A Hydrophobin of the Chestnut Blight Fungus, Cryphonectria parasitica, Is Required for Stromal Pustule Eruption

Hydrophobins are abundant small hydrophobic proteins that arepresent on the surfaces of many filamentous fungi. The chestnutblight pathogen Cryphonectria parasitica was shown to producea class II hydrophobin, cryparin. Cryparin is the most abundantprotein produced by this fungus when grown in liquid culture.When the fungus is growing on chestnut trees, cryparin is foundonly in the fungal fruiting body walls. Deletion of the geneencoding cryparin resulted in a culture phenotype typical ofhydrophobin deletion mutants of other fungi, i.e., easily wettable(nonhydrophobic) hyphae. When grown on the natural substrateof the fungus, however, cryparin-null mutation strains wereunable to normally produce its fungal fruiting bodies. Althoughthe stromal pustules showed normal development initially, theywere unable to erupt through the bark of the tree. The hydrophobincryparin thus plays an essential role in the fitness of thisimportant plant pathogen by facilitating the eruption of thefungal fruiting bodies through the bark of its host tree.

INTRODUCTION

Top

Introduction

The exterior surfaces of plants are formidable barriers to pathogens.Most plant pathogens are unable to directly penetrate thesebarriers, and those that can penetrate these barriers have evolvedremarkable strategies for doing so. For example, Plasmodiophorabrassicae (1) penetrates the host wall by puncturing it witha specialized structure called the stachel. Magnaporthe grisea,the rice blast pathogen, produces another type of specializedstructure, the appressorium, which penetrates the plant cuticlewith a peg that develops unusually high hydrostatic pressures(26). To exit a plant, the same barriers that met the pathogenon entry must once again be crossed. Some pathogens, such asthe late-blight pathogen Phytophthora infestans, sporulate byexiting natural openings on the plant surface. Little is known,however, of how those fungi that sporulate by erupting throughhost tissues are able to breach the plant surface barriers.It is known that microbial pathogens must be able to exit theirhosts and gain access to new hosts in order to successfullycomplete the disease cycle.

The search for pathogenicity factors of fungal plant pathogenshas focused largely upon understanding how the fungus invadesits host (8, 13, 20, 29). Much less attention has been paidto the role of sporulation as a potential target for controlof disease. From our study of cryparin, a cell surface hydrophobinof the plant pathogen Cryphonectria parasitica, we report theinvolvement of this protein in the normal egress of the fungusfrom infected trees. This protein thus acts as a pathogenicityfactor since it is necessary for the successful completion ofthe pathogen's disease cycle.

The ascomycete C. parasitica is one of the most devastatingplant pathogens of recorded history; it has essentially eliminatedits host, the American chestnut, from its natural range. Thisfungus enters the tree through wounds and, in the process ofcolonizing the wound, forms a hyphal fan with which the fungusinvades healthy tissues (11). After colonizing the bark tissues,a stroma is formed directly beneath the outermost layer of bark,which is comprised of a combination of dead and dying tree tissuesand fungal hyphae. Within this stroma, the stromal pustules,also known as pycnidia, and perithecia develop. These structuresare the asexual and sexual fruiting bodies, respectively, ofthe fungus. These fruiting bodies erupt through the bark andrelease large numbers of spores into the environment.

Hydrophobins are widely found in fungi and lichens but havenot yet been reported from other organisms (30, 34). Hydrophobinsare a group of small, secreted hydrophobic cell surface proteinsthat have highly conserved properties but limited sequence conservation;they are similar, however, in each having eight cysteines locatedwithin the protein sequences in conserved positions, and theyare all highly surface-active molecules. Cryparin is a classII hydrophobin, a classification that is based on the ease ofits isolation from cell surfaces and on the relative spacingof four of its conserved cysteine residues (30). Antibodiesprepared against cryparin demonstrated that it is found exclusivelyin the fungal fruiting body walls when the fungus is growingon its natural substrate, but when grown in artificial culture,cryparin is also found on undifferentiated hyphal cell surfaces(5).

Although there has been a conservation of structure and chemicalproperties in the evolution of hydrophobins, there has not beena conservation of function: all hydrophobins are located onhyphal surfaces and confer hydrophobic properties to the surfaces,but they appear to have evolved different functions in variousfungi (33). The functions include being involved in penetrationof host surfaces by appressoria of the rice blast pathogen M.grisea (23), protecting the human opportunistic pathogen Aspergillusfumigatus against attack by macrophages (18), improving thedispersal of conidia of A. nidulans (22) and Cladosporium fulvum(31), and producing aerial hyphae of Schizophillum commune (27)and facilitating the attachment of these hyphae to solid surfaces(35).

To investigate the role of cryparin in the biology of C. parasitica,mutants of the fungus were produced by directed mutagenesisto produce strains that lack a functional cryparin gene. Theresults demonstrate that cryparin is required for the fruitingbodies of this fungus to erupt through host bark tissue. Althoughstromal pustules develop normally in strains lacking cryparin,they are unable to erupt through the bark.

MATERIALS AND METHODS

Top

Materials and Methods

Strains. Strain EP155 (ATCC 38755) of C. parasitica was used for thegene deletion experiments. This strain is virulent on chestnuttrees, and its mating type is MAT-2. Strain EP44 (ATCC 22511),which was used in sexual crosses with a deletion mutant of EP155,is also virulent, but its mating type is MAT-1.

Vector construction. The vector p ${Delta}$ Crp, which was used to delete the cryparin gene,was constructed from the genomic clone of the cryparin gene(37) and the vectors pDH25 (7) and pBT3 (17). A 1.1-kb fragmentbetween the XhoI and NcoI sites of the coding region of thecryparin gene (crp) was replaced with a 3.8-kb fragment containingthe bacterial hygromycin phosphotransferase gene (hph) (7).This construct retained portions of the 5' and 3' flanking sequencesfrom the cryparin gene (Fig. 1A). A second selectable marker,the gene encoding a benomyl-resistant form of Neurospora crassaß-tubulin (tub-2) (17), was added to the vector asshown in Fig. 1B.

View larger version (15K):
[in this window]
[in a new window]
FIG. 1. Construction of cryparin deletion vector p ${Delta}$ Crp and rescue vector pWT6. (A) Plasmid pCrp contains the entire crp gene on a 3.6-kb EcoRI fragment that was inserted into the polylinker region of pBSKS–. The solid bar represents the coding region, a 1.1-kb fragment between the XhoI and NcoI sites. (B) Plasmid p ${Delta}$ Crp was constructed by replacing the 1.1-kb XhoI/NcoI fragment containing the crp gene coding region with a 3.8-kb cassette containing the bacterial hygromycin phosphotransferase (hph) gene of E. coli under control of the A. nidulans trpC promoter and terminator (7) while retaining portions of the 5' and 3' flanking regions of the crp gene. In addition, a 2.5-kb fragment containing the benomyl resistance gene tub-2 from N. crassa (17) was blunt end ligated into the KpnI site upstream of the 5' flanking crp sequence. (C) For wild-type restoration, plasmid pWT6 was constructed by the insertion of tub-2 upstream from the intact 3.6-kb genomic sequence of pCrp.

A vector, pWT6, was constructed to complement the mutation createdby the deletion of crp. This vector was constructed by removingthe tub-2 from pP₂P₃ß (25) and inserting it into theClaI site upstream of the genomic clone of the cryparin genein vector pCrp (36) (Fig. 1C).

Mutation of the cryparin gene. Methods for spheroplast preparation, transformation, and selectionfor drug-resistant transformants were previously described (6).Transformants were initially selected for hygromycin resistanceon PDAmb (Difco potato dextrose agar supplemented with methionine[100 mg/liter] and biotin [1 mg/liter]) (2) and with 50 µg/mlof hygromycin (Calbiochem, La Jolla, CA). Those colonies thatconsistently grew on media containing hygromycin were screenedfor sensitivity to PDAmb containing 0.5 µg/ml of benomyl(ChemService, West Chester, PA). Transformants exhibiting hygromycinresistance and benomyl sensitivity were then screened for theproduction of cryparin by extraction of lyophilized myceliafrom individual colonies with 60% ethanol. The proteins in theextract were separated using polyacrylamide gel electrophoresis(PAGE) (5), and detection of cryparin on the gels was done byprotein staining and comparison of the gels with a cryparinstandard. Deletion of cryparin from those isolates that didnot express cryparin was confirmed using Southern and Northernblot analyses, as described previously (36).

Cryparin antibody production. Polyclonal antibody was prepared in rabbits against expressedHis tag-purified cryparin. The 800-bp cryparin cDNA (37) wascloned into the XhoI/EcoRI site of pRSET A (Invitrogen Corporation,Carlsbad, CA). Expressed protein was purified according to manufacturer'sprotocols. Resulting antibodies were purified against C. parasiticastrain ${Delta}$ 119 (cryparin deletion strain) to remove nonspecificcross-reacting proteins by suspending 0.1 g of lyophilized myceliain 10 ml of Tris-buffered saline and 0.2% Tween 20. The suspensionwas homogenized using a HandiShear AC (Gardiner, NY) on setting5 for 1 min after which 100 µl of undiluted antibody wasadded and incubated at room temperature for 5 h with gentleagitation. The suspension was centrifuged for 2 min at 5,000x g. The supernatant was removed and stored at –20°Cuntil use.

Protein extraction from conidia. Conidia were collected from plates 3 weeks after inoculationand then lyophilized (Virtis, Gardiner, NY). The lyophilizedconidia were suspended in 100% trifluoroacetic acid (TFA), incubatedat room temperature for 1 h, and then sonicated (Branson, Shelton,Conn.) for 1 min (9). The extracted spores were pelleted bycentrifugation at 12,000 x g for 2 min. The supernatant wasdiscarded, and the pellet was air dried overnight. The pelletwas resuspended in 60% ethanol and with an equal volume of 2xPAGE loading buffer containing 4x glycerol. The sample was separatedon a 15% PAGE gel for 1.5 h at 20 mA and stained with Coomassieblue.

For Western blot analysis of proteins of the conidia, the sporeswere collected and lyophilized as described above, and the driedspores were suspended in 200 µl buffer A (0.1 M MES [2-(N-morpholino)ethanesulfonicacid], pH 6.5, 1 mM EDTA, 0.5 mM MgCl₂, 1 mM dithiothreitol,0.5 mM phenylmethylsulfonyl fluoride, and 1 µg/ml leupeptine)(14) and homogenized with 0.5-mm glass beads in a Mini-BeadBeater (Biospec Products, Inc., Bartlesville, OK) for 50 s.The homogenate was centrifuged at 20,000 x g for 1 min, and50 µl was removed and retained as the soluble fraction.The pellet was further processed by adding 150 µl of 2xTris-Tricine sample buffer (0.1 M Tris, pH 6.8, 24% glycerol,8% SDS, 0.2 M dithiothreitol, and 0.02% Coomassie blue) (3).This pellet fraction was homogenized in the bead beater for50 s, boiled for 5 min, and centrifuged at 20,000 x g for 1min, and the supernatant was retained. Equal volumes of 2x Tris-Tricinebuffer were added to all samples. Proteins were separated byPAGE and blotted onto nitrocellulose membranes (Osmonics, Inc.,Minnetonka, MN) according to manufacturer's directions for proteindetection using antibodies.

Phenotype analysis of mutant strains. Asexual sporulation was quantified as described previously (36).Germination of conidia was tested after 3 weeks of storage at4°C and 22°C. The resistance of conidia to UV radiationwas assessed by measuring the percentage of conidia that germinatedafter treatment compared to that of controls (19). Conidia wereirradiated with 300 Jm^–2 or 600 Jm^–2 at 254 nm usinga Stratalinker UV Cross-Linker 2400 (Stratagene, La Jolla, CA)and then plated onto PDAmb in the dark. Virulence was measuredon chestnut trees as described previously (36).

Stromal pustule production on chestnut stems was measured byinoculating sterile pieces of chestnut stems containing barkthat were approximately 7 cm long by 1.5 cm wide and split inhalf lengthwise. These stems were placed in disposable petriplates (9 by 2.5 cm) containing 40 ml 3% Bacto agar supplementedwith methionine (100 mg/liter) and biotin (1 mg/liter). Plugsof mycelia that were 0.25 cm in diameter were removed from PDAmbplates using a standard cork borer and placed on either sideof the stem pieces. These plates were incubated under low fluorescentlight at 26°C. Stromal pustule production and eruption throughthe bark was assessed 90 days after inoculation. Numbers ofpycnidia produced on malt extract agar were counted by inoculating2% malt extract agar with plugs of mycelia as described previously(36).

RESULTS

Top

Results

Cryparin-null mutants. Over 1,000 transformants that were resistant to hygromycin werescreened for resistance to benomyl. Of these, 200 that weresusceptible to benomyl were tested for lack of cryparin productionand collapse of aerial mycelia when treated with water. Fivetransformants had no detectable cryparin protein production;initial Northern and Southern blots (data not shown) indicatedthat the cryparin gene was disrupted. All showed Southern blotevidence of multiple ectopic integrations of the deletion vector,so one of these strains, 194-7, was sexually crossed with EP44to obtain progeny with a single insertion of p ${Delta}$ Crp.

One hundred and forty-three progeny of the sexual cross between194-7 and EP44 were screened for resistance to hygromycin andthe lack of cryparin production. A recombinant ( ${Delta}$ 119) was foundthat had the expected restriction endonuclease digestion patternof a single insertion at the cryparin gene locus (Fig. 2). Thecryparin gene of EP155 is contained on a 3.6-kb EcoRI fragmentshown in Fig. 2A and B, in the lanes containing EP155 DNA. Bothpanels contain DNA digested with EcoRI and probed with the 5'and 3' region, respectively, of the cryparin gene (Fig. 2C).When probed with both regions of the cryparin genomic clone,digested ${Delta}$ 119 DNA showed the absence of this 3.6-kb EcoRI fragmentcontaining the coding region of the cryparin gene. In Fig. 2A,the EcoRI digestion of ${Delta}$ 119 DNA showed a 2.8-kb band when probedwith the 5' flanking region of the cryparin gene (probe A).This band contains the 1.4-kb EcoRI/XhoI 5' flanking sequenceof cryparin as well as the cloned portion of the hph gene upto the EcoRI site at approximately 1.4 kb into the hph gene(Fig. 1B). The original deletion transformant, 194-7, whichserved as a parent to ${Delta}$ 119, showed this band and also multipleectopic insertions of the transforming vector. In Fig. 2B, thesame EcoRI digestion of ${Delta}$ 119 yielded a 1.2-kb fragment with homologyto the 1.0-kb 3' flanking region of crp and 200 bp of the 3'end of hph (Fig. 1B). Probing the parental strain 194-7 withthis 3' flanking portion of the DNA shows the same 1.2-kb bandas well as ectopic insertions. Additional restriction enzymes,PstI and NcoI, were used to further characterize the flankingregions of the insertions (data not shown).

View larger version (28K):
[in this window]
[in a new window]
FIG. 2. Southern analysis of deletion mutants. Panels A and B contain EcoRI-digested DNA of the crp deletion mutation strain ${Delta}$ 119, the parental deletion mutation strain 194-7, and wild-type strain EP155, respectively. The DNA in panel A was probed with a 1.2-kb EcoRI/PstI fragment containing the 5' end of the genomic cryparin clone (probe A). The DNA in panel B was probed with a 600-bp NcoI/BamHI fragment downstream of the coding region of the gene (probe B).

Absence of cryparin from conidia. Since both class I and class II hydrophobins are often presenton spores (12), attempts were made to isolate cryparin fromthe spores of EP155 and the cryparin deletion strain ${Delta}$ 119. Becauseour standard method of cryparin isolation from mycelia using60% ethanol failed to detect the hydrophobin on conidia, wealso tried the trifluoroacetic acid method used to isolate classI hydrophobins (9) from fungal tissues. This method also failedto isolate cryparin from the conidia (data not shown). Totalprotein was also isolated from conidia by homogenization withglass beads and boiling in SDS. Western analysis was performedusing cryparin antibody, and the results again failed to detectany cryparin in the conidia (Fig. 3).

View larger version (17K):
[in this window]
[in a new window]
FIG. 3. Western analysis of conidial proteins. Lyophilized spores were homogenized and separated into soluble and insoluble fractions. These fractions, along with mycelial controls, were blotted and probed with polyclonal anti-CRP (cryparin) (upper panel) and anti-GPDH (glyceraldehyde-3-phosphate dehydrogenase) (lower panel). Lane 1, size markers are protein standards (in kDa). Lanes 2 and 3 contain soluble conidial proteins from EP155 and ${Delta}$ 119, respectively. Lanes 4 and 8 contain undiluted mycelial proteins of ${Delta}$ 119. Lanes 5 to 7 contain 10^–2, 10^–3, and 10^–4 dilutions, respectively, of EP155 mycelial proteins. Lanes 9 and 10 contain insoluble conidial proteins from EP155 and ${Delta}$ 119, respectively.

Phenotype of cryparin-null mutant strains. We found no evidence that deletion of the cryparin gene affectedthe number of conidia, viability of conidia, or resistance ofconidia to UV irradiation (data not shown). There was also nosignificant change in virulence of ${Delta}$ 119 compared with EP155 whenthey were inoculated into chestnut trees (data not shown). Whengrown on PDAmb solid medium, however, deletion strains showedthe hydrophilic hyphae phenotype typical of other hydrophobindeletion mutants (4, 21, 22, 27). Normally, the surface of fungalcolonies growing on agar medium is so hydrophobic that dropsof water added to the surface remain as discrete drops, whereaswhen water is applied to hydrophobin deletion mutant strains,the water is quickly absorbed. Another observable phenotypeassociated with deletion of cryparin is that older areas ofaerial mycelia collapse with age. To confirm that the phenotypewas caused by the mutation, this mutation was complemented bytransformation of the mutant strain using vector pWT6, whichcontains an intact cryparin gene. The wild-type phenotype ofsurface hydrophobicity was restored when the mutation was complemented(data not shown).

There was no observable or quantitative difference in developmentor morphology of pycnidia of the null mutant strain comparedwith EP155 when grown on malt agar (data not shown). However,when the strains were inoculated onto sterile chestnut wood,strain ${Delta}$ 119 had significantly fewer stromal pustules that haderupted through the bark than EP155 (Fig. 4A). When the infectedwood was sectioned, we observed that the pustules of ${Delta}$ 119 hadformed under the bark but were unable to break through it (Fig.4B). Stromal pustules of the deletion mutant erupted only wherethe bark was broken or missing. In addition, when grown as afemale and crossed with the opposite mating type, ${Delta}$ 119 producedperithecia only beneath stromal pustules that had erupted throughthe bark (data not shown). To further test that the lack ofstromal pustule eruption was associated with the deletion ofcryparin, four other cryparin deletion strains were tested forthis phenotype. Because these disrupted strains also had multipleectopic copies of the deletion vector inserted into their chromosomes,mutant phenotypes other than that for cryparin deletion couldoccur. Each of them, however, showed the same inability to eruptthrough the bark of chestnut stems, as was observed for strain ${Delta}$ 119 (data not shown). The wild-type phenotype of erupted stromalpustules was restored in strain WT6, created by complementingthe cryparin-null mutation with an intact cryparin gene (Fig.4A and B).

View larger version (58K):
[in this window]
[in a new window]
FIG. 4. Stromal pustule eruption on chestnut wood. Panel A shows stromal pustule eruption on stem pieces of sterile chestnut wood 90 days after inoculation with strains EP155, ${Delta}$ 119, and WT6. The magnification is x6.4, and the bar represents 10 mm. Panel B shows free-hand cross-sectional areas of the individual stromal pustules from panel A that were excised from these stem pieces. These cross-sectional areas were magnified (x40), and the bar represents 1 mm.

DISCUSSION

Top

Discussion

Cryparin is produced in abundance during growth of C. parasiticain artificial culture and is largely responsible for the extensivefoaming of the culture fluid when the fungus is grown in liquidshake culture. Previous antibody localization experiments, however,showed that cryparin was present on undifferentiated hyphalcell surfaces only when the fungus was grown in artificial culture.When growing on its natural host, chestnut trees, the proteinwas detectable only in fruiting bodies (5). In those proteinlocalization studies using antibody techniques for protein visualizationat the light microscope level, we were unable to eliminate thepossibility that cryparin is also found on spores, as has beenreported for many hydrophobins (10, 12, 18, 22, 30-33). In thisstudy, we demonstrated that cryparin is not present on the asexualspores. Using extraction procedures that have successfully dissolvedboth type I and type II hydrophobins (33), we were unable todetect cryparin in the spore extracts using sensitive antibodydetection methods. These results were also consistent with ourfinding that a null mutation of cryparin created by site-directedmutagenesis did not affect asexual spore viability in storageor after UV irradiation.

The antibody visualization results that showed the presenceof cryparin in the fruiting body walls (5) are consistent withthe results of this study which clearly show that this proteinis necessary for the eruption of the fungal fruiting bodiesthrough the outer layer of bark of infected trees. Five differentnull mutant strains were tested, and each was found to formwhat appear to be normal pycnidia within the stroma, but thesefruiting bodies were unable to erupt through the bark (Fig.4). Only those few pycnidia that were able to erupt throughwounds or lenticels in the bark formed perithecia. Since thesenull mutant strains all contained multiple ectopic copies ofportions of the knockout vector, sexual crosses were done usingone strain to isolate a mutant strain ( ${Delta}$ 119) with a single insertionat the cryparin gene locus. Analysis of Southern blots clearlyshowed that the cryparin gene was knocked out (Fig. 2). An intactgene encoding cryparin was reintroduced into this strain, andthe ability of the fruiting bodies to erupt was restored bythis complementation experiment (Fig. 4). The cryparin geneis thus necessary for the fruiting bodies of this fungus toerupt through bark, the natural substrate of the fungus.

Other minor phenotypic effects of the null mutation were noticedwhen the fungus was grown in artificial culture: the aerialhyphae of the fungus were wettable, as reported for other hydrophobin-nullmutations (4, 21, 22, 27), and the hyphae of null mutants wereeasier to cut with transfer needles than were those of the wild-typeor complemented strains. Our studies have not yet identifieda mechanism whereby cryparin is able to facilitate the eruptionof the fruiting bodies through the host's bark, but we assumethat it must either strengthen the fruiting body walls againstthe compression that would occur during a physical eruptionor provide structural strength to the fruiting body sufficientfor hydrostatic pressure to facilitate eruption through thebark. A hydrophobic surface and a strengthening of the fruitingbody walls may both be the result of the binding of cryparinto the walls, the result being that the fruiting body has thestructural strength necessary to push through the bark. Theclass I hydrophobin SC3 has been shown to affect hyphal wallcomposition in S. commune (28). Additional studies are necessaryto determine how cryparin alters the properties of the fungalcell and fruiting body walls after it binds to them (15).

The eruption of fungal fruiting bodies through the surface layersof the plant for dissemination of spores is vital to the reproductionand spread of the pathogen. The outer bark layers of a treeprovide a formidable barrier to this process of egress. Few,if any, pathogens are able to directly penetrate through intactbark of woody plant species. Most, including C. parasitica,require wounds in the bark for entry into the host. Bark, similarly,is a major barrier to the eruption of fruiting bodies, eventhough this eruption may be facilitated by the enzymatic degradationof host tissues that occurs during the colonization process(16). Our results indicate that such digestion of host tissueby C. parasitica was insufficient to allow eruption of the fruitingbodies.

We tested the cryparin-null mutant strain for its ability toinvade chestnut trees from wounds and found that the null mutantgrew and formed cankers in the trees that were indistinguishablefrom those of wild-type strains. These results show that cryparinis not necessary for growth of the fungus in its host. Hydrophobinsof a few plant pathogens have been shown to play important rolesin various aspects of pathogenesis (23, 24, 31, 32). The hydrophobinMPG1, produced by the plant pathogen Magnaporthe grisea, isnecessary for appressorium formation as well as conidiogenesisand spore coat formation and thus can be described as a pathogenicityfactor important for fungal infection (23). Unlike MPG1, deletionof cryparin does not visibly disrupt development of any tissueproduced by the fungus.

The hydrophobin most closely related to cryparin is producedby the Dutch elm disease pathogen Ophiostoma ulmi (4). Cerato-ulminis closely related to cryparin at the amino acid level, andboth hydrophobins have similar properties that are typical oftype II hydrophobins (4, 37). Genetic evidence indicates thatcerato-ulmin is not involved in virulence of O. ulmi, as wasoriginally proposed. Cerato-ulmin, however, has been found toaccumulate on the surface of yeast-like cells, the main propagulesof the fungus (24), while the conidia of C. parasitica do notcontain cryparin. The genetic mutants lacking the respectivehydrophobins cryparin and cerato-ulmin had similarly alteredcolony morphology; i.e., older regions of the colony lackedaerial hyphae (4). It is postulated that cerato-ulmin aids inthe dispersal of yeast-like cells by causing them to stick tothe surface of the beetles that vector Dutch elm disease. Itwas also demonstrated that cerato-ulmin protects these cellsfrom desiccation. Because of these biological functions, ithas been suggested that cerato-ulmin is a parasitic fitnessfactor (24).

Cryparin plays a unique role in the biology of C. parasitica,and its role in the biology of this fungus is one that has notpreviously been reported for any protein. Its absolutely essentialrole in the eruption of the fungal fruiting bodies through thebark of its host makes this protein essential to the normallife cycle of this fungus. Without cryparin, spread of the diseasefrom host to host would not be possible or would be greatlyreduced. This study also expands the list of known functionsof hydrophobins in fungi, functions that are remarkably differentfor a group of proteins with conserved properties and structuralsimilarities.

ACKNOWLEDGMENTS

This research was supported by the California Agricultural ExperimentStation.

We thank Lei Zhang for her help with the Southern analysis andLynn Epstein for critical review of the manuscript.

FOOTNOTES

* Corresponding author. Mailing address: Department of Plant Pathology, University of California—Davis, One Shields Ave., Davis, CA 95616. Phone: (530) 754-5500. Fax: (530) 752-5674. E-mail: pjkkaz{at}ucdavis.edu.

REFERENCES

Top