 2 mg FeS0₄, 2 mg ZnS0₄, 42 g moist purified colloidal chitin, 5 g sucrose, in 1 L distilled water, final pH 6. The flasks were inoculated with conidia collected from freshly grown PDA plates providing 10⁷ conidia ml^"1 medium, and incubated at 25°C for 5 days on a rotary shaker at 200 rpm. The culture filtrate was harvested by centrifugation at 8000 X g for 15 minutes and particulate removed by filtration through a glass fiber filter.

The purified chitin 1, 4-β-chitobiosidase was isolated from the culture filtrate as described below with all steps being carried out at 4oc except for concentration which was carried out at room temperature.

The filtered culture filtrate was transferred into dialysis tubing (6,000 to 8,000 Da cut-off) and dialyzed overnight against 50 mM potassium phosphate buffer pH 6.7 containing 200 mM NaCl (8 L buffer L^"1 culture filtrate) , then concentrated approximately 25-fold by placing the tubing in solid polyethylene glycol (20,000 MW; Fisher Scientific) . The concentrate was applied, in two sample runs, to a gel filtration column (5 X 60 cm) containing Sephacryl S-300 (Pharmacia LKB Biotechnology, Uppsala, Sweden) equilibrated with 50 mM potassium phosphate buffer pH 6.7 containing 200 mM NaCl and 0.05% NaN₃. Fractions, approximately 10 ml, were eluted with the same buffer using reverse gravitational flow at 140 ml h^"1. Fractions were assayed for various enzyme activity. Fractions 85-100, which contained high levels of chitobiosidase, endochitinase and glucanase activities, were pooled and concentrated as described above to approximately 25 ml, then dialyzed overnight against a 20-fold volume of 25 mM ethanolamine-HCl buffer pH 8.7. The sample was applied to a chromatofocusing column (1 X 30 cm) packed with PBE 94 medium (Pharmacia LKB) , and equilibrated with the same buffer used for dialysis. Approximately 8 ml fractions were eluted at a flow rate of 50 ml h^"1 with Polybuffer 96 (Pharmacia LKB) , diluted 1:10 and adjusted to pH 7.0 with HCI according to the manufacturer's directions. Fractions 2-8 contained glucanase activity and fractions 12-16 contained endochitinase activity. The chitobiosidase activity was not eluted from the column at the pH conditions used above. The chitin 1, 4- -chitobiosidase enzyme was removed from the column by eluting with 1 M NaCl. The bulk 1 M chitin 1, 4-3-chitobiosidase fraction was dialyzed against a 20-fold volume of distilled water, concentrated to approximately 10 ml as described above and dialyzed against a 40-fold volume of 25 mM imidazole- HCl buffer, pH 6.7. The sample was applied to the same chromatofocusing column used above that had been equilibrated to pH 6.7 with the imidazole-HCl buffer.

Fractions were collected by eluting with Polybuffer 74

(Pharmacia) , diluted 1:13 and adjusted to pH 3.6 with

HCl . Approximately 8 ml fractions were collected with fractions 12-17 containing chitobiosidase activity. These fractions were pooled, concentrated, dialyzed against a 20-fold volume of 1 M NaCl, then against a 100- fold volume of distilled water. At this point the chitin 1, 4-β-chitobiosidase was determined to be pure by the presence of only a single band using sodium dodecyl sulfate poiyacrylamide gel, native poiyacrylamide gel and isoelectric focusing gel electrophoresis. A single fluorescent activity band was observed following overlay of the native gel with the methylumbelliferyl substrate. The sample was concentrated to dryness in a Speedvac apparatus and stored at -20oc. Reference Example 3

A glucan 1, 3-/^-glucosidase isolated from Trichoderma harzianum strain PI having accession No. ATCC 74058 as described above was prepared as described below.

Crude enzyme solutions were prepared using the growth conditions described for Reference Example 1, except that strain PI of T. harzianum was substituted for G. virens ATCC 20906.

All procedures except concentration steps were carried out at 4° C. Enzyme solutions were concentrated at room temperature. The filtered culture filtrate was transferred into dialysis tubing (6,000 to 8,000 Da cut¬ off) and concentrated 20-fold by placing the tubing in solid polyethylene glycol (35,000 MW; Fluka Chemika- Bioche ika, Buchs, Switzerland) . The concentrate was dialyzed overnight against 50 mM potassium phosphate buffer pH 6.7 (5 L buffer L^"1 culture filtrate) and applied to a gel filtration column (5 x 60 cm) containing Sephacryl S-300 HR (Pharmacia LKB Biotechnology, Uppsala, Sweden) equilibrated with 50 mM potassium phosphate buffer pH 6.7 containing 200 mM NaCl. Fractions (10 ml) were eluted with the same buffer using reverse flow at a rate of 140 ml h^"1. The material from 1 liter of culture medium was chromatographed separately in two samples on Sephacryl S-300 HR, and fractions (numbers 72 to 84 from the first sample, similar ones from the second) containing glucosidase activity were pooled (approximately 230 ml) and concentrated to about 20 ml as described above. They were then dialyzed overnight against a 20-fold volume of 25 mM Tris-CH₃COOH buffer, pH 8.0. The sample was then applied to a chromatofocusing column (1 x 30 cm) packed with PBE 94 (Pharmacia LKB) , and equilibrated with the same buffer used for dialysis. The column was eluted at a flow rate of 50 ml h^'1 with Polybuffer 96 (Pharmacia LKB) , diluted 1:13 and adjusted to pH 7.0 with CH₃COOH according to the manufacturer's directions. Fractions of interest (fractions 13 to 16; 24 ml total) were pooled, dialyzed first against a 20- fold volume of 1 M NaCl and then against a 40-fold volume distilled water to remove Polybuffer, and concentrated to a volume of 2 ml in a collodion bag system (10,000 Da cutoff; UH 100/1, Schleicher & Schuell Inc., Keene, NH) . The enzyme solution was stored at -20oC until use.

The results of the purification from 1 L culture filtrate are summarized in Table 3 below.

Table 3

Step Total Enzyme Specific Purifi protein activity activity cation

(mg) (%)

(nka al) (nkatal mg '} (fold)

Dialyzed culture filtrate 450 1938 4 3 1 0 100 Sephacryl S-300 HR 144 907 6 } 1 5 47 Chromatofocusing 2 3 351 153 36 18

The invention is illustrated by the specific examples set forth below. For the examples, the B . cinerea was B . cinerea strain 12 isolated from grapes by R. Pearson of Cornell University and the F. oxysporum was F. oxysporum strain FOPl isolated from beans in Naples, Italy. For the examples, the osmotin I and endochitinase from Nicotiana tabacum were gifts from R. Bressan of Purdue University and the trichorzianines were a gift from B. Bodo of Paris, France. Example I The antifungal fungal cell membrane affecting compounds tested were flusilazole (E.I. duPont de Nemours) and miconazole (Sigma Chemical Co.) . For the assays herein, the flusilazole was dissolved in acetone and the miconazole was dissolved in 50% ethanol .

The fungal cell wall degrading enzymes tested were endochitinase from Trichoderma harzianum strain PI (ATCC 74058) prepared as described in Harman et al U.S. Patent

No. 5,173,419 and also in S.N. 07/919,784, filed July 27,

1992; the endochitinase from Gliocladium virens strain 41

(ATCC 20906) prepared as described in Reference Example

1; the 40 kDa chitobiosidase from Trichoderma harzianum strain PI (ATCC 74058) prepared as described in Harman et al in S.N. 07/919,784, filed July 27, 1992; and the glucan 1, 3-/3-glucosidase from Trichoderma harzianum strain PI (ATCC 74058) prepared as described in Reference Example 3. The fungal cell wall degrading enzymes tested were dissolved in deionized water.

Assay mixtures were prepared that contained 20 μl of a conidial suspension (IO⁵ to 10⁶ conidia/ml) of the test fungus, which was the plant pathogen Botrytis cinerea , 20 μl of a nutrient solution, potato dextrose broth (Difco Laboratories, Detroit, MI) made at 3 times the standard rate, 18 μl of a fungal cell wall degrading enzyme solution made to the appropriate concentration (for controls, water was substituted) , and 2μl of the solution or suspension of the antifungal fungal cell membrane affecting compound. As controls, the various solvent solutions were tested at the final concentrations employed, but they had no effect on the test fungus.

Mixtures as indicated above were placed in sterile Eppendorf tubes and incubated 24 to 30 hours at 25oC. Portions of the mixtures were placed on a microscope slide and the germination of the first 100 conidia seen were evaluated. Each treatment was done in triplicate in each experiment, and each experiment was repeated. Percent inhibition was calculated according to the following equation: %I = (l-%S_t/%S_c) x 100, where %I represents the percentage inhibition, %S_C represents percentage germination of spores in the treatment of interest, and %S_C represents the percentage of spores germinating in ζhe control (i.e., with neither antifungal fungal cell membrane affecting compounds, nor enzyme) . Appropriate concentrations of the various fungal cell wall degrading enzymes and antifungal fungal cell membrane affecting compounds were evaluated by preliminary experiments which determined the dosage response curve for each substance singly. Concentrations of fungal cell wall degrading enzyme solutions were employed ranging from 0 to 100 μg/ml, and the concentration of antifungal fungal cell membrane affecting compound was chosen that provided about 20% inhibition of spore germination of B . cinerea . The concentration for each chemical is provided in the tables below.

Results for combinations of antifungal fungal cell membrane affecting compound and endochitinase from Trichoderma harzianum strain PI are set forth in Table 4 below.

Table 4

% inhibition of different concentrations of enzyme l g ml

Fungicides 25

Flusilazole (5 ng ml ') 18 99 100 100 100 Miconazole (0 5 μg ml ') 19 100 100 100 100 None 0 29 61 65 73

Results for combinations of antifungal fungal cell membrane affecting compound and endochitinase from Gl iocladium virens strain 41 (ATCC 20906) are set forth in Table 5 below. Table 5

% inhibition at different concentrations of enzyme (<ιq ml'')

Fung cides 50 100

Flusilazole (5 ng ml ') 18 89 100 100 100 Miconazole (0 5 μg ml ') 19 99 100 100 100 None 0 0 0 24

Results for combinations of antifungal fungal cell membrane affecting compound and chitobiosidase from Trichoderma harzianum strain PI (ATCC 74058) are set forth in Table 6 below.

Table 6

% inhibition at different concentrations of enzyme (»q ml ')

Fungicides 25 50 75 100

Flusilazole (5 ng ml ') 18 60 100 100 100

Miconazole (0.5 μg ml ') 19 85 100 100 100

None 0 11 20 24 28

Results for combinations of antifungal fungal cell membrane affecting compound glucan 1, 3-β-glucosidase from Trichoderma harzianum strain PI (ATCC 74058) are set forth in Table 7 below.

Table 7

% inhibition at different concentrations of enzyme (μg ml ^:

Fungicides 25 50 75 100

Flusilazole (5 ng ml"¹) 18 49 63 89 100

Miconazole (0.5 μg ml ') 19 50 70 88 100

None 0 17 32 40

Every combination provided a synergistic interaction. The most appropriate equation to test for synergy was described by Richer (Richer, D.L. Pestic. Sci. 19, 309-315, 1987) as Limpel's formula E_e=X+Y- XY/100, where E_e is the expected effect from additive responses of the chemicals and X and Y are percentage inhibition of the chemicals. Thus, if X provides 20% inhibition and Y provides 30%, the expected additive effect is 20 + 30- (20x30) /100=44%. Any value greater than 44% is evidence of synergy. This equation is based on the consideration that if X kills 20% of the organisms available, then only 80% of the total are available to Y.

The tables above provide values for each antifungal fungal cell membrane affecting compound/fungal cell wall degrading enzyme combination, with the antifungal fungal cell membrane affecting compound used at the ED₂₀ level

(dose effective to cause 20% inhibition when antifungal fungal cell membrane affecting compound is used alone) and fungal cell wall degrading enzyme concentrations ranging from 0 to 100 μg/ml. The levels of synergy shown are substantial; for example, miconazole alone at 0.5 μg/ml gave 19% inhibition and the endochitinase from G . virens alone at 25 μg/ml gave 0% inhibition, while the combination gave 99% inhibition.

Example II The assay procedure, fungal cell wall degrading enzymes, and antifungal fungal cell membrane affecting compounds were the same as those used in Example I . As in Example I, the test fungus was B . cinerea .

Results are presented in Figs. 1 and 2 which are bar graphs showing % inhibition at antifungal fungal cell membrane affecting compound concentrations as recited, in the presence of fungal cell wall degrading enzyme and in the absence of fungal cell wall degrading enzyme, wherein the totally black bars denote no fungal cell wall degrading enzyme, the hatched bars with alternating black and dotted lines denote endochitinase from Trichoderma harzianum strain PI (ATCC 74058) at 10 μg/ml, the bars filled in with dots denote 40 kDa chitobiosidase from Trichoderma harzianum strain PI (ATCC 74058) at 25 μg/ml, the bars hatched with alternating open areas and lines denote glucan 1, 3-?-glucosidase from Trichoderma harzianum strain PI (ATCC 74058) at 25 μg/ml and the bars that are open denote endochitinase from Gliocladium virens strain 41 (ATCC 20906) at 25 μg/ml . In the graphs, error bars indicate standard deviations. The values for inhibition are means across two experiments with three replicates for each experiment.

Results are also set forth in Table 8 below wherein E_e is the expected effect from an additive response according to Limpel's formula expressed as percent inhibition and I₀ is the percent inhibition observed and

"Endoc. (Pl)" stands for endochitinase from Trichoderma harzianum strain PI (ATCC 74058), "Chitob. (PI) " stands for chitobiosidase from Trichoderma harzianum strain PI

(ATCC 74058) , "Glucan. (PI) " stands for glucan 1,3-/3- glucosidase from Trichoderma harzianum strain PI (ATCC 74058) and "Endoc. (4D" stands for endochitinase from Gliocladium virens strain 41 (ATCC 20906) , and each of the fungal cell wall degrading enzymes was used at a concentration of 25 μg/ml. Table 8

Endoc IP1! Chitob (PI) Glucan (PI) Endoc (41) 25 μg ml ' 25 μg ml ' 25 μg ml ' 25 μg ml '

Toxins (concentration) E. I. E. I. E. I. E. I.

Flusilazole (5 ng ml ') 41 99 27 60 22 1 65 IB 67 Miconazole (0 5 μg ml ') 42 5 100 27 9 85 23 70 13 ee

Table 9 below shows ED₅₀ values for endochitinase from Trichoderma harzianum strain PI (ATCC 74058) and for the antifungal fungal cell membrane affecting compounds tested with no endochitinase and with endochitinase at concentrations indicated. ED₅₀ is the dose effective for 50% inhibition.

Table 9

Endochitinase (ED₆₀ of 41 μg ml^"1) (μg ml^'1) Fungicides 10 25

Flusilazole 0.06 0.0007 0.0003 Miconazole 3.0 0.06 0.04

Table 10 below shows ED₅₀ values for chitobiosidase from Trichoderma harzianum strain PI (ATCC 74058) and for the antifungal fungal cell membrane affecting compounds tested with no chitobiosidase and with chitobiosidase at concentrations indicated. Table 10

Chitobiosidase (ED_S0 of 152μg ml^"1) (ctg ml^"1) Fungicides 0 25 50 75 100

Flusilazole 0.06 0.001 0.0004 0.00035 0.0003

Miconazole 3.0 0.06 0.045 0.04 0.03

Table 11 below shows ED₅₀ values for glucan 1,3-/3- glucosidase from Trichoderma harzianum strain PI (ATCC 74058) and for the antifungal fungal cell membrane affecting compounds tested with no glucosidase and with glucosidase at concentrations indicated.

Table 11

Glucosidase (ED₅₀ of 90 μg ml^-1) (μσ ml^"1)

Fungicides 0 25 50 75

Flusilazole 0.06 0.0033 0.001 0.00017 Miconazole 3.0 0.27 0.04 0.0018

Table 12 below shows ED₅₀ values for endochitinase from Gliocladium virens strain 41 (ATCC 20906) and for the antifungal fungal cell membrane affecting compounds tested with no endochitinase and with endochitinase at concentrations indicated. Table 12

Endochitinase (ED₅₀ of 195 μg ml^"1) (ug ml^"1)

Fungicides 0 25 50 75 100

Flusilazole 0.06 0.0025 0.001 0.0045 0.00022 Miconazole 3.0 0.092 0.06 0.04 0.025

In data not depicted in figures or shown in tables, the addition of 1 ng ml^"1 of flusilazole reduced the ED₅₀ value 6.8-fold for endochitinase from Trichoderma harzianum strain PI (ATCC 74058) and 4.6-, 1.3- and 3.9- fold for chitobiosidase and glucanase from Trichoderma harzianum strain PI (ATCC 74058) and endochitinase from Gliocladium virens strain 41 (ATCC 20906) , respectively. The addition of the cell wall degrading enzymes to the preparations of antifungal fungal cell membrane affecting compounds at the concentrations tested increased the occurrence of morphological changes such as lysis of the mycelium and hyphal tips in the germinating spores of B . cinerea .

Example III This experiment tested the effect of endochitinase from Gliocladium virens strain 41 (endoc G) on the inhibition of spore germination of Botrytis cinerea strain 12 by osmotin I.

Assays were carried out by the procedure described in the summary of the invention section. Dose response curves were obtained by probit analysis of data collected from two separate experiments with each experiment involving three treatment replicates.

Fig. 3 shows the results. In Fig. 3, the filled in squares represent data points where endoc G was present in the assay mixture at a concentration of 0 μg/ml, the filled in circles represent data points where endoc G was present in the assay mixture at a concentration of 5 μg/ml and the filled in triangles represent data points where endoc G was present in the assay mixture at a concentration of 10 μg/ml.

As indicated in Fig. 3, the combination of fungal cell wall degrading enzyme and antifungal fungal cell membrane affecting compound inhibit spore germination of B . cinerea more strongly than each agent alone. Example IV

This experiment tested the effect of glucosidase from Trichoderma harzianum strain PI (glucos T) on the inhibition of spore germination of Botrytis cinerea strain 12 by osmotin I. Assays were carried out by the procedure described in the summary of the invention section. Dose response curves were obtained by probit analysis of data collected from two separate experiments with each experiment involving three treatment replicates. Fig. 4 shows the results. In Fig. 4, the filled in squares represent data points where glucos T was present in the assay mixture at a concentration of 0 μg/ml, the filled in triangles represent data points where glucos T was present in the assay mixture at a concentration of 10 μg/ml and the open squares represent data points where glucos T was present in the assay mixture at a concentration of 25 μg/ml.

As indicated in Fig. 4, the combination of fungal cell wall degrading enzyme and antifungal fungal cell membrane affecting compound inhibit spore germination of b . cinerea more strongly than each agent alone.

Example V This experiment tested the effect of endochitinase from Tri choderma harzianum strain PI (endoc T) on the inhibition of spore germination of Fusarium oxysporum strain F0P1 by osmotin I.

Fig. 5 shows the results. In Fig. 5, the filled in squares represent data points where endoc T was present in the assay mixture at a concentration of 0 μg/ml, the filled in triangles represent data points where endoc T was present in the assay mixture at a concentration of 10 μg/ml and the open squares represent data points where endoc T was present in the assay mixture at a concentration of 25 μg/ml.

As indicated in Fig. 5, the combination of fungal cell wall degrading enzyme and antifungal fungal cell membrane affecting compound inhibit spore germination of F. oxysporum more strongly than each agent alone. Example VI

This experiment tested the effect of glucosidase from Trichoderma harzianum strain PI (glucos T) on the inhibition of spore germination of Fusarium oxysporum strain FOP1 by osmotin I . Assays were carried out by the procedure described in the summary of the invention section. Dose response curves were obtained by probit analysis of data collected from two separate experiments with each experiment involving three treatment replicates. Fig. 6 shows the results. In Fig. 6, the filled in squares represent data points where glucos T was present in the assay mixture at a concentration of 0 μg/ml, the filled in triangles represent data points where glucos T was present in the assay mixture at a concentration of μg/ml and the open squares represent data points where glucos T was present in the assay mixture at a concentration of 25 μg/ml . As indicated in Fig. 6, the combination of fungal cell wall degrading enzyme and antifungal fungal cell membrane affecting compound inhibit spore germination of F. oxysporum more strongly than each agent alone. Example VII

This experiment tested the effect of endochitinase from G. virens strain 41 (endoc G) on the inhibition of spore germination of B . cinerea strain 12 by gramicidin.

Fig. 7 shows the results. In Fig. 7, the filled in squares represent data points where endoc G was present in the assay mixture at a concentration of 0 μg/ml, the open triangles represent data points where endoc G was present in the assay mixture at a concentration of 2.5 μg/ml and the filled in circles represent data points where endoc G was present in the assay mixture at a concentration of 5 μg/ml.

As indicated in Fig. 7, the combination of fungal cell wall degrading enzyme and antifungal fungal cell membrane affecting compound inhibit spore germination of B . cinerea more strongly than each agent alone. As further indicated in Fig. 7, gramicidin alone was able to provide 20-25% inhibition of spore germination of B . cinerea and fungal cell wall degrading enzyme was able to improve the inhibition to a level of about 100%. Example VIII

This experiment tested the effect of glucosidase from T. harzianum strain PI (glucose T) on the inhibition of spore germination of F. oxysporum strain FOPl by gramicidin. Assays were carried out by the procedure described in the summary of the invention section. Dose response curves were obtained by probit analysis of data collected from two separate experiments with each experiment involving three treatment replicates.

Fig. 8 shows the results. In Fig. 8, the filled in squares represent data points where glucos T was present in the assay mixture at a concentration of 0 μg/ml, the filled in triangles represent data points where glucos T was present in the assay mixture at a concentration of 10 μg/ml and the open squares represent data points where glucos T was present in the assay mixture at a concentration of 25 μg/ml.

As indicated in Fig. 8, the combination of fungal cell wall degrading enzyme and antifungal fungal cell membrane affecting compound inhibit spore germination of F. oxysporum more strongly than each agent alone. As further indicated in Fig. 8, gramacidin alone was able to provide 20-25% inhibition of spore germination of F. oxysporum and fungal cell wall degrading enzyme was able to improve the inhibition to a level of about 100%.

Example IX This experiment tested the effect of endochitinase from G. virens strain 41 (endoc G) on the inhibition of spore germination of B . cinerea strain 12 by valinomycin.

Fig. 9 shows the results. In Fig. 9, the filled in squares represent data points where endoc G was present in the assay mixture at a concentration of 0 μg/ml, the open triangles represent data points where endoc G was present in the assay mixture at a concentration of 2.5 μg/ml and the filled in circles represent data points where endoc G was present in the assay mixture at a concentration of 5 μg/ml.

As indicated in Fig. 9, the combination of fungal cell wall degrading enzyme and antifungal fungal cell membrane affecting compound inhibit spore germination of B . cinerea more strongly than each agent alone.

Example X This experiment tested the effect of glucosidase from T. harzianum strain PI (glucos T) on the inhibition of spore germination of F. oxysporum strain FOPl by valinomycin.

Fig. 10 shows the results. In Fig. 10, the filled in squares represent data points where glucos T was present in the assay mixture at a concentration of 0 μg/ml, the filled in triangles represent data points where glucose T was present in the assay mixture at a concentration of 10 μg/ml and the open squares represent data points where glucos T was present in the assay mixture at a concentration of 25 μg/ml.

As indicated in Fig. 10, the combination of fungal cell wall degrading enzyme and antifungal fungal cell membrane affecting compound inhibit spore germination of F. oxysporum more strongly than each agent alone. Example XI

This experiment tested the effect of endochitinase from T. harzianum strain PI (endoc T) on the inhibition of spore germination of B . cinerea strain 12 by phospholipase B. Assays were carried out by the procedure described in the summary of the invention section. Dose response curves were obtained by probit analysis of data collected from two separate experiments with each experiment involving three treatment replicates. Fig. 11 shows the results. In Fig. 11, the filled in squares represent data points where endoc T was present in the assay mixture at a concentration of 0 μg/ml, the filled in triangles represent data points where endoc T was present in the assay mixture at a concentration of 10 μg/ml and the open squares represent data points where endoc T was present in the assay mixture at a concentration of 25 μg/ml.

As indicated in Fig. 11, the combination of fungal cell wall degrading enzyme and antifungal fungal cell membrane affecting compound inhibit spore germination of B . cinerea more strongly than each agent alone. As further indicated in Fig. 11, phospholipase B alone was able to provide 15-20% inhibition of spore germination of B . cinerea and fungal cell wall degrading enzyme was able to improve the inhibition to a level of about 100%.

Example XII This experiment tested the effect of chitobiosidase from S . albidoflavus NRRL B-16746 (chitob S) on the inhibition of spore germination of B. cinerea strain 12 by phospholipase B.

Fig. 12 shows the results. In Fig. 12, the filled in squares represent data points where chitob S was present in the assay mixture at a concentration of 0 μg/ml, the open triangles represent data points where chitob S was present in the assay mixture at a concentration of 2.5 μg/ml and the filled in circles represent data points where chitob S was present in the assay mixture at a concentration of 5 μg/ml.

As indicated in Fig. 12, the combination of fungal cell wall degrading enzyme and antifungal fungal cell membrane affecting compound inhibit spore germination of B . cinerea more strongly than each agent alone. As further indicated in Fig. 12, phospholipase B alone was able to provide 15-20% inhibition of spore germination of B . cinerea and fungal cell wall degrading enzyme was able to improve the inhibition to a level of about 100%.

Example XIII This experiment tested the effect of endochitinase from G. virens strain 41 (endoc G) on the inhibition of spore germination of F. oxysporum strain FOPl by trichorzianine Al.

Fig. 13 shows the results. In Fig. 13, the filled in squares represent data points where endoc G was present in the assay mixture at a concentration of 0 μg/ml, the open triangles represent data points where endoc G was present in the assay mixture at a concentration of 2.5 μg/ml and the filled in squares represent data points where endoc G was present in the assay mixture at a concentration of 5 μg/ml.

As indicated in Fig. 13, the combination of fungal cell wall degrading enzyme and antifungal fungal cell membrane affecting compound inhibit spore germination of F. oxysporum more strongly than each agent alone. Example XIV

This experiment tested the effect of nagase from T. harzianum strain PI (NAGas T) on the inhibition of spore germination of F. oxysporum strain FOPl by trichorzianine Al. Assays were carried out by the procedure described in the summary of the invention section. Dose response curves were obtained by probit analysis of data collected from two separate experiments with each experiment involving three treatment replicates. Fig. 14 shows the results. In Fig. 14, the filled in squares represent data points where NAGas T was present in the assay mixture at a concentration of 0 - 51 -

μg/ml, the filled in triangles represent data points where NAGas T was present in the assay mixture at a concentration of 10 μg/ml and the open squares represent data points where NAGas T was present in the assay mixture at a concentration of 25 μg/ml.

As indicated in Fig. 14, the combination of fungal cell wall degrading enzyme and antifungal fungal cell membrane affecting compound inhibit spore germination of F. oxysporum more strongly than each agent alone. Example XV

This example compares the effect of varying the time of application of fungal cell wall degrading enzyme and antifungal fungal cell membrane affecting compound on inhibition of spore germination of B . cinerea . Assays were carried out by the procedure described in the summary of invention section. Time response curves were obtained by probit analysis of data collected from two separate experiments with three treatment replicates in each experiment. The fungal cell wall degrading enzyme was endochitinase activity from Nicotiana tabacum cv. Havana 425 as described above (endoc P) .

The antifungal fungal cell membrane affecting compound was osmotin I . The endoc P was used at a concentration of 32 μg/ml. The osmotin I was used at a concentration of 0.5 μg/ml. At these concentrations, the two agents applied concurrently gave 100% inhibition and the agents applied singly gave low inhibitions. In one experiment, in one run, endoc P and osmotin

I were applied at time 0; in another run, endoc P was applied at time 0 and osmotin I 8 hours later; and in another run, osmotin I was applied at time 0 and endoc P was applied 8 hours later. The results are shown in Fig. 15. In Fig. 15, the open squares denote data points for the run where endoc P and osmotin I were applied at time 0, the filled in triangles denote data points where endoc P was applied at tim 0 and osmotin I was applied 8 hours later and the filled in circles represent data points where osmotin I was applied at time 0 and endoc P was applied 8 hours later. As indicated in Fig. 15, if the antifungal fungal cell membrane affecting compound is applied 8 hours after the application of the fungal cell wall degrading enzyme, the resulting curve is substantially the same as if the two kinds of agents were both applied at the beginning of the assay. On the other hand, when the cell wall degrading enzyme was applied 8 hours after the application of the antifungal fungal cell membrane affecting compound, there is reduction in percentage inhibition obtained compared to when the two kinds of agents were both applied at the beginning of the assay until both kinds of agents have been in contact with the fungus for at least about 16 hours.

Consistent results were obtained for the inhibition of either spore germination or germ tube elongation with 15 different combinations of antifungal fungal cell membrane affecting compound and fungal cell wall degrading enzyme including besides that set forth above, gliotoxin/endochitinase from G. virens strain 41, gliotoxin/glucosidase from T. harzianum strain PI, osmotin I/endochitinase from T. harzianum strain PI, trichorzianine Al/endochitinase from T. harzianum strain PI, trichorzianine Bl/nagase from T. harzianum strain PI, flusilazole/endochitinase from G. virens strain 41 and miconazole/endochitinase from T. harzianum strain PI.

In another experiment, inhibitions obtained at the end of the standard bioassay (24 hours) versus time of application for both agents together, endoc P at time 0 and osmotin I at different times, and osmotin I applied at time 0 and endoc P applied at different times, were determined. The results are set forth in Fig. 16 wherein the open boxes denote data points for application of both compounds together, the filled in triangles denote data points for application of endoc P at time 0 and osmotin I applied at different times and the filled in circles denote data points for application at time 0 and endoc P at different times. Similar curves to those in Fig. 16 were obtained for other antifungal fungal cell membrane affecting compound-fungal cell wall degrading enzyme combinations .

Example XVI Each of 8 antifungal fungal cell membrane affecting compounds were used in combination with each of 8 or 6 fungal cell wall degrading enzymes to provide 50% inhibition (ED₅₀) of B . cinerea strain 12 spore germination. Each fungal cell wall degrading enzyme was used at a concentration giving 10% inhibition of spore germination when used alone in combination with an amount of antifungal fungal cell membrane affecting compound so the combination gave 50% inhibition of spore germination.

The antifungal fungal cell membrane affecting compounds employed were osmotin I, gramicidin, valinomycin, phospholipase B (phospholip in tables below) , trichorzianine Al (trich. Al in tables below) , trichorzianine Bl (trich. Bl in tables below), flusilazole and miconazole.

The fungal cell wall degrading enzymes employed were endochitinase from T. harzianum strain PI as described above (Endoc T below) , endochitinase from G. virens strain 41 as described above (Endoc G below) , endochitinase from Nicotiana tabacum cv. Havana 425 as described above (Endoc P below) , chitobiosidase from T. harzianum strain PI as described above (Chitob T below) , chitobiosidase from G. virens strain 41 as described above (Chitob G below) , chitobiosidase from S . albidoflavus NRRL B-16746 as described above (Chitob S below) , nagase from Trichoderma harzianum strain PI as described above (NAGas T below) and glucosidase from Trichoderma harzianum strain PI as described above (Glucos T below) .

Dose response curves were determined for each antifungal fungal cell membrane affecting compound and for each fungal cell wall degrading enzyme singly. Then each fungal cell wall degrading enzyme at a concentration giving 10% inhibition when used alone was combined with with increasing amounts of antifungal fungal cell membrane affecting compound. Then the concentration of antifungal fungal cell membrane affecting compound causing the ED₅₀ for each combination was calculated by probit analysis of data collected from two experiments with three treatment replicates for each experiment. Assays for inhibition of spore germination were carried out as described in the summary of invention section above .

The relative level of synergism (RS below) for each combination was calculated by applying Limpel's formula: Ee = X+Y- (XY/100) as described in Richer, D. L, Pestic. Sci. 19, 309-315, 1987. In Limpel's formula, X and Y are the percentages of inhibition relative to each compound used alone and Ee is the expected effect for an additional (non-synergistic) effect. If the combination of the two agents produces an observed effect (Eo) greater than Ee, synergism exists, and as the difference between Eo and Ee increases, the level of synergism is greater. In this experiment, the X value (percentage inhibition by fungal cell wall degrading enzyme when used alone was chosen to be 10 (10% inhibition of spore germination when used alone) and the observed effect, Eo, was 50 (i.e., 50% inhibition of spore germination) . The Y values were obtained from the determined concentrations of antifungal cell membrane affecting compound causing the ED₅₀ for a combination and the dose response curves for the antifungal cell membrane affecting compound used alone - i.e., the Y value was the percent inhibition for the antifungal fungal cell membrane affecting compound when used alone for the concentration of antifungal fungal cell membrane affecting compound that was determined to cause the ED₅₀ for the combination. Ee values were then calculated from Limpel's formula. The relative level of synergism, Eo-Ee, is Eo- (X+Y- (XY/100) ) or in this case 50- (10+Y-Y/lO) where Y is as described above. If no synergism exists, the RS value approaches zero while 40 is the highest possible value.

Results for ED₅₀ and RS are given in tables below. The upper and lower 95% fiducial limits for 95% probability for ED₅₀ and RS are given in parenthesis after the value listed. The concentration of each cell wall degrading enzyme used (i.e., corresponding to an ED₁₀ for the enzyme alone) were as follows: Endoc T, 11 μg/ml; Endoc G, 3.5 μg/ml; Chitob T, 25 μg/ml; Chitob G, 29 μg/ml; NAGas T, 17 μg/ml; Glucos T, 20 μg/ml; Endoc P, 32 μg/ml; Chitob S, 4 μg/ml.

The results are given in Table 14 below for the antifungal fungal cell membrane affecting compound osmotin I.

Table 14

mixture of ED₅₀ compounds (μg ml^"1' RS

Osmotin÷ 10.0 (8-12)

Endoc T 0.7 (0.6-0.8) 30. .0 (34.6-27.4)

Endoc G 0.5 (0.2-0.6) 37. .3 (40-34)

Chitob T 1.0 (0.9-1.2) 22. .0 (23.8-21)

Chitob G 3.0 (2.5-3.4) 15. .7 (17.5-13)

NAGas T 0.6 (0.4-0.7) 34, .6 (36-30)

Glucos T 0.3 (0.2-0.5) 40. .0 (40-37.3)

Endoc P 0.3 (0.1-0.5) 40 .0 (40-37.3)

Chitob S 5.0 (4.2-55) 11. .0 (13-10.4)

Thus, for example, the concentration of osmotin for 50% inhibition of spore germination was reduced from 10.0 μg/ml to 0.3 μg/ml by the inclusion of 32 μg/ml of Endoc P.

The results are given in Table 15 below for the antifungal fungal cell membrane affecting compound gramicidin.

Table 15

mixture of ED 50 compounds μ^<3 ml RS

Gramicidin+ >300

Endoc T 25 (20-29) 34.6 (36.4-30.1)

Endoc G 28 (25-30) 31.0 (34.6-29.2)

Chitob T 47 (46-49) 22.9 (24.7-22.1)

Chitob G 60 (57-62) 17.5 (21.2-16.6)

NAGas T 30 (29-31) 29.2 (30.1-28.3)

Glucos T 27 (24-29) 32.8 (35.5-30.1)

Endoc P 47 (44-49) 22.9 (24.7-22.1)

,Chitob S 30 (27-33) 29.2 (32.8-28.3)

The results are given in Table 16 below for the antifungal fungal cell membrane affecting compound valinomycin. Table 16

mixture of ED₅₀ compounds (μg ml^"11 R£

Valinomycin+ 7.3 (7.0-7.6) _

Endoc T 1.0 (0.9-1/1) 40.0 (+0-39.1)

Endoc G 1.5 (1.3-1.7) 36.4 (39.1-35.5)

Chitob T 5.1 (4.8-5.3) 26.5 (28.3-22)

Chitob G 5.5 (5.4-5.7) 23.8 (24.7-18.4)

NAGas T 1.8 (1.7-1.9) 32.8 (35.5-31.9)

Glucos T 1.2 (1.0-1.4) 39.1 (40-37.3)

Endoc P 2.5 (2.2-2.6) 29.2 (31-28.3)

Chitob S 5.6 (5.5-5.7) 19.7 (23.8-18.4)

The results are given in Table 17 below for the antifungal cell membrane affecting compound phospholipase B.

Table 17

mixture of ED₅₀ compounds (μg ml^'1' RS

Phospholip+ >300 _

Endoc T 25 (21-28) 40. .0 (40-40)

Endoc G 40 (36-44) 38. .2 (40-31.9)

Chitob T 55 (51-58) 31. .0 (31.9-29.2)

Chitob G 43 (42-46) 32. .8 (34.6-31.9)

NAGas T 32 (29-34) 40. .0 (40-40)

Glucos T 31 (30-32) 40, .0 (40-40)

Endoc P 120 (107-139) 22, .9 (23.8-18.4)

Chitob S 63 (60-66) 26, .1 (26.5-25.6)

The results are given in Table 18 below for the antifungal fungal cell membrane affecting compound trichorzianine Al . Table ii

mixture of ED₅₀ compounds (μg ml^"11 RS

Trich. A1+ 90.5 (85- -95)

Endoc T 13.0 (11- -14.8) 20, .2 (23- -20)

Endoc G 13.1 (10, .5-15) 20. .0 (23. .8-19)

Chitob T 27.0 (24, .5-29) 15, .3 (17, .5-14)

Chitob G 29.0 (26- -32) 14. .0 (16, .6-13)

NAGas T 22.2 (20- -24) 18, .3 (19- -17)

Glucos T 22.0 (19- -25) 18. .5 (19- -17.5)

Endoc P 29 (26.5- -31) 14, .0 (16. .5-12)

Chitob S 20.1 (18- -22) 19 .1 (18- -18.5)

The results are given in Table 19 below for the antifungal fungal cell membrane affecting compound trichorzianine Bl .

Table 19

mixture of ED 5,0 compounds (μg ml^" RS

Trich. B1+ 95.2 (92-99) Endoc T 25 (23-27) 23 (25.6-221

Endoc G 30 (28-31.5) 19 (21-18)

Chitob T 35 (34-36) 16 (15-17)

Chitob G 36 (33-38.5) 16 (18-15.5!

NAGas T 28 (24-31) 21 (26-18) Glucos T 33 (32-34) 17 (18-17)

Endoc P 52 (50-55) 11 (13-10)

Chitob S 30 (27-31) 19 (22-18)

The results are given in Table 20 below for the antifungal fungal cell membrane affecting compound flusilazole. Table 20

mixture of ED 50 compounds (ng ml -1) RS

Flusilazole+ 68 (61-85]

Endoc T 0.6 (0.4-0.8) 40 (40-39) Endoc G 0.7 (0.7-0.8) 40 (40-39) Chitob T 1.9 (1.6-2.1) 32.8 (34-30) Chitob G 5 (4-6.5) 23 8 (26-20) NAGas T 0.9 (0.8-1) 38 8 (39-37) Glucos T 0.8 (0.7-1) 39 (40-37)

The results are given in Table 21 below for the antifungal fungal cell membrane affecting compound miconazole.

Table 21

mixture of ED₅₀ compounds (ng ml^"11 RS

Miconazole÷ 3.0 (2-4) Endoc T 0.07 (0.06-0.08) 37, (39-36) Endoc G 0.01 (0.01-0.03) 39. (40-39) Chitob T 0.06 (0.04-0.08) 38. (39-36) Chitob G 0.06 (0.05-0.07) 38 (39-36) NAGas T 0.1 (0.06-0.2) 36 (39-31) Glucos T 0.2 (0.1-0.3) 31 (32-30)

Example XVII An experiment was carried out the same as in Example XVI except that the fungus used in the inhibition testing was Fusarium oxysporum strain FOPl .

The concentrations of each fungal cell wall degrading enzyme used (i.e., corresponding to an ED₁₀ for the enzyme alone) were as follows: Endoc T, 10 μg/ml; Endoc G, 5 μg/ml; Chitob T, 30 μg/ml; Chitob G, 33 μg/ml; NAGas T, 12 μg/ml; Glucos T, 14 μg/ml; Endoc P, 38 μg/ml; and Chitob S, 8 μg/ml. The results- are given in Table 22 below for the antifungal fungal cell membrane affecting compound osmotin I .

Table 22

mixture of ED_B0 compounds (ng ml^"11 RS

Osmotin+ 11.0 (8-12)

Endoc T 0.2 (0.1-0.3) 38.0 (40-36.4)

Endoc G 0.2 (0.1-0.4) 38.5 (40-34.5)

Chitob T 0.9 (0.7-1.1) 24.0 (26-21)

Chitob G 1.0 (0.9-1.1) 23.1 (24-21)

NAGas T 0.6 (0.4-0.7) 28.8 (34.5-26)

Glucos T 0.1 (0.1-0.2) 40.0 (40-38.5)

Endoc P 0.1 (0.1-0.1) 40.0 (40-40)

Chitob S 2.3 (2.2-2.5) 8.0 (11.2-6.5)

The results are given in Table 23 below for the antifungal fungal cell membrane affecting compound gramicidin.

Table 23

mixture of ED₅₀ compounds (ng ml^"1' RS

Gramicidin+ >300 Endoc T 30 (26-33) 31 0 (32-22) Endoc G 18 (16-22) 40 0 (40-38) Chitob T 97 (95-100) 12 0 (13-11) Chitob G 102 (97-112; 10 0 (11-5.5) NAGas T 32 (29-35) 22 8 (25-20) Glucos T 26 (23-29) 32 0 (34-25) Endoc P 110 (104- 12i; 6.0 8- 5) Chitob S 27 (23-30) 30.4 (34-24)

The results are given in Table 24 below for the antifungal fungal cell membrane affecting compound valinomycin. Table 24

mixture of ED₅₀ compounds (ng ml^'11 RS

Valinomycin+ 8.5 (8.0-9.0)

Endoc T 1. .5 (1, .4-1.7) 31.9 (34-28.5)

Endoc G 1. .0 (0, .3-1.2) 40.0 (40-38.5)

Chitob T 2. .5 (2, .1-2.7) 25.0 (27-23.5)

Chitob G 3. .0 (2, .9-3.1) 20.4 (21.5-19)

NAGas T 3. .5 (3 .3-3.7) 18.2 (18.5-17)

Glucos T 1, .1 (1, .0-1.3) 40.0 (40-38)

Endoc P 4 .0 (3.9-4.1) 7.6 1 I 9- 6)

Chitob S 1, .6 (1, .5-1.7) 30.0 (31.9-28.5)

The results are given in Table 25 below for the antifungal fungal cell membrane affecting compound phospholipase B.

Table 25

mixture of ED_S0 compounds (ncr ml^"1' RS

Phospholip+ >300 _

Endoc T 29 (26-31) 40, .0 (40-38.5)

Endoc G 35 (31-38) 37, .5 (38.5-34)

Chitob T 49 (45-55) 30, .5 (31-27)

Chitob G 65 (57-70) 20. .4 (22-19)

NAGas T 40 (35-44) 32. .8 (37.5-31.2)

Glucos T 49 (40-55) 30, .5 (32.8-27)

Endoc P 70 (67-79) 19, .0 (19.5-18.5)

Chitob S 44 (38-50) 31, .2 (34-29) The results are given in Table 26 below for the antifungal fungal cell membrane affecting compound trichorzianine Al .

Table 26

mixture of ED₅₀ compounds (ng ml^'11 RS

Trich. Al 87 (83-91) Endoc T 10 (9-11.8) 17. .5 (23-15.5) Endoc G 3 (2.5-3.5) 30..5 (33-28.5) Chitob T 12 (10.5-13. 5) 15..8 (18-13) Chitob G 15 (11-19) 10,.2 (16- 7) NAGas T 14 (12-16) 13,.2 (15.8-9.5) Glucos T 8 ( 7- 9) 27,.5 (30-23) Endoc P 12 (10.5-14) 15.2 (18-13.2) Chitob S 11 (10-13) 16,.0 (18-13.5)

The results are given in Table 27 below for the antifungal fungal cell membrane affecting compound trichorzianine Bl .

Table 27

mixture of ED₅₀ compounds (ng ml^"11 RS

Trich. Bl 95.5 (92-99) _

Endoc T 15 (10-19) 16.0 (23-13)

Endoc G 10 ( 8-11.5) 23.0 (25.5-21)

Chitob T 22 (20-24) 10.0 (12-8.5)

Chitob G 24 (21-27) 8.5 (11-7.5)

NAGas T 20 (18-22) 12.0 (14-10)

Glucos T 11 (10-13) 22.0 (23-21.5)

Endoc P 20.5 (17-24) 11.4 (15- 8.5)

Chitob S 20 (16-24) 12.4 (16- 8.5) The results are given in Table 28 below for the antifungal fungal cell membrane affecting compound flusilazole.

Table 28

mixture of ED₅₀ compounds (ng ml^"1' RS

Flusilazole+ 59.0 (51-65) Endoc T 0.3 (0.1-0.5) 40 .0 (40-39) Endoc G 0.1 (0.1-0.2) 40..0 (40-40) Chitob T 1.0 (0.8-1.2) 34,.0 (37-30) Chitob G 4.0 ( 3- 4.5) 28,.5 (29-27) NAGas T 0.9 (0.8 - 1) 35,.5 (37-34) Glucos T 0.7 (0.6-0.8) 38,.5 (39-37)

The results are given in Table 29 below for the antifungal fungal cell membrane affecting compound miconazole.

Table 29

mixture of ED₅₀ compounds (ng ml^{" 1}' RS

Miconazole+ 5. .0 (4.5-5.5)

Endoc T 0. .1 (0.07-0.2) 38. .0 (39. .5-36. 5 )

Endoc G 0. .06 ; (0.05-0.07) 40, .0 (40- -40)

Chitob T 0. .6 (0.4-0.8) 33, .0 (34- -27.5)

Chitob G 0, .7 (0.6-0.8) 28, .5 (33- -27.5)

NAGas T 0 .4 (0.3-0.5) 34 .0 (35, .5-33)

Glucos T 0, .3 (0.1-0.5) 35, .5 (38- -33)

Example XVIII This example shows the effect of the application of L-sorbose on the inhibition of spore germination and germ tube elongation of B . cinerea .

L-Sorbose was applied together with each of the fungal cell wall degrading enzymes mentioned in Examples XVI and XVII. It did not improve the activity of the enzymes except in the case of glucosidase from T. harzianum strain PI where the ED₅₀ (dose effective to inhibit 50% of spore germination) was lowered about 1.5 fold.

The following applications were made to B . cinerea :

(a) osmotin I alone; (b) osmotin I plus 1% sorbose; (c) osmotin I plus 3% sorbose; (d) osmotin I plus 1% L- sorbose and 10 μg/ml glucosidase from T. harzianum strain PI; and (e) osmotin I plus 1% L-sorbose and 2.5 μg/ml of endochitinase from G. virens strain 41. The sorbose concentrations are percentage of the reaction mixture on a weight/volume basis. The results for (a) - (e) are shown in Fig. 17 wherein the filled in squares denote the data points for (a) ,- the open triangles denote the data points for (b) ; the filled in circles denote the data points for (c) ; the filled in triangles denote the data points for (d) ; and the open squares denote the data points for (e) .

Similar curves were obtained for other antifungal fungal cell membrane affecting compounds (flusilazole, gramicidin, valinomycin and trichorzianine Al were additionally tested) together with L-sorbose and for antifungal fungal cell membrane affecting compounds together with L-sorbose and fungal cell wall degrading enzymes.

Similar curves were obtained when inhibition of germ tube elongation was determined rather than inhibition of spore germination.

Synergistic, in both inhibiting spore germination and in inhibiting germ tube elongation, were the combinations L-sorbose-antifungal fungal cell membrane affecting compound and L-sorbose-antifungal cell membrane affecting compound-fungal cell wall degrading enzyme. Instances of Ee and Eo values for the inhibition of spore germination were 17 and 50%, respectively, for the combination L-sorbose-trichorzianine Al; and 10 and 50%, respectively, for the combination L-sorbose- trichorzianine Al-glucosidase from T. harzianum strain PI. Example XIX

Data for inhibition of germ tube elongation was consistent with data for inhibition of spore germination obtained in Examples I-XV.

Example XX The gene (ThEn42) encoding endochitinase from T. harzianum strain PI from genomic DNA, and whose exon sequence is provided in U.S. Patent 5,378,821, was present as an insert in the plasmid pCRII (the original plasmid is available commercially from Invitrogen Corporation, San Diego, CA) . This sequence was removed from the pCRII vector by restriction digestion and ligated into pBI121 (Clontech Laboratories, Palo Alto, CA) . Plasmid pBI121 contains the cauliflower mosaic virus 35S promoter which drives expression in higher plants, the NPT II gene for kanamycin resistance necessary for selection of transformed plants, and also a gene encoding -glucuronidase. This plasmid was digested with the restriction enzymes BairiH I and FcoR I; this linearized the plasmid and excised the gene encoding 3-glucuronidase. ThEn42 was obtained from the plasmid containing it by the following procedure : The plasmid was linearized by restriction with BamΑ I, since pCRII contains unique site just upstream from the ThEn42 gene. There were two restriction sites for EcoR I in the plasmid. One was very near the 5' end of the endochitinase gene, in the untranslated region, and the other was just downstream from the 3' end of the gene. Therefore, a partial digest with EcoR I was done of the linearized plasmid, and the fragment that was the size of the full gene, including the untranslated region, was obtained following electrophoresis and electroelution. The isolated ThEn42 gene and the linear pBI121 plasmid therefore had both BcoR I and BamU I sticky ends. The gene and linear plasmid were then mixed and ligated together, and the orientation of the sticky ends was such that only correct orientation of the gene relative to the 35S promoter was possible. Once the ligation was complete, the presence of the gene in the recircularized plasmid was verified by polymerase chain reaction with appropriate primers. Similar procedures could be used to place ThEn42 in other plasmids behind other promoters known to those skilled in the art. The plasmid so obtained (p35S-ThEn42) was used to transform either E. coli to maintain it, or into Agrobacterium tumifaciens for plant transformation.

The plasmid so obtained can be used to transform a variety of plants to contain and express ThEn42 . This specific example deals with tobacco. Agrobacterium- mediated transformation (Agrobacterium tumefaciens strain LBA4404 was used) of Nicotiana tabacum var Samsum NN leaf disks was carried out as described in the following reference (Horsh, R.B., Fry, J.E., Hoffman, N.C. Eichhltz, D., Rogers, S.G. and Fraley, R.T. 1985. A simple and general method for transferring genes into plants. Science. 227: 1229-1231) . Kanamycin resistant plantlets were scored for their ability to form roots in two consecutive steps of propagation in Murashige-Skoog

(MS) medium containing 3% (w/v) sucrose and lOOmg per liter of kanamycin sulfate, and full plants were regenerated from those which formed roots .

Expression of active endochitinase can be detected in extracts from transformed plants using standard enzyme assays (see U.S. Patent 5,173,419) using either the methylumbelliferyl or the nitrophenyl derivatives of N,N' ,N" triacetylchitotriose. The expression of the transgenic endochitinase was demonstrated in transformed tobacco by assay with the methylumbelliferyl substrate. The expressed endochitinase was found to be present in an amount of 1-3% of total cellular protein. The enzyme so expressed in the plant can then be admixed with membrane active compounds if such compounds are taken up and translocated within the plant. As an example, flusilazole is formulated into the commercial fungicide NuStar (E. I duPont Co., Newark, DE) . Therefore, transgenic plants can be sprayed with the fungicide either according to the label rates, or preferably at 0.5 or 0.1 times the recommended rate. Once plants transgenic plants are treated with the fungicide, the enzyme and fungicide will be present as a synergistic mixture within the plant .

Results of the examples show the reduction of the quantity of the antifungal fungal cell membrane affecting compounds necessary to obtain any level of inhibition when such compounds were used in combination with the fungal cell wall degrading enzymes tested.

Trichoderma harzianum strain PI was deposited on May 20, 1991 at the American Type Culture Collection and was assigned accession number ATCC 74058. Gliocladium virens strain 041 was deposited on Oct. 24, 1988 at the American

Type Culture Collection and was assigned accession number

ATCC 20906. Streptomyces albidoflavus NRRL B-16746

(also known as strain 10) was deposited on February 28,

1996 at the American Type Culture Collection and was assigned accession number ATCC 55742. The American Type Culture Collection is located at 12301 Parklawn Drive, Rockville, MD 20852, U.S.A. Nicotiana tabacum cv. Havana 425 was deposited in the USDA Plant Introduction inventory in 1961 under accession number PI 552350 and is maintained in the Tobacco Collection (Curator: Verne A. Sisson, Crop Science Department, North Carolina State University, Oxford Tobacco Research Station, P.O. Box 1555, Oxford, North Carolina 27565) .

Variations in the invention will be obvious to those skilled in the art. Therefore, the invention is defined by the claims.

Claims

WHAT IS CLAIMED IS:

1. System for inhibiting the germination or growth of a fungus, said system comprising (a) fungal cell wall degrading chitinolytic or glucanolytic enzyme, (b) antifungal fungal cell membrane affecting compound which is not expressed by the same organism as the fungal cell wall degrading enzyme in nature, in an amount to provide a concentration where it provides about 4 to 95% inhibition of spore germination when it is used without (a) ; the weight ratio of (a) to (b) being 0.005:1 to 500, 000 :1.

2. System as recited in claim 1 comprising an amount of said fungal cell wall degrading chitinolytic or glucanolytic enzyme to provide a concentration of said enzyme where it provides 2 to 50% inhibition of spore germination when it is used without antifungal fungal cell membrane affecting compound and an amount of said antifungal fungal cell membrane affecting compound to provide a concentration of said compound where it provides 10 to 70% inhibition of spore germination when it is used without fungal cell wall degrading chitinolytic or glucanolytic enzyme and the total of the percentage inhibitions provided by the fungal cell wall degrading chitinolytic or glucanolytic enzyme and the antifungal fungal cell membrane affecting compound when each is used without the other, is less than 100%.

3. System as recited in claim 2 comprising an amount of said fungal cell wall degrading chitinolytic or glucanolytic enzyme to provide a concentration of said enzyme where it provides 5 to 20% inhibition of spore germination when it is used without antifungal fungal cell membrane affecting compound and an amount of said antifungal cell membrane affecting compound to provide a concentration of said compound where it provides 15 to 60% inhibition of spore germination when it is used without fungal cell wall degrading chitinolytic or glucanolytic enzyme.

4. The system of claim 1 additionally comprising an inhibition improving amount of a fungal cell wall biosynthesis inhibitor.

5. A method for protecting from a fungus a plant which expresses fungal cell wall degrading chitinolytic or glucanolytic enzyme at a level of about 0.05 to 5% of total cellular protein, said method comprising contacting said plant with an antifungal effective amount of an antifungal fungal cell membrane affecting compound at a concentration where it individually provides about 4 to 95% inhibition of spore germination.