PHYLOGENY OF THE PERONOSPOROMYCETES (OOMYCOTA) BASED ON PARTIAL

Download The phylogeny of the nuclear large ribosomal subunit (LSU) rDNA from Peronosporomycetes (Oomycota) was studied. Five orders. Rhipidiales, ...

0 downloads 345 Views 276KB Size
Mycol. Res. 104 (11) : 1295–1303 (November 2000). Printed in the United Kingdom.

1295

Phylogeny of the Peronosporomycetes (Oomycota) based on partial sequences of the large ribosomal subunit (LSU rDNA)

Ann B. PETERSEN and Søren ROSENDAHL Department of Mycology, Botanical Institute, University of Copenhagen, Øster Farimagsgade 2D, DK-1353 Copenhagen K, Denmark. E-mail : soerenr!bot.ku.dk Received 16 June 1999 ; accepted 22 March 2000.

The phylogeny of the nuclear large ribosomal subunit (LSU) rDNA from Peronosporomycetes (Oomycota) was studied. Five orders Rhipidiales, Leptomitales, Saprolegniales, Pythiales and Peronosporales were included in the study in order to reveal phylogenetic relationships within the group. Based on maximum parsimony, neighbour-joining and maximum-likelihood the Peronosporomycetes seem to have evolved as two major lineages. One lineage includes members of the Rhipidiales, Leptomitales and Saprolegniales and the other includes the members of Pythiales and Peronosporales. However, the inclusion of Rhipidiales in Saprolegniomycetidae was not well supported. Leptomitales were placed basal to the Saprolegniales in the LSU rDNA based tree. In the Saprolegniaceae sensu lato the primary zoospore seems to have been lost several times in the evolution of this family. Species of Aphanomyces were placed on the most basal branch in this family. The obligate parasite, Albugo, was placed on the most basal branch within the Peronosporomycetidae. Phytophthora showed a closer relationship to Peronospora and Peronophythora than to Pythium suggesting that this genus should be removed from Pythiales to Peronosporales.

INTRODUCTION The Peronosporomycetes (Oomycota) are known to be related to the chromophyte algae and other heterokont protists. This relationship between Peronosporomycetes and heterokont protists is based on the ultrastructure of the zoospore, e.g. a straminipilous ornamentation of the anterior flagella and similarities in the flagellar rootlet structure (Barr 1981, 1992, Beakes 1987). DNA studies support this relationship and suggest that Peronosporomycetes is monophyletic (Gunderson et al. 1987, Fo$ rster et al. 1990). The traditional taxonomy of Peronosporomycetes distinguishes four orders : Lagenidiales, Leptomitales, Peronosporales, and Saprolegniales (Sparrow 1960, Dick 1973). The taxonomy of the Peronosporomycetes has been reorganized several times (Dick, Wong & Clark 1984, Dick et al. 1989, Dick 1990, 1995). Dick et al. (1984) recognised three additional orders : Rhipidiales, previously Rhipidiaceae, placed in Leptomitales, Pythiales including organisms formerly placed in Lagenidiales and finally Sclerosporales including species of Sclerosporaceae formerly placed in Peronosporales. In this revision Peronosporomycetes was subdivided in two subclasses : Saprolegniomycetidae with Saprolegniales as a single order, and Peronosporomycetidae with the orders Leptomitales, Rhipidiales, Sclerosporales, Pythiales and Peronosporales. Leptomitales and Sclerosporales were later transferred to Saprolegniomycetidae based on DNA studies (Dick et al. 1989, Klassen et al. 1988). Dick et al. (1989) and Dick (1995) proposed a third subclass,

the Rhipidiomycetidae with members of the Rhipidiales, and reorganized Saprolegniomycetidae and Peronosporomycetidae. The phylogenetic relationships between the groups in Peronosporomycetes have long been a matter of controversy (Bessey 1942, Barr 1983, Beakes 1987, Dick et al. 1984, Dick 1988, 1990). Bessey (1942), Barr (1983) and Beakes (1987) proposed a monophyletic origin of the Peronosporomycetes. Barr (1983) suggested that the obligate parasites on terrestrial plants evolved from Saprolegniales through primitive Pythium and Phytophthora. According to this theory Peronosporomycetes developed from saprobes in aquatic habitats. Terrestrial facultative soil-borne parasites on plant roots developed from these and later obligate parasites of the aerial parts of the plants evolved. The zoospore stages and the dependency on water were gradually reduced during this evolution. Barr (1983) also suggested that Leptomitales evolved as a separate lineage from Saprolegniales. This idea was based on the presence of chitin in some of the species in Leptomitales. Beakes (1987) supported the theory of a monophyletic origin of Peronosporomycetes, and proposed two different evolutionary models for the class. In one of these Saprolegniales was considered the main ancestral group due to the similarity between Saprolegnia and the alga Vaucheria. Leptomitales, Peronosporales and Lagenidiales evolved sequentially or independently from Saprolegniales. He further suggested that characteristics of wall synthesis during zoosporogenesis should be regarded as a key character. A polyphyletic origin of the Peronosporomycetes has also

Phylogeny of the Peronosporomycetes been proposed. Sparrow (1976) suggested two independent evolutionary lineages, a saprolegnian and a peronosporacean lineage. Dick et al. (1984) and Dick (1988, 1990) adopted this idea, and suggested different evolutionary hypotheses of Peronosporomycetes. Later, however, Dick (1995) stated that it is not yet possible to designate primitive or advanced phylogenetic criteria for Peronosporomycetes. In order to propose new phylogenetic theories, it is necessary to include additional characters. Additional characters for phylogenetic studies can be obtained from nucleic acid sequences. Such characters have been used in taxonomy and phylogeny of the economically important plant pathogens Pythium and Phytophthora (Lee & Taylor 1992, Briard et al. 1995, Crawford et al. 1996), and for species of the Saprolegniaceae (Daugherty et al. 1998). Ribosomal DNA (rDNA) sequences have been used extensively to reconstruct phylogenetic relationships among fungi and algae, where the evolutionary interpretation of morphological features is ambiguous (Taylor 1993, Bhattacharya & Medlin 1995, Hibbit et al. 1997), and lately Dick et al. (1999) used 18S rDNA to suggest a phylogeny for the Peronosporomycetes. So far the LSU (28S) rDNA has not been used to develop phylogenetic hypotheses for the major lineages of the Peronosporomycetes. The objective of the present study was to construct a phylogeny of members of the Peronosporomycetes based on partial sequences of LSU rDNA. The molecular phylogeny was compared with morphological and physiological characters in this group. We sequenced 1200 base pairs (bp) of the large subunit ribosomal DNA (LSU rDNA) gene from 24 members of the Peronosporomycetes representing the three major subclasses listed above and several families within these. MATERIALS AND METHODS Fungal material The taxa\isolates used in the study and the sources of these are given in Table 1. Peronospora farinosa and P. parasitica were obtained as sporangia from infested spinach (Spinacea oleracea) and infested white cauliflower (Brassica oleracea var. botrytis) respectively obtained from Dæhnfeldt A\S (Denmark). Albugo candida was obtained as sporangia from infested Shepherds purse (Capsella bursa-pastoris). Saprolegnia ferax (Sf5.6) and Achlya americana (am2) were isolated from a lake (Bagsværd sø, Denmark), Saprolegnia litoralis (Sl1) was isolated from forest soil (Grib skov, Denmark). Dictyuchus sp. no. 4.2 and Dictyuchus sp. no. 5.6 were isolated from twigs from fresh water (Smørmosen, Denmark) and wet soil (Hareskoven, Denmark) respectively. DNA extraction and PCR conditions The cultures were grown as 100 ml batches in liquid media (Fuller & Jaworski 1987). The batches were inoculated from agar cultures and incubated at 20 mC for 1 week. The resultant mycelia were washed in dilute salt solution (DS) (Dill & Fuller 1971), blotted on filter paper and freeze dried. Total genomic DNA was then extracted by the methods of Lee, Milgroom &

1296 Taylor (1988). The extracted DNA was checked on 1 % agarose gels, and used for PCR-amplifications of LSU rDNA, except for P. farinosa, P. parasitica and A. candida where resting spores were crushed with a pestle and used directly for PCRamplifications. The primers LSU-0025-F and LSU-1170-R (Table 2) were used to amplify the first approximately 1200 bp of the LSU rDNA. PCR conditions were : Initial denaturation for 1 min at 96 mC, followed by 30 cycles of denaturation 1 min at 95 mC, annealing 1 min at 52 m, extension 2 min at 72 m and a final extension of 7 min at 72 m.

DNA sequencing Amplifications were checked on 2 % Nusieve gels, and PCRproducts were purified from the amplification mixture with milipore Ultrafree-MC filters (Milipore Corporation, Bedford, MA) and used in cycle-sequencing reactions (ABI Dye Terminator Cycle Sequencing Ready Reactions Kit) together with specific primers (Table 2). Both strands of the LSU rDNA were sequenced on an automatic sequencer (ABI Prism TM 377 DNA Sequencer), and chromatograms of these sequences were checked using Sequencer 3.1 (Gene Codes Corporations Inc., Ann Arbor).

Sequence alignment and phylogenetic analysis Twenty-four new sequences from this study were aligned with two published LSU rDNA sequences obtained from GenBank, using the computer program ESEE (vers. 3.1 1997, Cabot, E). The database on the structure of large ribosomal subunit RNA was used for the alignment (De Rijk, van der Peer & De Wachter 1997). The two published sequences were a sequence of Phytophthora megasperma accession no. X75631 and Hyphochytrium catenoides accession no. X80345. For alignment of variable regions of the LSU rDNA the program Sequence Navigator (Applied Biosystems, version 1.0) was used. The sequences varied substantially and could not be unambiguously aligned. The alignment is available from the authors (soerenr!bot.ku.dk). Analyses were performed that included all organisms but with an exclusion of 180 characters (position 436–507, 580–625, 889–950) with an ambiguous alignment. Phylogenetic analyses were performed with the software PAUP* 4.0b2. All maximum parsimony searches were performed using Fitch parsimony (unordered, multistate characters) with gaps treated as missing data using the heuristic search option, TBR (tree bisection reconnection) branch swapping and random addition sequence with 300 replicates. To assess the robustness of the clades, bootstrap percentages were calculated. Bootstrap values (BV) were calculated for 200 replicates with simple entry of the data. Consistency index (CI) and retention index (RI) were calculated for all parsimony trees (Kluge & Farris, 1969 ; Farris, 1989). In addition to the parsimony analysis alternative topologies were searched by maximum-likelihood and neighbour-joining methods using PAUP* 4.0b2. Maximum-likelihood trees were constructed using the heuristic search algorithm. The transition\transversion ratio

A. B. Petersen and S. Rosendahl

1297

Table 1. List of species of the Peronosporomycetes sequenced in this study. Taxonomic placement of the species included (Dick 1995) and their source.

Major group Peronosporomycetes Peronosporomycetidae Peronosporales Peronosporaceae Albuginaceae Pythiales Pythiaceae

Rhipidiomycetidae Rhipidiales Rhipidiaceae Saprolegniomycetidae Saprolegniales Saprolegniaceae

Leptomitales Leptomitaceae

a b c d e f g

Species

Isolate and source

GenBank\EMBL accession no.

Peronospora farinosa P. parasitica Albugo candida

Spinaceae oleracea Brassica oleracea var. botrytis Capsella bursa-pastoris

AF235955 AF235957 AF235938

Pythium aphanidermatum Lagenidium chthamalophilum L. giganteum Phytophthora infestans Peronophythora litchii

170, Hockenhull, J.a 65-028b Novo Nordisk A\Sc Pi-3, Novo Nordisk A\Sc CBS 100.81

AF235956 AF235946 – – AF235949

Sapromyces elongatus

CBS 213.82

AF235950

Saprolegnia ferax S. litoralis Achlya americana Isoachlya toruloides Aphanomyces euteiches A. astaci A. piscicida Dictyuchus sp. Dictyuchus sp. Leptolegnia sp. Leptolegnia sp. Thraustotheca clavata Brevilegnia bispora

Sf5.6, Petersen, A. Bd SI1, Petersen, A. B.d am2, Petersen, A. B.d Tayseir, M. A.e ATCC 201684 (uv), So$ derha$ ll, K.f njm 9211, Hatai, K.g no. 5.6, Petersen, A. B.d no. 4.2, Petersen, A. B.d no. 3.1, Sørensen, D.d no. 1.6, Sørensen, D.d CBS 557.67 CBS 568.67

AF235953 AF235952 AF235943 AF235947 AF235939 AF235940 AF235941 AF235945 AF235944 AF235954 AF235948 AF235951 AF235942

Apodachlya minima A. brachynema

CR-55b 61-020b

AF235937 AF235936

The Royal Veterinary and Agricultural University, Copenhagen. University of California, Berkeley Microgarden. Novo Nordisk A\S, Bagsværd. Department of Mycology, Botanical Institute, University of Copenhagen. Botanical Department, Faculty of Science, Suez Canal University. Department of Comparative Physiology, University of Uppsala. Division of Fish Diseases, Nippon Veterinary and Animal Science University, Tokyo.

Table 2. Primers used for PCR and sequencing of LSU rDNA from organisms of Peronosporomycetes. Name

Sequence

Reference

LSU-0025-F LSU-0344-F LSU-0670-F LSU-0826-F LSU-1170-R LSU-0826-R LSU-0670-R LSU-0344-R

ACCCGCTGAACTTAAGCATAT CGATAGCGAACAAGTACCGTG GACTGAGGTGCCTACAAC CTTGAAACACGGACCAAGGAG GCTATCCTGAGGGAAATTTCGG CTCCTTGGTCCGTGTTTCAAG GTTGTAGGCACCTCAGTC CACGGTACTTGTTCGCTATCG

van der Auwera et al. (1994) This study This study This study van der Auvera et al. (1994) This study van der Auwera et al. (1994) This study

was estimated by maximum-likelihood. Standard settings for maximum-likelihood were used, and the Hasegawa-KishinoYano (HKY) two parameter model for unequal base frequencies was chosen. Branch support was determined by bootstrap analysis (Felsenstein 1985) calculated using 100 replicates. The settings for the bootstrap analysis were as above, except that the transition\transversion ratio estimated above was used. Hyphochytrium catenoides was chosen as outgroup to the

Peronosporomycetes based on its sistergroup relationship to this group (van der Auwera et al. 1995). RESULTS The 24 sequences obtained were aligned with the published sequences of the large ribosomal subunit of P. megasperma and H. catenoides, and the length of the sequences obtained,

Phylogeny of the Peronosporomycetes

1298

Fig. 1. Phylogenetic relationships within Peronosporomycetes (Oomycota) inferred from a maximum parsimony analysis of 981 nucleotides of nuclear LSU rDNA. A strict consensus tree of two equally most parsimonious trees requiring 984 steps with consistency index l 0n559 and retension index l 0n712 is shown. Bootstrap percentages higher than 50 % from 200 replicates are shown above each branch. The differences in the two equally parsimonious trees obtained were whether Peronophythora litchii clustered with Phytophthora infestans or P. megasperma. Arrows (6) indicate where the primary zoospore is suggested to be lost in Saprolegniales.

surrounded by the primers LSU-0025-F and LSU-1170-R, varied from 1030 to 1121 bp. The taxa belonging to the Saprolegniales had an approx. 60 bp large deletion in helix 2 corresponding to bp 432–492 in P. megasperma (van der Auwera, Chapelle & De Wachter 1994) or the excluded positions 436–507 in the alignment. A maximum parsimony tree based on 981 bp of the partial sequences of the LSU rDNA of the Peronosporomycetes is shown in Fig. 1. Of the 981 included sites 361 were variable, and 245 sites of these were parsimony-informative. The tree was a strict consensus tree of two equally parsimonious trees requiring 984 steps. The CI and the RI for the two trees were 0n559 and 0n712 respectively. Bootstrap percentages were

assigned above the branches. The tree divided the 25 taxa\isolates of Peronosporomycetes included in this study into two major lineages. One lineage consisted of species of Leptomitales, Saprolegniales and Sapromyces, and the other lineage contained the species of the Pythiales and Peronosporales. These lineages were supported by bootstrap values of 74 % and 95 % respectively (Fig. 1). In the former lineage, Sapromyces representing Rhipidiales was the most basal organism. The two species of Apodachlya representing Leptomitales were on the next branch to diverge (100 % BV). Saprolegniales seemed to be the most advanced group in this lineage with high bootstrap support (92 %). Three major clusters were seen in Saprolegniales. Species of

A. B. Petersen and S. Rosendahl

1299

Fig. 2. Phylogenetic relationships within Peronosporomycetes (Oomycota) inferred from a maximum likelihood analysis of 981 nucleotides of nuclear LSU rDNA using the model of Hasegawa–Kishino–Yano. The tree is an optimum maximum likelihood tree obtained by a heuristic search. The –Ln likelihood is 6366.15977 and the estimated transition\transversion ratio is 1.651202. (kappa l 3.249597). Bootstrap percentages higher than 50 % from 100 replicates are shown above each branch.

Aphanomyces formed a cluster on the most basal branch with high bootstrap support (97 %), resulting in a cluster containing the remaining species of Saprolegniales with very low bootstrap support (55 %). This cluster consisted of two major clusters. One of these contained species of Saprolegnia and Leptolegnia (93 % BV), and the other major cluster were composed of species of Achlya, Thraustotheca and Isoachlya in one internal cluster (95 % BV) and Brevilegnia and Dictyuchus in another internal cluster (100 % BV). Albugo diverged on the most basal branch within the species representing the Peronosporomycetidae. The species of Peronospora, Phytophthora and Peronophythora formed a strongly

supported clade (100 % BV), with the species of Peronospora on the most basal line. The positions of Pythium and Lagenidium in the LSU rDNA based tree were less clear. Pythium was resolved as basal to Lagenidium (75 % BV) while Lagenidium was resolved as a sister to the clade including Phytophthora but with bootstrap support lower than 50 %. Minor differences were seen between the two most parsimonious trees. In the two trees Peronophythora litchii clustered with either Phytophthora infestans or P. megasperma. The overall topology of the trees obtained by maximum likelihood (Fig. 2) was similar to the tree based on maximum parsimony seen in Fig. 1 except for the position of Lagenidium.

Phylogeny of the Peronosporomycetes The optimum maximum likelihood tree had a -Ln likelihood of 6366.15977 and the estimated transition\transversion ratio of this tree was 1.651202 (kappa l 3.249597) (Fig. 2). The inclusion of Sapromyces as representing the Rhipidiales in the Saprolegniomycetidae was supported by only 67 % bootstrap in the maximum likelihood tree. Aphanomyces was also suggested as basal in the Saprolegniales in the maximum likelihood tree. The resulting clade including the remaining species of Saprolegniales obtained a bootstrap support of 79 %. Isoachlya clustered with Achlya and Thraustotheca, but this position was not well-supported (56 %). In Peronosporomycetidae, Albugo was placed on the most basal branch with high bootstrap support (96 %) as in the maximum parsimony based trees. The species of Lagenidium were placed as basal to a clade including Pythium, Phytophthora and Peronospora but supported by low bootstrap values (57 %). In this clade Pythium was evolving on a separate line with low bootstrap supports (53 %), showing sistergroup relationships to a cluster including Peronospora, Phytophthora and Peronophythora. This cluster was wellsupported (100 %). The topology of the neighbour-joining tree was identical with the trees inferred from maximum likelihood method (data not shown). DISCUSSION The phylogenetic trees based on partial sequences of the LSU rDNA presented here confirm the two major lineages in the Peronosporomycetes representing the Saprolegniomycetidae and the Peronosporomycetidae as suggested by Dick et al. (1984). One lineage includes Sapromyces representing Rhipidiales, Apodachlya representing Leptomitales, and the species representing Saprolegniales (Saprolegniomycetidae). The other lineage includes species representing Pythiales and Peronosporales (Peronosporomycetidae). In a restriction analysis of rDNA (Klassen et al. 1988), and in the taxonomic revisions by Dick et al. (1989) and Dick (1995), Peronosporomycetes consisted of three subclasses : Saprolegniomycetidae, Peronosporomycetidae and Rhipidiomycetidae. However, the inclusion of Sapromyces, the single representative for the Rhipidiales in this study, in Saprolegniomycetidae is only moderately supported in the maximum parsimony analysis (74 % BV) and in the maximumlikelihood analysis (67 % BV). Several characters support the existence of two major lineages in the LSU rDNA based phylogenetic trees. The members of the Leptomitales and the Saprolegniales are able to synthesise sterols de novo while the members of the Pythiales are not (Domnas, Srebo & Hicks 1977, Nes 1987, Weete 1989). K -bodies (kinetosome-associated vesicles) present in # the zoospore of some members of the Peronosporomycetes might also be a key character. K -bodies have so far been # found in Sapromyces, Apodachlya and members of the Saprolegniales, but not in the Peronosporomycetidae (Holloway & Heath 1977, Hoch & Mitchell 1972, Olson et al. 1984, Powell, Lehnen & Bortnick 1985, Beakes 1987, Gotelli & Hanson 1987, Randolph & Powell 1992, Powell & Blackwell 1995). The oosporogenesis and oospore organisation is also an important key character. According to Dick (1969, 1995) Saprolegniales and probably Leptomitales have a centrifugal

1300 oosporogenesis resulting in non-periplasmic oospores, while the members of the Peronosporomycetidae have a centripetal oosporogenesis resulting in oospores with a persistent periplasm. Sapromyces has morphological features that place it between the Saprolegniomycetidae and the Peronosporomycetidae. Gotelli & Hanson (1987) noted that the ultrastructure of Sapromyces zoospores suggests that Rhipidiales are more closely related to Saprolegniales than to Pythiales. Beakes (1987) noted that the secondary cysts of Sapromyces are coated with an outer layer of fibrillar material and lack the electron-dense coat similar to the secondary cysts of the Pythiales. The oospore morphology is similar to that found in Peronosporomycetidae, e.g. the presence of periplasm (Dick 1969, 1996). On the basis of the present study a placement for members of Rhipidiales can not be suggested. Additional and independent molecular markers will be needed to place this group of organisms with certainty. The phylogenetic trees resulting from sequences of the LSU rDNA strongly suggest that Leptomitales should be included in the Saprolegniomycetidae as suggested by Dick et al. (1989). Leptomitales and Saprolegniales are placed as sistergroups in all LSU rDNA based trees with very high bootstrap support (100 %). Morphological characters also support this placement. In both orders the presence of a pyriform zoospore with a pair of flagella anchored near the spore apex (the primary zoospore) has been described (Sparrow 1960, Dick 1973, Jacobs 1982). Saprolegniales and Leptomitales share traits in the oospore morphology as both produce oospores without a periplasm (Dick 1969, 1995), and similarities are seen in the zoospore morphology (Randolph & Powell 1992). A proximal relationship between the Leptomitales and the Saprolegniales was also suggested by Klassen et al. (1988) based on a restriction analysis of ribosomal DNA, and lately by a maximum parsimony analysis based on sequences of the SSU rDNA (Dick et al. 1999). The species of Saprolegniales included in the present study form a monophyletic clade. The genus Aphanomyces is suggested to be the first lineage to diverge in the maximum parsimony trees, the neighbour-joining tree and in the maximum likelihood trees. In many ways Aphanomyces is different from other genera of Saprolegniales. The K -bodies # examined in the genus Aphanomyces are of different morphotypes than the K -bodies seen in other species of the # Saprolegniales (Powell & Blackwell 1995). The genus Aphanomyces also has a different oospore organisation. In Aphanomyces the ooplasm is non-fluid and shows no Brownian movement of the granules as it is seen in the other genera of Saprolegniales (Dick 1971, Howard 1971, Traquair & McKeen 1980, Dick 1995). Antibodies produced against oospores of A. euteiches showed almost no cross-reactions with other genera of Saprolegniales and Pythiales, suggesting that Aphanomyces has a different wall chemistry compared to other groups of Peronosporomycetes (Petersen, Olson & Rosendahl 1996). Aphanomyces was suggested to be included in a newly erected family of Saprolegniales, the Leptolegniaceae, together with species of Leptolegnia and Plectospira, by Dick et al. (1999) based on morphological and molecular data. In the phylogenetic trees based on LSU rDNA presented here an unnamed

A. B. Petersen and S. Rosendahl isolate of Leptolegnia clustered with Saprolegnia and not with Aphanomyces in the maximum parsimony trees, in the maximum-likelihood tree and in the neighbour-joining tree. Our data are thus not supporting the erection of the family Leptolegniaceae. However, we did not include the type species Leptolegnia caudata. It would be interesting to include LSU rDNA sequences of this species in future studies to evaluate the proposed family Leptolegniaceae. Morphologically based studies of these organisms indicate that they should be grouped together (Powell & Blackwell 1998, Dick et al. 1999). It has been postulated several times that Saprolegnia is the most basal organism in the Saprolegniaceae (Humphrey 1893, Atkinson 1909, Ho$ hnk 1933, Barr 1983), mainly because of its mechanism of asexual reproduction. In Saprolegnia two morphologically different zoospores are formed. The first formed zoospore, the primary zoospore, is pyriform and has two flagella anchored in the spore apex. The secondary zoospore is reniform with the flagella inserted laterally in a grove. This zoospore is generally a much better swimmer and it is in some species able to perform repeated emergence (Die! guez-Uribeondo, Cereneus & So$ derha$ ll 1994). The different mechanisms of spore release seen in the Saprolegniaceae have been viewed as a transition series with Saprolegnia exhibiting the ancestral mechanism (Humphrey 1893, Atkinson 1909, Ho$ hnk 1933). In this series the primary zoospore stage is reduced. Achlya releases pyriform spores from its sporangia showing some similarities with the primary zoospore of Saprolegnia, but they encyst at the apex of the zoosporangia, and have in some species been shown only to possess rudimentary flagella with stunted axonemes (Money et al. 1987). In Thraustotheca the primary spores encyst in the sporangium before they are released by rupture of the sporangial walls. Later secondary zoospores are released from the cysts. In Dictyuchus the spores encyst in the sporangia. Protoplasts are released through individual papilla that project through the sporangial wall. These are reorganized into secondary zoospores on the surface of the sporangia. The only remnant of the primary zoospore in this genus is the cyst walls in the sporangia (Coker & Matthews 1937). In Aphanomyces protoplasts are cleaved out in hypha-like sporangia. These protoplasts move to the orifice of sporangia and encyst. Later the cysts germinate with a secondary zoospore. It has been shown that the primary spores, the protoplasts, of the plant pathogenic species Aphanomyces euteiches do not possess flagella (Hoch & Mitchell 1972). Daugherty et al. (1998) confirmed this theory by using internal transcribed spacer sequences (ITS) of four genera of Saprolegniaceae. In their phylogenetic analysis Saprolegnia merged as the most basal organism, sister to Achlya, Thraustotheca and Dictyuchus, with Achlya and Thraustotheca as most closely related, while Dictyuchus appeared to have evolved independently of these. The topology of the LSU rDNA based phylogenetic trees of the Saprolegniaceae sensu lato is consistent with the topology of the phylogenetic tree based on sequences of ITS published by Daugherty et al. (1998). In both studies a suppression of the flagella of the primary spores is seen during evolution of the family. But our trees include more species, and a new pattern emerges. The LSU based phylogenetic trees suggest several

1301 independent losses of the primary zoospore during the evolution of the Saprolegniales (Fig. 1). The primary zoospore is first lost at the branch with Aphanomyces, then at the branch with Achlya and Thraustotheca, and the third time it is lost at the branch with Brevilegnia and Dictyuchus. In our trees Saprolegnia is not the most basal organism of Saprolegniales, but Aphanomyces is suggested to contain the most basal organisms in this family. Dick et al. (1999) suggests that the primary zoospore is an apomorphic state. According to Dick (1990) the primary zoospore is known only with certainty in Saprolegnia sensu lato (including Isoachlya). It is thus most likely that the primary zoospore has been lost several times during evolution. The alternative, that this zoospore with restricted motility has evolved twice, is doubtful. The topology of the part of the LSU rDNA based phylogenetic trees including members of the Peronosporomycetidae is non-congruent with most phylogenetic theories of the Peronosporomycetes. In most theories primitive Pythium and\or Phytophthora are seen as ancestors to the obligate parasites in Peronosporales (Bessey 1942, Shaw 1981, Barr 1983). This is partly based on the view that parasitism represents the apomorpic state in contrast to saprotrophy. In the phylogenetic trees presented here Albugo is seen as basal in the Peronosporomycetidae. Albugo is a biotrophic parasite on hosts of relatively early origin like Amaranthaceae and Convolvulaceae (Dick 1988). This might support a basal placement of this genus compared to species of Pythium, which contains parasites on monocots. In the true Fungi studies on molecular phylogeny suggest that obligate parasites such as Taphrinales and Uredinales are basal in Ascomycota and Basidiomycota respectively (Swan & Taylor 1993, Nishida & Sugiyama 1994). A basal placement might also be the case for some of the obligate parasites of the Peronosporomycetes. Phytophthora and Pythium have traditionally been placed together in Peronosporaceae in Peronosporales (Waterhouse 1973) or lately in Pythiales in Peronosporomycetidae based on similarities in morphology and in details of the sexual reproduction (Waterhouse 1973, Dick 1995). In the phylogenetic tree based on partial sequences of LSU rDNA the species of Phytophthora included in this study cluster with Peronospora rather than with Pythium, suggesting that Phytophthora should be removed from Pythiales to Peronosporales. This alternative placement is strongly supported by bootstrap percentages (100 %). Several differences between Pythium and Phytophthora support such a movement. Species of Phytophthora produce well formed sporangia on distinct sporangiophores, compared to the hypha-like or spherical sporangia of Pythium, and the zoosporogenesis of the two genera follows two different patterns. In Pythium the protoplasm of the sporangia is emitted into a vesicle, where the differentiation of the zoospores takes place, and the zoospores are released by rupture of the vesicle, while in Phytophthora the zoospores are fully differentiated in the sporangia within a boundary plasmamembrane as seen in the Peronosporales. Several species of Phytophthora, e.g. the type species Phytophthora infestans, are dispersed aerially and show parallels to the species of Peronosporales in the production of caducous sporangia and use

Phylogeny of the Peronosporomycetes of direct germination via a germ-tube. Ecological differences between Pythium and Phytophthora are well known. The species of Pythium are saprobes in soil and water, and they are weak to moderate parasites (Waterhouse 1973). Species of Phytophthora are poor competitive saprotrophs but aggressive pathogens mainly on dicots, like the species found in Peronosporales (Waterhouse 1973). A key character used to place Phytophthora together with Pythium in Pythiales is the nature of the exospore wall layer of the oospore (Dick 1969, Waterhouse 1973). In Peronosporales the mature oospore contains a well-defined exposure wall layer derived from a persistent periplasm. In Pythium and Phytophthora a thin layer of periplasm is present outside the membrane of the oospore, and this thin layer later disappears in some species. The LSU rDNA based phylogenetic trees do not cluster Albugo and Peronospora together, but Albugo is placed on the most basal branch in the Peronosporomycetidae. In both genera a conspicuous exospore wall layer is present. The LSU rDNA based trees thus suggest that the exospore is not the key character for uniting these organisms. Instead the exospore wall layer might represent a plesiomorph, shared with the Rhipidiales, and reduced or lost at some of the internal branches in the Peronosporomycetidae. Lagenidium is a genus that has been placed differently by several authors. Sparrow (1960) placed it in Lagenidiaceae in Lagenidiales together with the families Olpidiopsidaceae and Sirolpidiaceae based on their endobiotic, simple thalli with sexuality involving fusion of contents of two thalloid bodies. Dick et al. (1984) transferred Lagenidiaceae to Pythiales, recognising the affinity between Pythium and Lagenidium. Lagenidium sensu stricto was later transferred to Pythiaceae (Dick 1995). In both genera zoosporogenesis by external cleavage in a vesicle is seen, and they share several traits in their sexual reproduction. The trees based on maximum parsimony suggest that Pythium is basal to Lagenidium, but with bootstrap values below 50 % (Fig. 1). The maximum likelihood tree (Fig. 2) and the neighbour-joining tree suggest that Lagenidium evolved on a line basal to Pythium, but this is not well supported by bootstrap values. Both trees suggest that the included species of Lagenidium should be placed in a separate order Lagenidiales. Unfortunately the present data can not resolve the phylogenetic placement of Lagenidium and Pythium. Inclusion of sequences of a wide range of species of Myzocytium, Myzocytiopsis and additional species of Lagenidium and Pythium might help to suggest a placement of the genus Lagenidium. ACKNOWLEDGMENTS J. W. Taylor is acknowledged for his advice on sequencing and alignments of the LSU rDNA sequences. We wish to thank the following people for supplying us with cultures : J. W. Taylor for the isolates of Lagenidium chthamalophilum, Apodachlya minima and Apodachlya branchynema from U. C., Berkeley Microgarden, M. A. Tayseir for the culture of Isoachlya toruloides, K. Hatai for the culture of Aphanomyces piscicida, K. So$ derhall for the culture of Aphanomyces astaci, J. Hockenhull of the culture of Pythium aphanidermatum, D. Sørensen for the cultures of Leptolegnia sp., B. D. Jensen for plants infected with Peronospora farinosa and Peronosora parasitica, and Novo Nordisk A\S for the isolates of Lagenidium giganteum and Phytophthora infestans. We thank C. Hansen for technical help, and N. Daugbjerg for his advice on alignments and maximum-likelihood analyses. M. W. Dick is acknowledged for his comments

1302 on the manuscript, and for correcting the English text. This research was supported by a grant from The Faculty of Science, University of Copenhagen, to A. B. Petersen.

REFERENCES Atkinson, G. F. (1909) Some problems in the evolution of the lower fungi. Annales Mycologici 7 : 441–472. Barr, D. J. S. (1981) The phylogenetic and taxonomic implications of flagellar rootlet morphology among zoosporic fungi. BioSystems 14 : 359–70. Barr, D. J. S. (1983) The zoosporic groupings of plant pathogens. Entity or non-entity ? In Zoosporic Plant Pathogens. A modern perspective (S. T. Buczacki, ed.) : 43–83. Academic Press, London. Barr, D. J. S. (1992) Evolution and kingdoms of organisms from the perspective of a mycologist. Mycologia 84 : 1–11. Beakes, G. W. (1987) Oomycete phylogeny : ultrastructural perspectives. In Evolutionary Biology of the Fungi (A. D. M. Rayner, C. M. Brasier & D. Moore, eds) : 405–421. Cambridge University Press, Cambridge, UK. Bessey, E. A. (1942) Some problems in fungus phylogeny. Mycologia 34 : 355–379. Bhattacharya, D. & Medlin, L. (1995) The phylogeny of plastids based on comparisons of small-subunit ribosomal RNA coding regions. Journal of Phycology 31 : 489–498. Briard, M., Dutertre, M., Rouxel, F. & Brygoo, Y. (1995) Ribosomal RNA sequence divergence within the Pythiaceae. Mycological Research 99 : 1119–1127. Coker, W. C. & Matthews, V. D. (1937) Saprolegniales. North American Flora 2(1) : 15–76. Crawford, A. R., Bassam, B. J., Dreth, A., Maclean, D. J. & Irwin, J. A. G. (1996) Evolutionary relationships among Phytophthora species deduced from rDNA sequence analysis. Mycological Research 100 : 437–443. Daugherty, J., Evans, T. M., Skillom, T., Watson, L. E. & Money, N. P. (1998) Evolution of spore release mechanisms in the Saprolegniaceae (Oomycetes) : evidence from a phylogenetic analysis of internal transcribed spacer sequences. Fungal Genetics and Biology 24 : 354–363. de Rijk, P., van de Peer, Y. & de Wachter, R. (1997) Database on the structure of large ribosomal subunit RNA. Nucleic Acids Research 25 : 117–122. Dick, M. W. (1969) Morphology and taxonomy of the Oomycetes, with special reference to Saprolegniaceae, Leptomitaceae and Pythiaceae. I. Sexual reproduction. New Phytologist 68 : 751–775. Dick, M. W. (1971) Oospore structure in Aphanomyces. Mycologia 63 : 686–688. Dick, M. W. (1973) Saprolegniales. In The Fungi. An advanced treatise (G. C. Ainsworth, F. K. Sparrow & A. S. Sussman, eds) 4A : 109–114. Academic Press, New York. Dick, M. W. (1988) Coevolution in the heterokont fungi (with emphasis on the downy mildews and their angiosperm host). In Coevolution of Plants and Animals (K. A. Pirozynski & D. L. Hawksworth, eds) : 31–62. Academic Press, London. Dick, M. W. (1990) Phylum Oomycota. In Handbook of Protoctista (L. Margulis, J. O. Corliss, M. Melkonian & D. Chapman, eds) : 661–685. Jones & Bartlet, Boston. Dick, M. W. (1995) Sexual reproduction in the Peronosporomycetes (chromistan fungi). Canadian Journal of Botany 73 (suppl. 1) : S712–S724. Dick, M. W., Croft, B. J., Magarey, R. C., de Cock, A. W. A. M. & Clark, G. (1989) A new genus of the Verrucalvaceae (Oomycetes). Botanical Journal of the Linnean Society 99 : 97–113. Dick, M. W., Vick, M. C., Gibbings, J. G., Hedderson, T. A. & Lopez Lastra, C. C. (1999) 18S rDNA for species of Leptolegnia and other Peronosporomycetes : justification for the subclass taxa Saprolegniomycetidae and Peronosporomycetidae and division of the Saprolegniaceae sensu lato into the Leptolegniaceae and Saprolegniaceae. Mycological Research 103 : 1119–1125. Dick, M. W., Wong, P. T. W. & Clark, G. (1984) The identity of the oomycete causing ‘ Kikuyu Yellows ’, with a reclassification of the downy mildews. Botanical Journal Linnean Society 89 : 171–197. Die! guez-Uribeondo, J., Cereneus, L. & So$ derha$ ll, K. (1994) Repeated zoospore emergence in Saprolegnia parasitica. Mycological Research 98 : 810–815. Dill, B. C. & Fuller, M. S. (1971) Amino acid immobilization of fungal motile cells. Archiv fuW r Mikrobiologie 78 : 92–98.

A. B. Petersen and S. Rosendahl Domnas, A. J., Srebo, J. P. & Hicks, B. F. (1977) Sterol requirement for zoospore formation in the mosquito parasitizing fungus Lagenidium giganteum. Mycologia 69 : 875–886. Farris, J. S. (1989) The retention of index and the rescaled consistency index. Cladistics 5 : 417–419. Felsenstein, J. (1985) Confidence limits on phylogenies : an approach using the bootstrap. Evolution 39 : 666–670. Fo$ rster, H., Coffey, M. D., Elwood, H. & Sogin, M. L. (1990) Sequence analysis of the small subunit ribosomal RNAs of three zoosporic fungi and implications for fungal evolution. Mycologia 82 : 306–312. Fuller, M. S. & Jaworski, A. (1987) Zoosporic Fungi in Teaching and Research. Southeastern Publishing, Athens, GA. Gotelli, D. & Hanson, L. C. (1987) An ultrastructural investigation of the zoospore of Sapromyces androgynus (Oomycetes, Leptomitales). Mycologia 79 : 745–752. Gunderson, J. L., Elwood, H., Ingold, A., Kindle, K. & Sogin, M. L. (1987) Phylogenetic relationships between chlorophytes, chrysophytes, and oomycetes. Proceedings of the National Academy of Sciences, USA 84 : 5823–5827. Hibbit, D. S., Pine, E. M., Langer, E., Langer, G. & Donoghue, M. (1997) Evolution of gilled mushrooms and puffballs inferred from ribosomal DNA sequences. Proceedings of the National Academy of Sciences, USA 94 : 12002–12006. Hoch, H. C. & Mitchell, J. E. (1972) The ultrastructure of zoospores of Aphanomyces euteiches and their encystment and subsequent germination. Protoplasma 75 : 113–138. Holloway, S. A. & Heath, I. B. (1977) An ultrastructural analysis of the changes in organelle arrangement and structure between the various sporetypes of Saprolegnia. Canadian Journal of Botany 55 : 1328–1339. Ho$ hnk, W. (1933) Polyplanetism and zoo$ spore germination in Saprolegniaceae and Pythium. American Journal of Botany 20 : 45–62. Howard, F. L. (1971) Oospore types in Saprolegniaceae. Mycologia 63 : 679–686. Humphrey, J. E. (1893) The Saprolegniaceae of the United States, with special notes on other species. Transactions of the American Philosophical Society, n.s. 17 : 63–148. Jacobs, R. P. W. M. (1982) The genus Apodachlya. Antonie van Leeuwenhoek 48 : 389–399. Klassen, G. R., McNabb, S. A., Whitmore, E., Gray, M. & Dick, M. W. (1988) Restriction sites in ribosomal DNA as characters for the evolutionary study of zoosporic fungi. Abstracts, 16th International Congress of Genetics. Kluge, A. G. & Farris, J. S. (1969) quantitative phyletics and the evolution of anurans. Systematic Zoology 18 : 1–32. Lee, S. B., Milgroom, M. G. & Taylor, J. T. (1988) A rapid, high yield miniprep method for isolation of total genomic DNA from fungi. Fungal Genetics and Biology 35 : 23–24. Lee, S. B. & Taylor, J. W. (1992) Phylogeny of five fungus-like protistian Phytophthora species, inferred from the internal transcribed spacers of ribosomal DNA. Molecular Biology and Evolution 9 : 636–653. Money, N. P., Beakes, G. W., Webster, J. & Wakely, G. (1987) Rudimentary flagella in sporangiospores of Achlya. Transactions of the British Mycological Society 89 : 108–114.

1303 Nishida, H. & Sugiyama, J. (1994) Archiascomycetes : detection of a major new lineage within the Ascomycota. Mycoscience 35 : 361–366. Nes, W. D. (1987). Biosynthesis and requirement for sterols in the growth and reproduction of Oomycetes. In Ecology and Metabolism of Plant Lipids (G. Fuller & W. D. Nes, eds) : 305–328. [ACS Symposium Series No. 325.] American Chemical Society, Washington, DC. Olson, L. W., Cerenius, L., Lange, L. & So$ derha$ ll, K. (1984) The primary and secondary spore cysts of Aphanomyces (Oomycetes, Saprolegniales). Nordic Journal of Botany 4 : 681–696. Petersen, A. B., Olson, L. W. & Rosendahl, S. (1996) Use of polyclonal antibodies to detect oospores of Aphanomyces. Mycological Research 100 : 495–499. Powell, M. J., Lehnen jr., L. P. & Bortnick, R. N. (1985) Microbody-like organelles as taxonomic markers among Oomycetes. BioSystems 18 : 321–334. Powell, M. J. & Blackwell, W. H. (1995) Searching for homologous ultrastructural characters in zoosporic fungi. Canadian Journal of Botany 73 (suppl. 1) : S693–S700. Powell, M. J. & Blackwell, W. H. (1998) Phenetic analysis of the Saprolegniaceae (Oomycetes). Mycotaxon 68 : 505–516. Randolph, L. R. & Powell, M. (1992) Ultrastructure of zoospores of the oomycete Apodachlya pyrifera. Mycologia 84 : 768–780. Shaw, G. C. (1981) Taxonomy and evolution. In The Downy Mildews (Spencer, D. M., ed.) : 17–29. Academic Press, London. Sparrow, F. K. (1960) Aquatic Phycomycetes. 2nd edn. University of Michigan Press, Ann Arbor. Sparrow, F. K. (1976) The present status of classification in biflagellated fungi. In Recent Advances in Aquatic Mycology (E. B. G. Jones, ed.) : 193–213. Elek Science, London. Swan, E. C. & Taylor, J. W. (1993) Higher taxa of basidiomycetes : an 18S gene perspective. Mycologia 85 : 923–936. Taylor, J. W. (1993) A contemporary view of the holomorph. Nucleic acid sequences and computer databases and changing fungal classification. In The Fungal Holomorph : mitotic, meiotic and pleomorphic speciation in fungal systematics (D. R. Reynolds & J. W. Taylor, eds) : 3–15. CAB International, Wallingford. Traquair, J. A. & McKeen, W. E. (1980) Electron microscopy of developing Aphanomyces oogonia and oospores. Mycologia 72 : 378–393. van der Auwera, G., Chapelle & De Wachter, R. (1994) Structure of the large ribosomal subunit of Phytophthora megasperma, and phylogeny of the Oomycetes. FEMS Letters 338 : 133–136. van der Auwera, G., De Baere, R., van de Peer, Y., De Rijk, P., van den Broeck, I. & De Wachter, R. (1995) The phylogeny of the Hyphochytriomycota as deduced from ribosomal RNA sequences of Hyphochytrium catenoides. Molecular Biology and Evolution 12 : 671–678. Waterhouse, G. M. (1973) Peronosporales. In The Fungi. An advanced treatise (G. C. Ainsworth, F. K. Sparrow & A. S. Sussman, eds) 4B : 165–183. Academic Press, New York. Weete, J. D. (1989) Structure and function of sterols in fungi. Advances in Lipid Research 23 : 115–162. Corresponding Editor : M. W. Dick