Estimating polyketide metabolic potential among nonsporulating fungal endophytes of Vaccinium macrocarpon

Mycol. Res. 106 (4) : 460–470 (April 2002).

# The British Mycological Society

460

DOI : 10.1017\S095375620200566X Printed in the United Kingdom.

Estimating polyketide metabolic potential among nonsporulating fungal endophytes of Vaccinium macrocarpon

Michelle SAUER1, Ping LU1, Rajinder SANGARI1, Sarah KENNEDY1, Jon POLISHOOK1, Gerald BILLS2 and Zhiqiang AN1* " Merck Research Laboratories, RY 80Y-225, PO Box 2000, Rahway, NJ 07065, USA. # Centro de InvestigacioT n BaT sica, Merck Sharp & Dohme de Espang a SA, Josefa ValcaT rcel 38, Madrid 28027, Spain. E-mail : zhiqiangIan!merck.com Received 29 July 2001 ; accepted 8 December 2001.

A set of 23 non-sporulating, unidentifiable endophytic fungi associated with wild Vaccinium macrocarpon were examined for their polyketide metabolite producing potential. Using two degenerate polymerase chain reaction primers (FPKSKSU-2 and FPKSKSD-1), we cloned and sequenced 12 ketosynthase domains from 11 of the 23 cranberry endophytic fungi. Phylogenetic analyses segregated the 12 ketosynthase domains into three groups. One group of four sequences was clustered with polyketide synthase genes involved in melanin formation. The second group of two ketosynthase sequences clustered with aflatoxin encoding fungal polyketide synthases. The remaining six ketosynthase domains were not clustered with any of the known fungal polyketide synthase groups. Of the 12 ketosynthase fragments, five contained one or more introns in the " 800 bp DNA region. In order to locate the phylogenetic origin of the polyketide synthase genes, phylogenetic relationships of the strains were inferred from small subunit ribosomal DNA sequences. Analyses of small subunit ribosomal DNA sequences showed that all but one of the strains grouped among the major clades of the ascomycetes. The exceptional strain, CR70, was probably an oomycete. Thirteen of the 22 ascomycetous fungi appeared within a clade that included Oidiodendron tenuissimum of the Myxotrichaceae.

INTRODUCTION Plant endophytic fungi are known to produce secondary metabolites and often are targets of searches for bioactive secondary metabolites in pharmaceutical and agrochemical discovery programs (Boddy & Griffith 1989, Dombrowski et al. 1992, Petrini et al. 1992, Dreyfuss & Chapela 1994). Although some endophytic fungi are easily cultured and readily manipulated in laboratory fermentations, others are slow-growing and can be difficult to grow in laboratory- or factory-scale fermentors. Fungi isolated directly from healthy living plants can be cultivated from vegetative cells that are quiescent in host tissues. Such fungi often produce only vegetative hyphae when cultured in vitro, and therefore do not display the morphological taxonomic characteristics necessary to permit identification, limiting prediction of their secondary metabolic profiles. Although many endophytic fungi are prolific producers of secondary metabolites, selection of the fungi for secondary metabolite screening remains an arbitrary * Corresponding author.

process because of the absence of rational selection criteria. Studies are needed to determine whether phylogenetic information can be used to predict which groups of fungi are rich in secondary metabolic pathways. To explore the relationship between phylogenetic diversity and secondary metabolic pathways, we selected a test set (23 strains) of unidentifiable endophytic fungi. Strains from healthy Vaccinium macrocarpon (large cranberry) were chosen because plants of the Ericaceae are rich in endophytic fungi, many of which are host-specific (Petrini, Stone & Carroll 1982, Petrini 1984, Widler & Mu$ ller 1984, Petrini 1987, Dalpe! , Litten & Sigler 1989). With the exception of pathogenic species found in production bogs, few studies have been done characterizing fungi associated with cranberry plants (Shear, Stevens & Bain 1931, Farr et al. 1989), and the diversity of fungi associated with wild populations of V. macrocarpon has not been established. Our goal was to evaluate the taxonomic range and diversity of some nonsporulating endophytes of V. macrocarpon plants from southern New Jersey bogs. Isolates that lacked distinctive sporulation and that were generally con-

M. Sauer and others sidered too slow-growing for batch metabolite production fermentations were selected for further molecular genetic analysis. For the sample set of slowgrowing, nonsporulating strains, small subunit (SSU) rDNA nucleotide sequences were analyzed to elucidate their phylogenetic relationships relative to reference sequences from taxa of major fungal groups. A similar approach using the internal transcribed spacers (ITS) of rDNA was recently used to characterize the ericoid mycorrhizal endophytes of Woollsia pungens (Chambers, Liu & Cairney 2000). Phylogenetic diversity alone may not predict the genetic potential for secondary metabolite production. During the last decade, genes encoding major classes of fungal secondary metabolites have been cloned and characterized, most notably, genes involved in polyketide biosynthesis (Wang, Reeves & Gaucher 1991, Mayorga & Timberlake 1992, Bedford et al. 1995, Chang et al. 1995 ; Feng & Leonard 1995, Takano et al. 1995, Yu & Leonard 1995, Fujii et al. 1996, Yang et al. 1996 ; Takano et al. 1997 ; Ehrlich et al. 1998 ; Tsai et al. 1998, Bingle, Simpson & Lazarus 1999, Fulton et al. 1999, Kennedy et al. 1999, Proctor et al. 1999). Polyketides are perhaps the largest family of secondary metabolites produced by fungi. The primary structure of the genes encoding polyketide synthases (PKS) are conserved. Therefore, DNA probes can be designed based on cloned PKS genes to predict the existence of polyketide pathways of a fungus. Using PKS probes designed for this study, we determined whether the slow-growing and nonsporulating cranberry endophytes possess polyketide pathways. Since the SSU sequence for these fungi also was determined, we investigated their phylogenetic relationships of these fungi as determined by their rDNA genes and PKS genes. MATERIALS AND METHODS

461 was an abandoned commercial bog. Leaves were cut in half, and stems and roots were cut into pieces 3–5 mm long. Root, stem, and leaf pieces were surface sterilized by an initial wetting in 95 % ethanol for 1 min, followed by immersion in 66 % laundry bleach (Chlorox, Oakland, CA) for 2 min, and rinsing in 95 % ethanol for 30 s. Surface-sterilized tissues were placed on plates of YM agar (20 g agar, 10 g malt extract, 2 g yeast extract, 10 mg cyclosporin, 50 mg streptomycin sulfate, and 50 mg chlortetracycline 1−") and water agar (20 g agar, 50 mg streptomycin sulfate, and 50 mg chlortetracycline l−"). Some root samples were processed by repeated washing and centrifugation (4000 g, 5 min). Roots were washed free of debris with water, diced to 3 mm pieces with a scalpel, and placed into a sterile test tube. Sterile water was added and roots were shredded further with a hand-held homogenizer (Biospec Products, Bartlesville, OK). The homogenized root suspension was placed in a 50 ml centrifuge tube. Sterile H O was # added to bring the total volume to 40–50 ml. Tubes were centrifuged for 5 min at 4000 g. The supernatant was discarded, the pellet resuspended in sterile H O, # and the procedure repeated 4 times. After the fourth centrifugation, the pellet was resuspended in sterile H O. Aliquots (0n1 and 0n2 ml) were plated onto YM # agar plates and Bandoni’s sorbose plates (4 g -sorbose, 50 mg yeast extract, 15 g agar, 50 mg streptomycin sulfate and 50 mg chlortetracycline l−"). Plates were incubated at 15 mC. Fungi were dissected from tissues as they grew from the host material and transferred to YM slants. Individual isolates were later transferred to oatmeal agar, corn meal agar, YME agar (10 g malt extract, 4 g yeast extract, 4 g dextrose, and 20 g agar l−"), and cranberry decoction agar (200 ml cranberry fruit extract, 15 g agar l−") for morphological identification. Isolates that failed to produce recognizable spore-forming structures after 4–6 weeks were considered to be nonsporulating.

Strains, media and reagents Yeast extract, malt extract, oatmeal agar, and corn meal agar were from Difco Laboratories (Detroit, MI). All remaining media components and chemicals, except where specified, were obtained from Sigma (St Louis, MO) or Fisher Scientific (Pittsburg, PA). The Escherichia coli strain JM109 was from Promega (Madison, WI). E. coli colonies were grown on Luria-Bertani (LB), LB-ampicillin or LB-ampicillin X-gal\IPTG solid media as described (Sambrook et al. 1989). All the fungal strains reported in this study have been deposited in the Merck Culture Collection in Rahway, NJ ; they are available to other researchers upon request. Isolation of fungi from cranberry plant tissues Wild cranberry (Vaccinium macrocarpum) plants were collected from three bogs near Chatsworth in southern New Jersey. Two of the sites were natural bogs and one

DNA procedures To obtain mycelium for genomic DNA extraction, the isolates were grown on a layer of cellophane (dialysis membrane, Bel-Art, Pequannock, NJ) on top of YME agar. Total genomic DNA was isolated either by a phenol-chloroform extraction procedure (Bruns, Fogel & Taylor 1990, Byrd et al. 1990) or by using a DNeasy Plant Mini kit (Qiagen, Valencia, CA). PCR reactions were carried out with Ready-To-Go PCR beads according to manufacturer’s instructions (Amersham Pharmacia Biotech, Piscataway, NJ). PCR products were cleaned using the QIAquick PCR Purification kit (Qiagen). Sequence reactions were performed using an ABI Big Dye Terminator Cycle Sequencing Ready Reaction kit (PE Applied Biosystems, Foster City, CA). Nucleotide sequences were determined using an ABI 377 DNA Sequencer (PE Applied Biosystems).

Polyketide synthase ketosynthase domains of endophytic fungi

462

Fig. 1. Phylogenetic relationship of 22 strains of non-sporulating cranberry endophytic fungi and 56 reference fungi based on nucleotide sequence from the nuclear small (18S) ribosomal gene. The 22 strains of nonsporulating cranberry endophytic fungi are in bold. The phylogenetic tree presented was generated from parsimony analysis (Swofford 2000) and is one of 3240 most parsimonious trees. Pneumocystis carinii and Protomyces inouyei were used as an outgroup. Of the 1610 characters, 396 were parsimony-informative, 1012 characters were constant, and the remaining 202 variable characters were parsimony-uninformative. The frequencies with which the clades were represented
50 % in 500 bootstrap replications (Felsenstein 1988) are indicated on their respective branches. The total length of the tree was 1804. The consistency index (CI), excluding uninformative characters, was 0n3879. The retention index (RI) was 0n6661.

Plasmid DNA was extracted using QIAprep Miniprep kits (Qiagen). rDNA nucleotide sequencing DNA encoding small subunit ribosomal RNA (SSU rDNA) was amplified by PCR with primers NS1-NS8 (White et al. 1990) and SR2, SR4, SR5, SR7R, SR9R, and SR10R (Spatafora, Mitchell & Vilgalys 1995). PCR conditions were : 94 m, 2 min ; 30i(94 m, 1 min, 55 m, 1 min, 72 m, 1 min) ; and 72 m, 5 min. Nucleotide

sequence homology searches were performed using BLAST (Altschul et al. 1990). To construct phylogenetic trees, corresponding sequences from representative genera of major classes of ascomycetes were obtained from GenBank (Fig. 1). Accession numbers for the SSU rDNA sequence of the 56 reference fungi are : Ajellomyces capsulatus, X58572 ; Alternaria alternata, U05194 ; Auxarthron zuffianum, U29395 ; Blumeria graminis, L26253 ; Botryosphaeria rhodina, U42476 ; Bulgaria inquinans, AJ224362 ; Candida albicans, X53497 ; Capronia pilosella, U42473 ; Ceramothyrium

M. Sauer and others

Table 1. Published fungal PKS genes used in primer designa, probeb, and tree constructionc. Fungus

Gene

Product

Activity

Symbol

Accession No.

Reference

Aspergillus fumigatus A. nidulans A. nidulans

alb1abc pksSTabc wAabc

Dihydroxynaphthalene Sterigmatocystin Naphthopyrone

Pigment Carcinogenic mycotoxin Pigment

AFALB1 ANST ANWA

AF025541 L39121 X65866

A. parasiticus A. parasiticus A. parasiticus A. parasiticus A. parasiticus A. terreus A. terreus A. terreus A. terreus Cochliobolus heterostrophus Colletotrichum lagenarium Gibberella fujikuroi G. fujikuroi Nodulisporium sp. Penicillium patulum P. griseofulvum P. griseofulvum Phoma sp. Phoma sp.

pksL1abc pksL2abc pksAc wAc wAc atXabc pksMc LDKSc LNKSc pks1ac pks1abc fum5c GFPKS4c pks1abc msasabc pks2abc wAc msasc wAc

Aflatoxin

Carcinogenic mycotoxin

Aflatoxin

Carcinogenic mycotoxin

6-methyl salicylic acid

Antibiotic

Lovastatin (partial) Lovastatin (partial) T-toxin Dihydroxynaphthalene Fumonisin Bikaverin Dihydroxynaphthalene 6-methyl salicylic acid

HMG-CoA reductase inhibitor HMG-CoA reductase inhibitor T cytoplasm corn toxicity Melanin Mammalian toxicity Pigment\antibiotic Melanin Antibiotic

APPKSL1 APPKSL2 APPKSA APWA2 APWA1 ATATX ATPKSM ScPKS NPKS CHPKS1 CLPKS1 GFFUM5 GFPKS4 NSPKS1 PPMSAS PGPKS2 PGWA PSMSAS PSWA

L42766 U52151 Z47198 AJ132276 AJ132275 D85860 U31329 AF141925 AR003668 U68040 D83643 AF155773 AJ278141 AF151533 X55776 U89769 AJ132274 AJ132278 AJ132277

Tsai et al. 1998 Yu & Leonard 1995 Mayorga & Timberlake 1992, Watanabe et al. 1999 Feng & Leonard 1995 Feng & Leonard 1998 Chang et al. 1995 Bingle et al. 1999 Bingle et al. 1999 Fujii et al. 1996 Pazoutova et al. 1997 Kennedy et al. 1999 Vinci et al. 1998 Yang et al. 1996 Takano et al. 1995 Proctor et al. 1999 Fulton et al. 1999 Beck et al. 1990 Bingle et al. 1999 Bingle et al. 1999 Bingle et al. 1999

463

Polyketide synthase ketosynthase domains of endophytic fungi

464

Table 2. Primer pairs used in the amplification of KS domains from each of the 10 published fungal PKS genes. Fungus

Strain

Gene

Primer name

Sequence

Colletotrichum lagenarium

ATCC10534

pks1

Nodulisporium sp.

MF5331

pks1

Aspergillus nidulans

MF5999

wA

A. fumigatus

MF5669

alb1

A. nidulans

MF5999

pksST

A. parasiticus

ATCC15517

pksL1

A. terreus

ATCC20542

atX

A. parasiticus

ATCC15517

pksL2

Penicillium patulum

MF4096

msas

P. griseofulvum

ATCC10120

pks2

CLPKS1a CLPKS1b NSPKS1a NSPKS1b ANWAa ANWAb AFALB1a AFALB1b ANSTa ANSTb APPKSL1a APPKSL1b ATATXa ATATXb APPKSL2a APPKSL2b PPMSASa PPMSASb PGPKS2a PGPKS2b

5h-CGTATCAACTACCACTTTG 5h-GGCATCGTCCAGGCGCTTG 5h-CGGATCAATTACCACTTC 5h-GGCGTCATCAAGGCGCTTG 5h-CGGATCAACTATTACTTC 5h-AGCATCTTCAAGCCGCTTC 5h-CGTATCAACTACTACTTC 5h-AGCATCCTCCAGGCGCTTC 5h-CGGATCAATTTCTGCTTTG 5h-TGCATCCTCAAGACGTTTG 5h-CGCATTAACTTCTGTTTCG 5h-AGCATCTTCCAGCCGTTTG 5h-CGCATCTCCTACCACCTG 5h-GGCGGTGGACAGCCGTTTC 5h-CGAATATCTTACCATCTG 5h-TGCTTCTGCCATGTTTTTAAG 5h- CGCATCTCATATCACCTCAAC 5h- TGCACGGTGCAGACTTTTG 5h-CGCATATCATATCTGCTAG 5h-AGCTTTTTCCAGACGCTTTAG

linnaeae, AF022715 ; Chaetomium elatum, M83257 ; Cladonia bellidiflora, U60900 ; Coniosporium sp., Y11712 ; Cudonia confusa, Z30240 ; Debaryomyces hansenii, X58053 ; Dendryphiopsis atra, AF053731 ; Dipodascopsis uninucleata, U00969 ; Dothidea insculpta, U42474 ; Eremascus albus, M83258 ; Eurotium rubrum, U00970 ; Exophiala mansonii, X78480 ; Geomyces pannorum var. asperulatus, AB016174 ; Geosmithia lavendula, D14405 ; Gyromitra esculenta, Z30238 ; Hypocrea lutea, D14407 ; Hypomyces chrysospermus, M89993 ; Inermisia aggregata, Z30241 ; Kirschsteiniothelia elaterascus, AF053728 ; Lecanora dispersa, L37535 ; Leotia lubrica, L37536 ; Leptosphaeria bicolor, U04202 ; Leucostoma persoonii, M83259 ; Lophiostoma crenatum, U42485 ; Microascus cirrosus, M89994 ; Monascus purpureus, M83260 ; Morchella elata, L37537 ; Mycoarachis inversa, AB012953 ; Mycocalicium albonigrum, L37538 ; Neurospora crassa, X04971 ; Oidiodendron tenuissimum, AB015787 ; Ophiostoma stenoceras, M85054 ; Ophiostoma ulmi, M83261 ; Phyllactinia guttata, AF021796 ; Pleospora rudis, U00975 ; Pneumocystis carinii, X12708 ; Podospora anserina, X54864 ; Porpidia crustulata, L37540 ; Protomyces inouyei, D11377 ; Pseudallescheria boydii, M89782 ; Sclerotinia sclerotiorum, X69850 ; Siphula ceratites, U72712 ; Sordaria fimicola, X69851 ; Spathularia flavida, Z30239 ; Sporormia lignicola, U42478 ; Sporothrix schenckii, M85053 ; Urnula hiemalis, Z49754 ; and Zygosaccharomyces rouxii, X58057.

PKS KS domain cloning Conserved nucleotide sequences flanking the ketosynthase (KS) domain of fungal PKS genes were identified by sequence alignment of 12 cloned fungal PKS genes (Tables 1–2). Nested degenerate primers were designed and used to amplify KS domains from genomic DNA

of cranberry endophytic fungi. PCR conditions were : 94 m, 5 min ; 5i(94 m, 30 s, 55 m, 45 s, 72 m, 1 min) ; 35i(94 m, 30 s, 60 m, 45 s, 72 m, 1 min) ; and 72 m, 5 min. Final concentration for each of the two primers in the PCR reactions was 4 µ. PCR products were ligated into pGEM-T easy vectors (Promega, Madison, WI) and then transformed into E. coli (strain JM109) using the protocol recommended by the manufacturer. E. coli colonies that contained putative KS domains were selected by colony blotting using a DNA mixture of 10 KS domains as a probe (Table 1). DNA probes were labeled with $#P[dCTP] Ready-to-Go DNA labeling beads (Amersham, Piscataway, NJ). Colonies were transferred to Immobilon-Nyj transfer membranes (Millipore Corporation, Bedford, MA). Colony blotting was performed as previously described (Sambrook, Fritsch & Maniatis 1989). Colonies that hybridized to the mixed KS domains probe were selected for nucleotide sequence determination. Both DNA and deduced protein sequences were compared against those in the databases to verify that they were putative PKS KS domains.

Phylogenetic analyses Small subunit ribosomal nucleotide sequences (SSU rDNA) were aligned using ClustalX (Thompson 1997) with manual manipulation using GeneDoc (Nicholas & Deerfield 1997). Introns were excised and excluded from tree construction. Relative to the Neurospora crassa SSU rDNA sequence (X04971), the aligned sequence begins at base 107. Regions between bases 1349–1355 and 1485–1488 were deleted from the analyses due to ambiguity, and the alignment ends at base 1668. All phylogenetic analyses were conducted with PAUP Version 4n0b6 (Swofford 2000). Maximum

M. Sauer and others parsimony trees were obtained by heuristic searching with random addition using the tree bisection reconnection algorithm. All characters had equal weight and gaps were treated as missing data. Sequence sets were bootstrapped 500 times and branches that appeared in 50 % or more of the resampled trees were used to assess their support to the overall tree construction. The tree was rooted with the Archiascomycetes, Pnuemocystis carinii and Protomyces inouyei. Similarly, deduced amino acid sequences of PKS KS domains were aligned using ClustalX with manual manipulation using GeneDoc. Phylogenetic analyses were also conducted with PAUP Version 4n0b6 (Swofford 2000) and parameters used in the analyses were the same as described above. The aligned sequences are deposited in TreeBase in NEXUS format, with the accession numbers SN645 and SN647.

RESULTS

465 PKS KS domain primers design To clone polyketide synthase KS domains from nonsporulating endophytic isolates, we designed a series of degenerate primers based on both amino acid and nucleotide sequence alignments of 12 fungal PKS genes (Table 1). Two primers, FPKSKSU-2 and FPKSKSD1, were selected as the optimal pair to amplify PKS KS domains from diverse fungal groups. The sequence of FPKSKSU-2 (5h-ATSTCKCCYMRRGARGC) represents nucleotide positions of 3029–3045 in the Aspergillus nidulans wA gene (GenBank Accession number X65866). The sequence of FPKSKSD-1 (5h-CHMSRT GRCCRAYRTTKG) represents nucleotide positions of 3833–3851 in the A. nidulans wA gene. The degenerate nucleotide codes are as follows : H l TjA ; K l GjT ; M l AjC ; R l GjA ; S l GjC ; and Y l CjT. The overall degeneracy for FPKSKSU-2 and FPKSKSD-1 are 128 and 512, respectively. The distance between FPKSKSU-2 and FPKSKSD-1 in the wA gene of A. nidulans is 822 bp.

Fungal isolation and characterization Many of the cranberry sporulating endophytic isolates were morphologically identified as genera of ascomycetes, zygomycetes or mitosporic fungi and these fungi were further analyzed in downstream natural products screening programs, but a significant portion of fungal endophytes were non-sporulating and could not be identified (data not shown). From the pool of nonsporulating isolates, 23 were selected for phylogenetic analysis because they exhibited distinct colony morphology and pigmentation and relatively slow radial growth rates (e.g., colony diameter 5 cm in 1 month on YM agar). Of the 23 strains, ten were isolated from cranberry roots (CR38, CR68, CR70, CR101, CR121, CR133, CR137, CR203, CR228, and CR310) twelve were isolated from stems (CR47, CR97, CR132, CR216, CR265, CR271, CR325, CR475, CR478, CR483, CR493, and CR513), and strain CR61 was isolated from leaves. Maximum parsimony analyses based on SSU rDNA sequences placed 13 of the 23 non-sporulating isolates within a clade that included Oidiodendron tenuissimum of the Myxotrichaceae (Fig. 1). Many of the isolates in this group have low SSU sequence divergence. For example, SSU sequences of CR475, CR478 and CR483 were identical, and CR216 and CR265 differed by one base pair. Strains CR121, CR310, CR132, CR101, CR68, and CR203 could not be placed in one of the major ascomycetes clades, although CR310 and CR121 formed a bootstrap supported (67 %) clade and CR132 and CR101 formed a clade not supported by bootstrap (Fig. 1). CR38 grouped within the pyrenomycetes, among the Diaporthales and Sordariales. CR513 grouped with black yeasts of the Chaetothyriales. CR61 grouped adjacent to Dothideales but without bootstrap support (Fig. 1). Morphological and SSU rDNA sequence analyses identified one of the 23 strains, CR70, as an oomycete (data not shown).

KS domains of polyketide synthases from the nonsporulating cranberry endophytic fungi Using PCR primers FPKSKSU-2 and FPKSKSD-1, we surveyed 23 non-sporulating cranberry endophytic fungi for the presence of KS domains of PKS genes. For each fungal strain, we first ligated the PCR products into the pGEM-T easy vector, and about 300 Escherichia coli colonies were then selected for colony Southern hybridization. Nucleotide sequences of clones that hybridized with the ten mixed KS-domain probe were determined. The ten KS domains were amplified from the producing fungi using gene specific primers (Table 2). The ten PCR products were mixed in equal quantities and the mixture was used as probe in the colony Southern hybridization. Twelve DNA fragments that encode KS domains of putative fungal PKS genes were identified from 11 of the 23 fungi tested (Table 3). Two KS domains were cloned from strain CR216 (CR216-1 and CR216-2). The size of the DNA fragments ranged from 825–959 bp (Table 3). Of the 12 cloned PKS KS domains, those from CR97, CR216-1, CR216-2, CR265, and CR493 contained one or two introns in the KS domains (Table 3). All seven introns have the consensus 5h-GT splice site ; six of the introns contain the consensus 3h-AG splice site while the intron in clone CR216-1 has a 3h-CA site (Ballance & Turner 1985, Ballance 1986, 1991, Bruchez, Eberle & Russo 1993, Radford & Parish 1997). Six of the seven introns contain the putative CTRAY (where R is a purine and Y is a pyrimidine) Lariat sequence, while clone CR2162 has a putative CTCAG Lariat sequence (Ballance & Turner 1985, Ballance 1986, Ballance 1991, Bruchez et al. 1993, Radford & Parish 1997). The two introns in clone CR265 splice at the same positions as those in clone CR493. The 3h-introns in both clones are 52 bp in length, while the 5h-introns in the two clones are

Polyketide synthase ketosynthase domains of endophytic fungi

466

Table 3. Comparisons of 12 KS domains cloned from 11 strains of cranberry endophytic fungi.

Isolate

Gene name

Genomic DNA size (bp)

cr38 cr47 cr61 cr97 cr216 cr216 cr265

pks1\38 pks1\47 pks1\61 pks1\97 pks1\216 pks2\216 pks1\265

829 826 825 889 878 876 959

cr310 cr475 cr483 cr493

pks1\310 pks1\475 pks1\483 pks1\493

826 832 832 934

cr513

pks1\513

826

Intron size (bp)

63 52 50 78 52

52 52

Intron 5h and 3h splicing sites and lariat sequence

GT(40 bp)CTGAC(14 bp)AG GT(36 bp)CTAAT(7 bp)CA GT(18 bp)CTCAG(23 bp)AG GT(61 bp)CTAAC(8 bp)AG GT(42 bp)CTGAC(1 bp)AG

GT(37 bp)CTAAC(8 bp)AG GT(34 bp)CTAAC(9 bp)AG

different in size (78 bp for CR265 and 52 bp for CR493). The intron in CR97 is spliced at the same site as that of the CR216-1 intron, but the sizes of the two introns are different (63 bp for CR97 and 52 bp for CR216-1) (Table 3). Phylogenetic relationship of the 12 KS domains from cranberry endophytic fungi Nucleotide sequence alignment of the predicted coding regions of the 12 KS domains cloned from cranberry endophytic fungi (Table 3) and 22 corresponding domains of fungal PKSs in the databases (Table 1) was performed using ClustalX (Thompson et al. 1997). The overall nucleotide identity among the 34 sequences ranged from 38 % to 83 %. Translation of these 34 deduced coding sequences gave protein products of 272–276 amino acids (Table 3). The amino acid sequence alignment indicated an overall similarity of 49–99 % and identity of 28–99 % among the 34 sequences. Phylogenetic analysis of protein sequences grouped the 34 proteins into three major clusters (Fig. 2). The clusters correlate highly with the types of products formed by the PKS genes. The melanin PKS cluster contains sequences from seven sequences, four of which were isolated from cranberry fungal endophytes (CR310, CR216-1, CR97, and CR38). PSWA is a fungal PKS homolog isolated from Phoma sp., however, the biological function of the PSWA is unknown (Table 1). Two of the sequences in this group are melanin PKSs from Nodulisporium sp. (NSPKS1) and Colletotrichum lagenarium (CLPKS1) (Table 1). A second cluster contains three groups : the aflatoxin PKS group, the pigment PKS group, and GFPKS4 (Fig. 2). The aflatoxin PKS group contains the Aspergillis nidulans sterigmatocystin biosynthetic gene ANST, the aflatoxin biosynthetic genes APPKSA and APPKSL1 isolated from A. parasiticus, and two PKS homologs, CR475 and CR483, from cranberry endophytic fungi (Fig. 2, Table 3). The pigment PKS group

Deduced coding sequence used in translation (bp)

Translation product (aa)

816 825 825 819 822 825 825

272 275 275 273 274 275 275

825 819 828 819

275 273 276 273

825

275

contains AFALB1, ANWA, PGWA and APWA2 (Fig. 2). AFALB1 and ANWA are involved in the biosynthesis of spore pigments in A. fumigatus and A. nidulans, respectively (Table 1). The biological functions of APWA2 and PGWA, isolated from A. parasiticus and Penicillium griseofulvum, respectively, are unknown (Table 1). GFPKS4 is involved in the biosynthesis of the red pigment bikaverin in G. fujikuroi (Kjaer et al. 1971). Bikaverin also has antitumor and cytotoxicity activity (Fuska, Proksa & Fuskova 1975). The third cluster contains two functional groups. The 6-methyl salicylic acid PKS group contains PPMSAS, ATPKSM, PSMSAS, PGPKS2, APPKSL2, and ATATX (Fig. 2, Table 1). Two of the six genes in this group, PPMSAS and ATATX are involved in the antibiotic 6-methyl salicylic acid biosynthesis in P. patulum and A. terreus, respectively (Table 1). The four PKSs in the second group of this cluster include ScPKS, CHPKS1, GFFUM5, and NPKS. All of these genes are involved in the biosynthesis of potent bioactive compounds (Table 1). ScPKS and NPKS are involved in the biosynthetic pathway of the HMG-CoA reductase inhibitor lovastatin in A. terreus (Table 1). GFFUM5 is responsible for the mycotoxin fumonisin biosynthesis in Gibberella fujikuroi, and CHPKS1 is involved in the phytotoxin T-toxin biosynthesis in Cochliobolus heterostrophus (Table 1). The remaining seven sequences (CR493, CR265, CR61, CR216-2, CR513, CR47, and APWA1) could not be grouped with any of the three functional clusters (Fig. 2). Even though the intron sequences were removed from the alignment and phylogenetic analyses, PKS genes that contain intron(s) in the analyzed region tended to be clustered (Fig. 2, Table 3). One of the intron clusters is located in the clade consisting of ScPKS, CHPKS1, GFFUM5, and NPKS in which all contain introns (Fig. 2). Two intron-containing KS domains (CR216-1 and CR97) are in the melanin PKS cluster. One or more introns in the analyzed KS region were found in CR493, CR265, APWA1, and CR216-2 (Table 3). No introns were found in the ten PKS

M. Sauer and others

467

Fig. 2. Phylogenetic relationship of the amino acid sequences of KS (ketosynthase) domains of 34 fungal polyketide synthases (PKSs). The 34 fungal PKS genes used in the analyses are listed in Table 1 and Table 3. The phylogenetic tree presented was generated from parsimony analysis (Swofford, 2000) and is one of nine most parsimonious trees. The outgroup aveA3 gene encodes one of the four PKS genes involved in the avermectin biosynthesis in Streptomyces avermitilis (Ikeda et al. 1999). The frequencies with which the clades were represented
50 % in 500 bootstrap replications (Felsenstein 1988) are indicated on their respective branches. The total number of characters was 238, all with equal weight. Among the 238 characters, 194 were parsimony-informative, 24 characters were constant, and the remaining 20 variable characters were parsimony-uninformative. The total length of the tree was 1379. The consistency index (CI), excluding uninformative characters, was 0n6135. The retention index (RI) was 0n7332. Sequences with one or more introns in the analyzed KS DNA fragments were identified by ‘‘ a ’’ and PKSs that contain a methyl transferase domain were identified by ‘‘ b ’’. PKS genes encoding known polyketide structures were underlined.

genes in the second cluster (Fig. 2). Three fungal PKSs containing methyl transferase domains (ScPKS, GFFUM5, and NPKS) are clustered together, even though the methyl transferase domains were not used in the sequence alignment and phylogenetic analyses. DISCUSSION Phylogenetic inferences based on SSU sequences Use of SSU rDNA sequences of the 23 non-sporulating isolates allowed their phylogenetic placement by comparison to SSU rDNA sequences available in GenBank.

The non-sporulating isolates generally did not represent taxa which usually produce recognizable conidiomata or ascomata in culture, such as the Hypocreales, Xylariales, or Eurotiales. The majority of these isolates (13 of 23) fell within the clade centered around O. tenuissimum in the Myxotrichaceae. Although the Myxotrichaceae has been traditionally classified in the Onygenales, recent phylogenetic studies of the SSU rDNA recognized that the Myxotrichaceae is placed in a different lineage, among the Leotiales and the Erysiphales (Sugiyama, Ohara & Mikawa 1999). A phylogenetically similar

Polyketide synthase ketosynthase domains of endophytic fungi assemblage of sterile isolates were observed among root isolates of W. pungens (Epacridaceae). Internal transcribed spacer (ITS) sequence data associated W. pungens root isolates with the Myxotrichaceae, Leotiaceae and other families of the Leotiales (Chambers, Liu & Cairney 2000). This study also confirmed the placement of the Myxotrichaceae among the families of Leotiales. Some of the isolates from cranberry that grouped with O. tenuissimum therefore could represent ericoid mycorrhizal fungi. However, the paucity of authenticated SSU sequences available in GenBank prevented further resolution of this clade. Two degenerate primers can amplify the ketosynthase domains of polyketide synthase genes from diverse fungal groups Primers FPKSKSU-2 and FPKSKSD-1 were designed by aligning the KS domains of 12 fungal PKS genes present in the databases at the beginning of this study (Table 1). The utility of the FPKSKSU-2 and FPKSKSD-1 primer set was validated by its ability to amplify PKS genes from diverse fungal groups which included 58 isolates of various Aspergillus species (Polishook et al. 2000), 12 cranberry endophytic fungi, 12 lignicolous dematiaceous hyphomycetes, and 10 fungi with known PKS genes (Table 1). Of the 92 fungi tested, hybridization signals were observed for 62 fungi at high washing stringency (data not shown). It is interesting to note that more than one hybridization band was observed for most of the fungi tested. Also, using degenerate primers designed from the KS domains, Bingle et al. (1999) observed that the fungi tested in their study had two types of PKSs : the wA-type and the MSAS-type. These results confirm that most fungi have more than one type of polyketide synthase gene. For example, multiple PKS genes have been identified from A. nidulans, A. parasiticus, A. terreus, Phoma sp., Gibberella fujikuroi, and Penicillium griseofulvum (Table 1). KS domains of polyketide synthases from cranberry endophytic fungi DNA fragments encoding putative PKS KS domains were cloned from about 50 % of the 23 cranberry endophytic fungi tested (Table 3). Of the 11 fungal isolates, only CR216 had two PKSs (Table 3). This low cloning frequency may result from the use of primers and probes that were not optimal for this group of fungi, because the Southern hybridization experiment showed that 12 of the 23 cranberry fungi had the weakest signals relative to that of the other three fungal groups (data not shown). The lengths of the 12 genomic DNA fragments amplified from 11 cranberry endophytic fungi varied from 825 to 959 bp. However, the removal of introns yielded exon sequences that were almost identical in length, ranging from 684 to 691 bp (Table 3). Even

468

though the overall nucleotide identity among the 34 sequences ranged from 38 % to 83 %, there are about 65 nucleotides conserved among all sequences. Similarly, the overall amino acid identity among the 34 sequences was between 28–99 %, and 17 amino acids were conserved in all the sequences. Among the 17 conserved amino acids is the ketosynthase active cysteine site that is involved in thioester formation. In their analyses of KS domains of 17 fungal PKS genes, Bingle et al. (1999) observed that a histidine residue about 135 amino acids C-terminal of the active cysteine site is conserved in all sequences. Not surprisingly, this histidine residue is also conserved in all 34 sequences analyzed in this study. This histidine residue has been hypothesized to play a role in increasing the nucleophilicity of the active site (Aparicio et al. 1996). The roles of the other 15 amino acid residues that are conserved in all 34 sequences are not known, but presumably they are important in enzyme folding, maintaining active secondary and tertiary structure, and\or in substrate specificity. Phylogenetic relationship of the 12 KS domains from cranberry endophytic fungi Phylogenetic analyses of nucleotide sequences grouped the 34 sequences into three major clusters and six functional groups (Fig. 2). The melanin PKS cluster contained seven sequences. Only two of the seven genes, NSPKS1 and CLPKS1 (Takano et al. 1995, Fulton et al. 1999), have been proven to encode polyketides that are polymerized to form melanins. However, it is reasonable to predict that the other five PKS genes, including three from cranberry endophytic fungi, are involved in the production of melanin-type PKSs. It is interesting to note that the spore pigment genes clustered with the aflatoxin PKS group, rather than the melanin genes. Of the five sequences in the spore color group, three genes are known to be involved in the spore color polyketide biosynthesis (Table 1). Two PKSs from cranberry endophytic fungi, CR475 and CR483, were clustered with three aflatoxin PKS genes (ANST, APPKSA and APPKSL1) (Fig. 2, Table 1). Although, it is possible that the CR475 and CR483 PKSs are involved in aflatoxin-type polyketide biosynthesis in these fungi, this needs to be verified by both genetic and chemistry studies. It is interesting to note that the red spore pigment, bikaverin, is also an antibiotic (Kjaer et al. 1971, Fuska et al. 1975). None of the 12 KS domains from the cranberry endophytic fungi grouped with the antibiotic and the 6methyl salicylic acid PKS group (Fig. 2). It is interesting that the 6-methyl salicylic acid PKS group of six sequences is closely related to the group which contains four genes (ScPKS, CHPKS1, GFFUM5, and NPKS) that are involved in the biosynthesis of polyketides that are very different on the basis of both structure and bioactivity. Pinpointing the relationships of the four

M. Sauer and others genes based on their product structure and bioactivity is difficult, but two interesting facts emerged from the analyses. The first is that three of the four genes (ScPKS, CHPKS1, and NPKS) contain introns (Fig. 2), even though their intron sequences were not included in the phylogenetic analyses. The second is that three of the four (ScPKS, GFFUM5, and NPKS) genes contain a methyl transferase domain and these three genes are the only fungal PKSs that have been identified to contain methyl transferase domains. Methyl transferase domains were not used in the sequence alignment and phylogenetic analyses. Several regions in both the amino acid and nucleotide alignments appear to be unique in these methyltransferase containing genes. Based on these findings, designing primers and probes for targeted cloning of methyltransferase containing PKSs from fungi may be possible. Similar to the ScPKS, CHPKS1, NPKS, and GFFUM5 cluster, two other intron containing genes were clustered in the melanin group (CR216-1 and CR97). It seems that intron containing PKSs tend to be clustered, but the significance of the observation is not clear. More studies are needed to determine whether these results reflect any biological and evolutionary relationships of fungal polyketide synthase encoding genes. Our initial survey of putative polyketide synthase genes from cranberry endophytes suggests that these fungi are a rich resource for polyketides, even though some of them may be difficult to ferment by conventional means. The results of this experiment did not suggest that the phylogenetic relationships of the cranberry endophytic fungi based on SSU rDNA sequences is predictive of those based on the protein sequences of the KS domains of PKS genes. Nor can we conclude that the SSU rDNA sequences are predictive of the number of PKS gene clusters or the products encoded by the PKS genes. This is largely because our search of PKS genes from the 23 cranberry endophytic fungi were far from exhaustive. Polyketide synthase genes have been detected in most of the major lineages of the ascomycetous fungi, except the Hemiascomycetes and Archiascomycetes. The more relevant question is which groups among the ascomycetes do not possess PKS gene clusters. Therefore, a more direct way to select fungi for production of targeted polyketide metabolites would be to search for PKS genes from a large pool of fungi across a taxonomic spectrum. This is especially true when one uses primers and probes which are derived from the PKS gene groups that encode biologically active polyketides or of those with unknown functions. This approach could be generalized to other major secondary metabolic pathways, such as nonribosomal peptides and isoprenoids. A C K N O W L E D G E M E N TS We thank Jan Tkacz, Richard Monaghan, and William Strohl for critically reading the manuscript, and Randy S. Currah for discussion on the Myxotrichaceae and ericoid mycorrhizal fungi.

469 REFERENCES Altschul, S. F., Gish, W., Miller, W., Myers, E. W. & Lipman, D. J. (1990) Basic local alignment search tool. Journal of Molecular Biology 215 : 403–10. Aparicio, J. F., Molnar, I., Schwecke, T., Konig, A., Haydock, S. F., Khaw, L. E., Staunton, J. & Leadlay, P. F. (1996) Organization of the biosynthetic gene cluster for rapamycin in Streptomyces hygroscopicus : analysis of the enzymatic domains in the modular polyketide synthase. Gene 169 : 9–16. Ballance, D. J. (1986) Sequences important for gene expression in filamentous fungi. Yeast 2 : 229–236. Ballance, D. J. (1991) Transformation systems for filamentous fungi and an overview of fungal gene structure. In Molecular Industrial Mycology : systems and applications for filamentous fungi (S. A. Leong & R. M. Berka, eds) : 1–30. Marcel Dekker, New York. Ballance, D. J. & Turner, G. (1985) Development of a high-frequency transforming vector for Aspergillus nidulans. Gene 36 : 321–331. Beck, J., Ripka, S., Siegner, A., Schiltz, E. & Schweizer, E. (1990) The multifunctional 6-methylsalicyclic acid synthase gene of Penicillium patulum. Its gene structure relative to that of other polyketide synthases. European Journal Biochemistry 192 : 487– 498. Bedford, D. J., Schweizer, E., Hopwood, D. A. & Khosla, C. (1995) Expression of a functional fungal polyketide synthase in the bacterium Streptomyces coelicolor A3(2). Journal of Bacteriology 177 : 4544 – 4548. Bingle, L. E., Simpson, T. J. & Lazarus, C. M. (1999) Ketosynthase domain probes identify two subclasses of fungal polyketide synthase genes. Fungal Genetics and Biology 26 : 209–223. Boddy, L. & Griffith, G. S. (1989) Role of endophytes and latent invasion in the development of decay communities in sapwood of angiospermous trees. Sydowia 41 : 41–73. Bruchez, J. J. P., Eberle, J. & Russo, V. E. A. (1993) Regulatory sequences involved in the translation of Neurospora crassa mRNA : Kozak sequences and stop codons. Fungal Genetics Newsletter 40 : 85–88. Bruns, T. D., Fogel, R. & Taylor, J. W. (1990) Amplification and sequencing of DNA from fungal herbarium specimens. Mycologia 82 : 175–184. Byrd, A. D., Schardl, C. L., Songlin, P. J., Mogen, K. L. & Siegel, M. R. (1990) The β-tubulin gene of Epichloe typhina from perennial ryegrass (Lolium perenne). Curruent Genetics 18 : 347–354. Chambers, S. M., Liu, G. & Cairney, J. W. G. (2000) ITS rDNA sequence comparison of ericoid mycorrhizal endophytes from Woollsia pungens. Mycological Research 104 : 168–174. Chang, P. K., Cary, J. W., Yu, J., Bhatnager, D. & Cleveland, T. E. (1995) The Aspergillus parasiticus polyketide synthase gene pksA, a homolog of Aspergillus nidulans wA, is required for aflatoxin B1 biosynthesis. Molecular and General Genetics 248 : 270–277. Dalpe! , Y., Litten, W. & Sigler, L. (1989) Scytalidium vaccinii sp. nov., an ericoid endophyte of Vaccinium angustifolium roots. Mycotaxon 35 : 317–377. Dombrowski, A. W., Bills, G. F., Sabnis, G., Koupal, L. R., Meyer, R., Ondeyka, J. G., Giacobbe, R. A., Monaghan, R. L. & Lingham, R. B. (1992) L-696,474, a novel cytochalasin as an inhibitor of HIV-1 protease I. The producing organism and its fermentation. Journal of Antibiotics 45 : 671–678. Dreyfuss, M. M. & Chapela, I. H. (1994) Potential of fungi in the discovery of novel, low-molecular weight pharmaceuticals. In The Discovery of Natural Products with Therapeutic Potential (V. P. Gullo, ed.) : 49–80. Butterworth-Heinmann, Boston, MAO. Ehrlich, K. C., Montalbano, B. G., Bhatnagar, D. & Cleveland, T. E. (1998) Alteration of different domains in AFLR affects aflatoxin pathway metabolism in Aspergillus parasiticus transformants. Fungal Genetics and Biology 23 : 279–287. Farr, D. F., Bills, G. F., Chamuris, G. P. & Rossman, A. Y. (1989) Fungi on plant and plant products in the United States. American Phytopathological Society Press, St Paul MN. Felsenstein, J. (1988) Phylogenies from molecular sequences : inference and reliability. Annual Review of Genetics 22 : 521–565.

Polyketide synthase ketosynthase domains of endophytic fungi Feng, G. H. & Leonard, T. J. (1995) Characterization of the polyketide synthase gene (pksL1) required for aflatoxin biosynthesis in Aspergillus parasiticus. Journal of Bacteriology 177 : 6246 –6254. Feng, G. H. & Leonard, T. J. (1998) Culture conditions control expression of the genes for aflatoxin and sterigmatocystin biosynthesis in Aspergillus parasiticus and A. nidulans. Applied and Environmental Microbiology 64 : 2275–2277. Fujii, I., Ono, Y., Tada, H., Gomi, K., Ebizuka, Y. & Sankawa, U. (1996) Cloning of the polyketide syntase gene atX from Aspergillus terreus and its identification as the 6-methylsalicylic acid synthase gene by heterologous expression. Molecular and General Genetics 253 : 1–10. Fulton, T. R., Ibrahim, N., Losada, M. C., Grzegorski, D. & Tkacz, J. S. (1999) A melanin polyketide synthase (PKS) gene from Nodulisporium sp. that shows homology to the pks1 gene of Colletotrichum lagenarium. Molecular and General Genetics 262 : 714–720. Fuska, J., Proksa, B. & Fuskova, A. (1975) New potential cytotoxic and antitumor substances I. In vitro effect of bikaverin and its derivatives on cells of certain tumors. Neoplasma 22 : 335–8. Ikeda, H., Nonomiya, T., Usami, M., Ohta, T. & Omura, S. (1999) Organization of the biosynthetic gene cluster for the polyketide anthelmintic macrolide avermectin in Streptomyces avermitilis. Proceedings of the National Academy of Sciences, USA 96 : 9509–95014. Kennedy, J., Auclair, K., Kendrew, S. G., Park, C., Vederas, J. C. & Hutchinson, C. R. (1999) Modulation of polyketide synthase activity by accessory proteins during lovastatin biosynthesis. Science 284 : 1368–1372. Kjaer, D., Kjaer, A., Pedersen, C., Bu’lock, J. D. & Smith, J. R. (1971) Bikaverin and norbikaverin, benzoxanthentrione pigments of Gibberella fujikuroi. Journal of the Chemical Society, Perkin Transactions, ser. I 16 : 2792–2797. Mayorga, M. E. & Timberlake, W. E. (1992) The developmentally regulated Aspergillus nidulans wA gene encodes a polypeptide homologous to polyketide and fatty acid synthases. Molecular and General genetics 235 : 205–212. Nicholas, K. B. jr., N. H. B. & Deerfield, D. W. (1997) GeneDoc : analysis and visualization of genetic variation. EMBnet News 4 : 14. Pazoutova, S., Linka, M., Storkova, S. & Schwab, H. (1997) Polyketide synthase gene pksM from Aspergillus terreus expressed during growth phase. Folia Microbiology 42 : 419– 430. Petrini, O. (1984) Endophytic fungi in British Ericaceae : a preliminary study. Transactions of the British Mycological Society 83 : 510–512. Petrini, O. (1987) Endophytic fungi of alpine Ericaceae. The endophytes of Loiseleuria procumbens. In Arctic and Alpine Mycology. Vol. 2 (G. A. Laursen, J. F. Ammirati & S. A. Redhead, eds) : 71–77. Plenum Press, New York. Petrini, O., Sieber, T. N., Toti, L. & Viret, O. (1992) Ecology, metabolite production, and substrate utilization in endophytic fungi. Natural Toxins 1 : 185–196. Petrini, O., Stone, J. & Carroll, F. E. (1982) Endophytic fungi in evergreen shrubs in western Oregon : a preliminary study. Canadian Journal of Botany 60 : 789–796. Polishook, J. D., Pela! ez, F., Platas, G., Asensio, F. J. & Bills, G. F. (2000) Observations on Aspergilli in Santa Rosa National Park, Costa Rica. Fungal Diversity 4 : 81–100. Proctor, R. H., Desjardins, A. E., Plattner, R. D. & Hohn, T. M. (1999) A polyketide synthase gene required for biosynthesis of

470

fumonism mycotoxins in Gibberella fujikuroi Mating Population A. Fungal Genetics and Biology 27 : 100–112. Radford, A. & Parish, J. H. (1997) The genome and genes of Neurospora crassa. Fungal Genetics and Biology 21 : 258–66. Sambrook, J., Fritsch, E. F. & Maniatis, T. (1989) Molecular Cloning : a laboratory manual. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY. Shear, C. L., Stevens, N. E. & Bain, H. F. (1931) Fungous diseases of the cultivated cranberry. Technical Bulletin of the USDA 258 : 1–56. Spatafora, J. W., Mitchell, T. G. & Vilgalys, R. (1995) Analysis of genes coding for small-subunit rRNA sequences in studying phylogenetics of dematiaceous fungal pathogens. Journal of Clinical Microbiology 33 : 1322–6. Sugiyama, M., Ohara, A. & Mikawa, T. (1999) Molecular phylogeney of onygenalean fungi based on small subunit ribosomal DNA (SSU rDNA) sequences. Mycoscience 40 : 251–258. Swofford, D. L. (2000) PAUP : phylogenetic analysis using parsimony. Version 4.0b4a. Sinauer Associates, Sunderland, MA. Takano, Y., Kubo, Y., Kawamura, C., Tsuge, T. & Furusawa, I. (1997) The Alternaria alternata melanin biosynthesis gene restores appressorial melanization and penetration of cellulose membranes in the melanin-deficient albino mutant of Colletotrichum lagenarium. Fungal Genetics and Biology 21 : 131–140. Takano, Y., Kubo, Y., Shimizu, K., Mise, K., Okuno, T. & Furusawa, I. (1995) Structural analysis of PKSI, a polyketide synthase gene involved in melanin biosynthesis in Colletotrichum lagenarium. Molecular and General Genetics 249 : 162–167. Thompson, J. D., Gibson, T. J., Plewniak, F., Jeanmougin, F. & Higgins, D. G. (1997) The ClustalX windows interface : flexible strategies for multiple sequence alignment aided by quality analysis tools. Nucleic Acids Research 25 : 4876–4882. Tsai, H. F., Chang, Y. C., Washburn, R. G., Wheeler, M. H. & Kwon-Chung, K. J. (1998) The developmentally regulated alb1 gene of Aspergillus fumigatus : its role in modulation of conidial morphology and virulence. Journal of Bacteriology 180 : 3031–3038. Vinci, V. A., Conder, M. J., McAda, P. C., Reeves, C. D., Rambosek, J., Davis, C. R. & Hendrickson, L. E. (1998) DNA encoding triol polyketide synthase. USA patent No. 5,744,350. Wang, I. K., Reeves, C. & Gaucher, G. M. (1991) Isolation and sequencing of a genomic DNA clone containing the 3h terminus of the 6-methysalicyclic acid polyketide synthetase gene of Penicillum urticae. Canadian Journal of Microbiology 37 : 86–95. Watanabe, A., Fujii, I., Sankawa, U., Mayorga, M. E., Timberlake, W. E. & Ebizuka, Y. (1999) Re-identification of product identification of Aspergillus nidulans wA gene to code for a polyketide synthase of naphthopyrone. Tetrahedron Letters 40 : 91–94. White, T. J., Bruns, T., Lee, S. & Tayler, J. (1990) Amplification and direct sequencing of fungal ribosomal RNA genes for phylogenetics. In PCR protocols : a guide to methods and applications (M. A. Innis, D. H. Gelfand, J. J. Sninsky & T. J. White, eds) : 315–322. Academic Press, San Diego. Widler, B. & Mu$ ller, E. (1984) Untersuchgen u$ ber endophytische Pilze von Arctostaphylos uva-ursi (L) Sprengel (Ericaceae). Botanica Helvetica 94 : 307–337. Yang, G., Rose, M. S., Turgeon, B. G. & Yoder, O. C. (1996) A polyketide synthase is required for fungal virulence and production of the polyketide T-toxin. Plant Cell 8 : 2139–2150. Yu, J. H. & Leonard, T. J. (1995) Sterigmatocystin biosynthesis in Aspergillus nidulans requires a novel type I polyketide synthase. Journal of Bacteriology 177 : 4792–4800. Corresponding Editor : R. S. Currah