J Gen Pathol (2005) 71:174–182 © The Phytopathological Society of Japan DOI 10.1007/s10327-005-0184-5 and Springer-Verlag Tokyo 2005

FUNGAL DISEASES

Koji Kageyama · Ayako Nakashima · Yuki Kajihara Haruhisa Suga · Eric B. Nelson Phylogenetic and morphological analyses of graminicola and related species

Received: May 19, 2004 / Accepted: September 19, 2004

Abstract Isolates of and related spe- matum with large oogonia and aplerotic oospores was cies were differentiated using restriction fragment length not related to the morphologically similar species P. polymorphism (RFLP) analyses of the internal transcribed myriotylum. Results suggest that P. graminicola and related spacer (ITS) regions of rDNA and the cytochrome c oxi- species are phylogenetically distinct, and molecular analy- dase subunit II (COX II) gene. These sequences were used ses, in addition to morphological analyses, are necessary in subsequent phylogenetic analyses. Finally, the phylo- for the accurate taxonomic placement of species in this genetic placement of species was compared to that deter- complex. mined from morphological characteristics. The 62 isolates tested were divided into seven groups, A–G, based on Key words Pythium graminicola · rDNA-ITS regions · RFLP analysis of the rDNA-ITS region. In the RFLP analy- Cytochrome oxidase gene · Phylogenetic relationship · sis of the COX II gene, isolates were divided into groups Morphology · Identification similar to those based on ITS-RFLP. Groups A and B were each separated into two additional subgroups. Grouping of isolates based on RFLP analyses agreed with the morphological differentiation. Groups A, B, D, E, F, and Introduction G were identified as P. graminicola, P. arrhenomanes, P. aphanidermatum, P. myriotylum, P. torulosum, and The Pythium is complex, containing more than 100 P. vanterpoolii, respectively. Group C was closely related species and consisting of plant and pathogens, to group B based on phylogenetic analysis of the rDNA- mycoparasites, and saprophytes. The identification of ITS region and the COX II gene and is similar to P. Pythium species has been based traditionally on mor- arrhenomanes. Each of the other species occupied their own phological characteristics (Dick 1990; Matthews 1931; individual clades. Although P. arrhenomanes is morpho- Middleton 1943; Van der Plaats-Niterink 1981; Waterhouse logically similar to P. graminicola, our phylogenetic analy- 1967). The major morphological criteria for species ses revealed that it was evolutionarily distant from P. identification are based on qualitative characteristics that graminicola and more closely related to P. vanterpoolii. Our may vary depending on the culture conditions and the iso- analysis also revealed that P. torulosum with smaller oogo- late tested (Dick 1990; Matthews 1931; Middleton 1943; nia is more closely related to P. myriotylum with large oogo- Van der Plaats-Niterink 1981; Waterhouse 1967). The event nia than to P. vanterpoolii, which forms smaller oogonia and makes accurate species placement extremely confusing and is morphologically similar to P. torulosum. P. aphanider- difficult, although species identification is one of the most important for diagnosing diseases and developing control strategies. K. Kageyama (*) · A. Nakashima · Y. Kajihara Pythium graminicola Subramaniam, one of the more River Basin Research Center, Gifu University, 1-1 Yanagido, Gifu important plant pathogenic species, is distributed world- 501-1193, Japan wide. The pathogen causes seed and root rots and seedling Tel. ϩ81-58-293-2063; Fax ϩ81-58-293-2062 e-mail: [email protected] damping-off especially in gramineous crops such as , corn, , and turf grasses. P. graminicola belongs to H. Suga Life Science Research Center, Gifu University, Gifu, Japan a large group of species with smooth-walled oogonia and inflated filamentous sporangia (Van der Plaats-Niterink E.B. Nelson 1981). This group contains other important plant patho- Department of , New York State College of Agriculture and Life Science, Cornell University, Ithaca, New York, genic species, such as P. arrhenomanes Drechsler, P. USA aphanidermatum (Edson) Fitzp, P. myriotylum Drechsler, 175 and P. vanterpoolii V. Kouyeas and H. Kouyeas. These were transferred to potato dextrose broth (Difco, Detroit, species are distinguished on the basis of various quantitative MI, USA). After 7–10 days of incubation in the dark characteristics, such as the number of antheridia per oogo- at 25°C, mycelial mats were collected on filter paper placed nium and the size of the oogonia, each of which often in a Buchner funnel and rinsed several times with distilled displays considerable variation among isolates, culture water. After removing excess water, mycelial mats were conditions, and growth media, making it difficult to assign frozen at Ϫ80°C for at least 1 day and ground to powder species to this group accurately. P. graminicola and P. with a sterilized, prefrozen mortar and pestle. The mycelial arrhenomanes have often been confused with one another powder was suspended in 250µl of extraction buffer because of the variability in the number of antheridia per (100mM Tris·HCl (pH 9.0), 40mM EDTA), 50µl of 10% oogonium. It is necessary in plant pathology and in Pythium sodium dodecyl sulfate (SDS), 10µl of 20% skim milk solu- to know whether these species are the same or tion (Difco), and 5µl of RNase A (10mg/ml) (Nippongene, they should be separated. Toyama, Japan) and then vigorously vortexed for 1min. Restriction fragment length polymorphic (RFLP) analy- Benzyl chloride (150µl) was added to the tube, which was sis of the internal transcribed spacer (ITS) regions between vigorously vortexed again for 2min and then incubated nuclear ribosomal RNA genes (rDNA) has become a useful at 50°C for 1h. Following this incubation, 150µl of 3M tool for identifying closely related Pythium species (Chen NaOAc was added to the suspension, lightly vortexed, and Hoy 1993; Matsumoto et al. 2000) that lack sexual and incubated on ice for 15min. This suspension was organs (Kageyama et al. 1998, 2003). Moreover, the analysis cleared by centrifugation at 18000g for 10min, and the of ITS sequences has provided a means of clarifying supernatant was transferred to a new tube. This step was phylogenic relationships in the genus Pythium (Matsumoto repeated twice. DNA was subsequently precipitated with et al. 1999). These data suggest that the sporangial form is two volumes of cold ethanol and collected by centrifugation the most important characteristic in Pythium taxonomy. at 18000g for 20min. The resulting pellet was rinsed with The cytochrome oxidase subunit II (COX II) gene is a 70% ethanol and dried under vacuum. The DNA pellet was housekeeping gene and is thought to accumulate mutations dissolved in 200µl of TE buffer/10mM Tris·HCl (pH 7.5), through evolution, indicating that this gene might be useful 1mM EDTA. for determining phylogenic relationships. Moreover, be- cause the gene is encoded in mitochondria, it may clarify phylogenic relationships among the species in relation to RFLP analyses of rDNA-ITS regions and COX II gene another genetic background. In species of both Pythium and , sequence analyses of the COX II gene The ITS region of nuclear rDNA was amplified with primer corroborates findings from sequence analysis of the ITS pair ITS1–ITS4 described by White et al. (1990) and the region (Martin 2000; Martin and Tooley 2003). COX II gene with FM58 and FM66 described by Martin The objectives of the present study were to distinguish (2000). The 50-µl reaction mixture contained 1µM of each isolates of P. graminicola from closely related species using primer, 1.25 units of rTaq DNA polymerase (Takara Shuzo, RFLP analyses of the ITS regions of rDNA and of the Shiga, Japan), 0.2mM dNTP mixture, 1ϫ polymerase chain COX II gene, determine phylogenic relationships based reaction (PCR) buffer (10mM Tris·HCl, pH 8.3, 50mM on these sequences, and compare sequence analyses with KCl, and 1.5mM MgCl2), and 200ng of DNA template. morphological characteristics as a means of identifying P. Reactions were carried out with a DNA Thermal Cycler graminicola and related species. (Applied Biosystems, Norwalk, CT, USA). The tempera- ture cycling parameters for the ITS regions were pro- grammed for one cycle of 3min at 94°C followed by 30 Materials and methods cycles of 1min at 94°C, 1min at 55°C, 2min at 72°C, and one cycle of 10min at 72°C. The program for the COX II Isolates was the same as the one for the ITS regions except for the annealing temperature at 52°C. PCR products were Thirty isolates of Pythium graminicola, seven of P. electrophoresed in 1.2% agarose LO3 (Takara Shuzo) gel aphanidermatum, five each of P. aristosporum and P. in TAE buffer (40mM Tris·HCl, pH 7.5, 19mM glacial arrhenomanes, ten of P. myriotylum, two of P. torulosum, acetic acid, and 2mM EDTA) and then stained with and three of P. vanterpoolii were used in this study (Table ethidium bromide. The amplified DNA was used for restric- 1). A P. dissotocum isolate was used as outgroup for phylo- tion enzyme analysis. Digestions with four restriction genetic analysis. Many isolates were collected from various enzymes – EcoRI, TaqI, HaeIII, HinfI (Toyobo, Osaka, hosts and geographic origins and were maintained on corn- Japan) – were carried out for the ITS regions according meal agar (CMA) at 25°C. to the manufacturer’s specifications and three additional enzymes – HhaI, MboI, RsaI – were used for the COX II gene. The restriction fragments were electrophoresed DNA extraction in 3.5% NuSieve (3:1) agarose gel (FMC BioProducts, Rockland, MN, USA) in TAE buffer followed by staining Three agar plugs were removed from the growing margin of with ethidium bromide and visualizing under ultraviolet 2-day-old cultures on CMA using a 1-cm cork borer and light and photography. 176

Table 1. Pythium isolates belonging to each DNA group (see Table 2) DNA Species Isolate Host/habitat Locale Sourcea Accession no. group ITS COX II

A-1 P. graminicola ATCC96234 (1425)bc Corn field soil Kumamoto, Japan 4 AB095045 AB160849 A-1 P. graminicola ATCC96600 (2948)bc Sugarcane Louisiana, USA 4 AB095044 AB095062 A-1 P. graminicola MAFF425415c Corn field soil Kumamoto, Japan 5 AB160836 AB095063 A-2 P. graminicola MAFF305860c Soil Kumamoto, Japan 5 AB160837 AB095064 B-1 P. arrhenomanes ATCC96526 (230)bc Sugarcane Louisiana, USA 4 AB095039 AB095053 B-1 P. arrhenomanes ATCC96598 (147)bc Sugarcane Louisiana, USA 4 AB160838 AB095056 B-1 P. arrhenomanes E-1 Sugarcane Kagoshima, Japan 1 B-1 P. arrhenomanes G-1 Sugarcane Kagoshima, Japan 1 B-1 P. graminicola ADO-6-1c Zoysia grass Gifu, Japan 1 AB160839 AB095055 B-1 P. graminicola KU9BGPc Bentgrass Shizuoka, Japan 6 AB160840 AB160850 B-1 P. graminicola P21-1c Bentgrass Korea 3 AB160841 AB160851 B-1 P. graminicola PRR5c Bentgrass New York, USA 1 AB160842 AB160852 B-1 P. graminicola P26-1c Bentgrass Korea 3 B-1 P. graminicola PRR13c Bentgrass New York, USA 1 B-1 P. graminicola SKGPc Bentgrass Okayama, Japan 1 B-1 P. graminicola UOP444c Bentgrass Toyama, Japan 2 B-1 P. graminicola TMF Sugarcane Kagoshima, Japan 1 AJ233444d AB095054 B-1 P. graminicola Ohmagari Rice Akita, Japan 1 B-1 P. graminicola Pma-2 Rice Iwate, Japan 1 B-1 P. graminicola HJ4 Japan 1 B-1 P. graminicola TM13 Sugarcane Kagoshima, Japan 1 B-1 P. graminicola TS2 Sugarcane Kagoshima, Japan 1 B-1 P. graminicola MAFF235840 Zoysia grass Kagawa, Japan 5 B-1 P. graminicola PRR6 Bentgrass New York, USA 1 B-1 P. graminicola PRR8 Bentgrass New York, USA 1 B-1 P. graminicola PRR9 Bentgrass New York, USA 1 B-1 P. graminicola PRR107 Bentgrass New York, USA 1 B-2 P. aristosporum ATCC11101c Saskatchewan, Canada 4 AB095042 AB095060 B-2 P. aristosporum PRR113c Bentgrass New York, USA 1 B-2 P. aristosporum PRR133c Bentgrass New York, USA 1 B-2 P. graminicola PRR115c Bentgrass New York, USA 1 AB160843 AB160853 C P. arrhenomanes ATCC96525 (1521)b Bermuda grass Hawaii, USA 4 AB095041 AB095059 C P. graminicola 32-1c Zoysia grass Hyogo, Japan 1 AB160844 AB095058 D P. aphanidermatum Toc-159c Soil Gifu, Japan 1 AJ233438d AB095072 D P. aphanidermatum GA-1c Soil Gifu, Japan 1 AB160845 AB160854 D P. aphanidermatum P36-3c Bentgrass Korea 3 AB095052 AB095073 D P. aphanidermatum TA114c Carrot field soil Korea 1 D P. aphanidermatum TJu132c Carrot field soil Korea 1 D P. aphanidermatum UOP286 Spinach Japan 2 D P. aphanidermatum UOP390 Japan 2 E P. myriotylum ATCC26082c Spinach Osaka, Japan 4 AB095047 AB095066 E P. myriotylum GF46c Kalanchoe Gifu, Japan 1 AB095051 AB095070 E P. myriotylum BeAT1c Kidney bean Hokkaido, Japan 1 AJ233450d AB095071 E P. myriotylum UOP431c Spinach Nara, Japan 2 AB095050 AB095069 E P. myriotylum KI-6c Kidney bean field soil Hokkaido, Japan 1 E P. myriotylum MkgS9807c Sweet Miyazaki, Japan 6 E P. myriotylum ATCC96232c Sugarcane Australia 4 E P. myriotylum MY-95c Anigozanthos species Israel 1 E P. myriotylum MY-149c Gyposophila paniculate Israel 1 E P. myriotylum MY-152c Gyposophila paniculate Israel 1 E P. aristosporum UOP284c Konnyaku Ibaraki, Japan 2 AB095048 AB095067 E P. aristosporum UOP394c Ryegrass Kagawa, Japan 2 AB095049 AB095068 F P. torulosum 60-2c Zoysia grass Hyogo, Japan 1 AJ233460d AB160855 F P. torulosum P31-2c Bentgrass Gifu, Japan 3 AB160846 AB160856 F P. graminicola PRR3c Bentgrass New York, USA 1 AB095046 AB095065 F P. graminicola PRR157c Bentgrass New York, USA 1 F P. graminicola PRR158c Bentgrass New York, USA 1 G P. vanterpoolii P39-1c Bentgrass Korea 3 AB160847 AB160857 G P. vanterpoolii UOP392c Zoysia grass Japan 2 AB160848 AB160858 G P. vanterpoolii K4-6-8Dc Zoysia grass Gifu, Japan 1 G P. graminicola PRR16c Bentgrass New York, USA 1 AB095043 AB095061 G P. graminicola PRR38c Bentgrass New York, USA 1 a 1, present study; 2, M. Tojo, Osaka Prefecture University; 3, J.W. Kim, Seoul National University; 4, American Type Culture Collection; 5, National Institute of Agrobiological Sciences; 6, A. Chikuo, Japanese National Institute of Floricultural Science b Isolate names used in the paper by Chen and Hoy (1993) c Morphological characteristics were observed d The sequence data were published by Matsumoto et al. (1999) 177

Fig. 1. Restriction banding patterns of poly- merase chain reaction (PCR)-amplified internal transcribed spacer of rDNA using four enzymes. a HhaI. b HinfI. c MboI. d TaqI. Lanes: M, 50-bp DNA ladder marker; 1–4, numbers of the band- ing patterns

Sequencing of rDNA-ITS regions and COX II gene oospores and the number of antheridia per oogonium were measured under the microscope. The first PCR for both the ITS regions and the COX II gene was performed as described earlier. The PCR products were purified using a Gene Elute PCR cleaning kit (Sigma, Results St. Louis, MO, USA). The Big Dye Terminater Cycle Sequencing Ready Reaction kit (Applied Biosystems) was RFLP of rDNA-ITS region and COX II gene used for the sequence reaction with primers ITS1, ITS2, ITS3, and ITS4 for the ITS regions and FM58 and FM66 for The size of the amplified rDNA-ITS region was 860–900bp. the COX II gene, according to the instructions of the manu- When the PCR products were digested with four restriction facturer. The sequencing reaction mixture was purified enzymes (HhaI, HinfI, MboI, TaqI), the digested fragments by ethanol precipitation and run on an ABI 3100 DNA of the rDNA-ITS regions showed four, three, two, and sequencer (Applied Biosystems). three patterns, respectively (Fig. 1). Based on the combina- tion of the banding patterns with these four enzymes, the 62 isolates were separated into seven groups, A–G (Table 2). Phylogenetic analysis P. graminicola isolates were divided into five groups, A, B, C, F, and G, whereas P. aphanidermatum, P. myriotylum, P. The sequence data of Drechsler was torulosum, and P. vanterpoolii isolates each belonged used as the outgroup (GenBank accession nos. AJ233443 to one group: D, E, F, and G, respectively (Table 1). P. for the rDNA-ITS region and AB095074 for the COX II aristosporum isolates were classified into group B together gene). All sequences were first aligned using the multiple with P. arrhenomanes and P. graminicola and into group E sequence alignment program CLUSTAL X version 1.81, with P. myriotylum. Group B represented the largest group, and the alignment was optimized manually. Alignment consisting of 27 isolates that had been identified as P. gaps were treated as missing data, and ambiguous positions arrhenomanes, P. graminicola, and P. aristosporum. Groups were excluded from the analysis. Nucleotide variation of A and D contained isolates from single species. Isolates these sequences was analyzed by the maximum-parsimony within all groups originated from diverse hosts and geo- method: phylogenetic analysis using parsimony (PAUP) graphic origins. version 4.0 10 (Swofford 2001). The bootstrap analysis The amplified COX II gene was approximately 570bp. was implemented using 1000 replicates of heuristic searches When PCR products were digested with seven restriction to determine the confidence levels of the inferred enzymes (HhaI, HinfI, MboI, TaqI, EcoRI, AfaI, AluI), no phylogenies. restriction sites for HhaI or HinfI were found in the COX II gene of any of the tested isolates (Fig. 2). The other restric- tion enzymes generated two to six profiles. Based on these Morphological observation profiles, the 62 isolates were divided into nine groups (Table 2). The isolates in groups C, D, E, F, and G, were Morphological characteristics were observed on grass blade identical to those placed in the same groups according to cultures (Waterhouse 1967). Grass blades were placed on the ITS-RFLP analysis. Groups A and B were additionally CMA inoculated with the Pythium isolate. After a 1-day separated into two subgroups by AfaI and TaqI digestion, incubation at 25°C, the colonized blades were transferred to respectively. 10ml of autoclaved pond water (pond water/distilled water 1:2) in a 9-cm petri dish. For each isolate, 30 oogonia were selected randomly, and the diameter of oogonia and 178

Table 2. Grouping of the tested isolates on the basis of banding patterns after restiction digestion of rDNA-ITS region and cytochrome c oxidase II gene DNA group Banding pattern

rDNA-ITS region COX II gene

HhaI HinfI MboI TaqI HhaI HinfI MboI TaqI EcoRI AfaI AluI

A-1 2 2 2 2 1 1 211 22 A-2 2 2 2 2 1 1 211 12 B-1 1 1 1 1 1 1 211 11 B-2 1 1 1 1 1 1 221 11 C32111123123 D43111121134 E22211111235 F11211121146 G12131121116 ITS, internal transcribed spacer; COX II, cytochrome c oxidase subunit II

Fig. 2. Restriction banding patterns of PCR-amplified cytochome oxi- dase II gene using five enzymes. a MboI. b TaqI. c EcoRI. d AfaI. e AluI. Lanes: M, 50-bp DNA ladder marker; 1–6, numbers of the banding patterns

Fig. 3. Phylogenetic tree of Pythium graminicola and related species using the rDNA-ITS region sequence data based on maximum parsi- mony analysis. The number of parsimony-informative characters was 177. The parsimony analysis produced one minimum length tree of 303 Phylogenetic analyses steps, with a consistency index, homoplasy index, and retention index of 0.7723, 0.2277, and 0.7244, respectively. DNA groups correspond to Phylogenetic trees generated using maximum parsimony P. graminicola (A), P. arrhenomanes (B, C), P. aphanidermatum (D), analysis of rDNA-ITS placed 30 isolates in seven clades P. myriotylum (E), P. torulosum (F), and P. vanterpoolii (G) (Fig. 3). Each clade agreed with the grouping established through RFLP analysis of rDNA-ITS sequences. Subgroups Group E isolates were associated with those in group F. of isolates in groups A and B could not be differentiated Group D isolates formed a single clade distant from the in the analysis. Isolates in group B were closely related other groups. to those in group C, and these isolates were clustered Phylogenetic analyses of COX II gene sequences together with isolates in group G within a large clade. gave rise to trees similar to those generated from rDNA- 179

Table 3. Comparison of morphological characteristics of the DNA groups of Pythium graminicola and related species DNA Oogoium Oospore Antheridium group sizea (µm) Sizea Status in No. per Attachment Attachment (µm) oogonium oogoniuma to hyphab to oogoniumc

A 23.6 (13–34) a 21.8 (13–29) a Plerotic 2.9 (1–7) c T Mo & Di B 24.4 (14–45) a 21.0 (12–34) a Plerotic 6.4 (1–16) ab T Di C 24.4 (20–33) a 22.0 (14–30) a Plerotic 3.2 (2–12) bc T Di D 23.4 (13–28) a 19.1 (13–25) ab Aplerotic 1.2 (1–3) c I and T Mo and Di E 26.0 (15–35) a 20.5 (12–32) a Aplerotic 6.8 (1–16) a T Di F 15.1 (11–20) b 13.9 (9–17) c Plerotic 1.3 (1–5) c T Mo G 18.7 (12–22) b 16.9 (10–19) bc Plerotic 1.3 (1–9) c T Mo a Values followed by different letters are significantly different (P ϭ 0.05) according to the Tukey test b T, terminal; I, intercalary c Mo, monoclinous; Di, diclinous

Morphology

Morphological characteristics among 47 representative iso- lates of the seven DNA groups were compared under the same cultural conditions (Table 1, Fig. 5). There were few differences in the morphological characteristics between the isolates in each group and between the subgroup iso- lates of the groups A and B. However, the size of oogonia and oospores and the number of antheridia per oogonium differed significantly among the DNA groups (Table 3). The characteristic features of group A isolates were larger oogonia, plerotic oospores, and fewer antheridia per oogonium. Group B isolates were similar to those in group A but differed in the number of antheridia per oogonium. Group C isolates were indistinguishable from those in groups A and B because the number of antheridia was intermediate between groups A and B, and other character- istics were similar. Group D isolates were characterized by aplerotic oospores, intercalary antheridia, and fewer antheridia per oogonium. Only group D isolates produced intercalary antheridia. Group E isolates differed from those in the other groups by having a more antheridia per oogonium and also aplerotic oospores. Group F and G iso- lates produced significantly smaller oogonia and oospores and fewer antheridia per oogonium than the other groups’ isolates. Isolates in group F were distinguished from group G isolates by the formation of branched sporangia and antheridial stalks that originated close to the oogonium (Fig. 5).

Fig. 4. Phylogenetic tree of Pythium graminicola and related species using the COX II gene sequence data based on maximum parsimony Discussion analysis. The number of parsimony-informative characters was 98. The parsimony analysis produced one minimum length tree of 172 steps, with a consistency index, homoplasy index, and retention index of A total of 62 isolates tested in the present study were 0.7151, 0.2849, and 0.9151, respectively. DNA groups correspond to the Pythium spp. shown in Fig. 3. divided into seven groups based on the RFLP analysis of the rDNA-ITS region. The banding patterns in groups A, B, D, F, and G were identical to those previously reported for P. graminicola, P. arrhenomanes, P. aphanidermatum, ITS sequences (Fig. 4). Group B isolates formed a large P. torulosum, and P. vanterpoolii, respectively (Chen et al. clade together with isolates from groups C and G. Group 1992; Wang and White 1997). Moreover, two isolates D isolates were quite distinct from those in the other (ATCC96234 and ATCC96600) of P. graminicola and three groups. (ATCC96526, ATCC96598, ATCC96525) of P. arrheno- 180

Fig. 5. Morphological character- istics of Pythium graminicola and related species on , oogonium, antheridium, and oospore. DNA groups were divided by restriction fragment length polymorphism (RFLP) analyses of the rDNA-ITS region and the COX II gene. DNA groups correspond to P. graminicola (A), P. arrheno- manes (B, C), P. aphaniderma- tum (D), P. myriotylum (E), P. torulosum (F), and P. vanterpoolii (G). Bars 30µm

manes (Table 1) had the same molecular analysis results as P. myriotylum isolates described by Chen and Hoy (1993). those found for P. arrhenomanes by Chen and Hoy (1993), This discrepancy may be due to intraspecific variation. A and the molecular characteristics were identical. The greater sampling of P. myriotylum isolates are required to patterns among group E isolates, which varied little in our resolve these discrepancies. experiment, were identical to those of P. myriotylum In the RFLP analysis of the COX II gene, our isolates provided by Wang and White (1997) but not to the segregated into groups similar to three established from the 181 ITS-RFLP analysis. We predicted that rDNA-ITS identified as P. arrhenomanes, indicating that P. arrheno- sequences would be more variable than COX II gene manes would be widely distributed and one of the more sequences because the rDNA-ITS region is found in the important pathogens of turf grasses. nucleus, whereas the COX II gene is found in mitochondria, Isolates that had been originally identified as P. which would be expected to evolve more slowly owing to graminicola (based on traditional morphological criteria) maternal inheritance of mitochondrial genes. Nonetheless, were divided into five and seven groups based on the RFLP the phylogenetic trees of the rDNA-ITS region and the and phylogenetic analyses of the rDNA-ITS region and the COX II gene had similar topologies, confirming the findings COX II gene, respectively. Because P. graminicola is mor- of Martin and Tooley (2000, 2003) in Pythium and phologically quite similar to P. arrhenomanes, isolates of P. Phytophthora species. graminicola are commonly misidentified as P. arrheno- Grouping of isolates based on RFLP analyses agreed manes (Chen and Hoy 1993), pointing to the difficulty with the groupings obtained from morphological chara- of relying solely on morphological characteristics for cteristics. Groups A, B, D, E, F, and G were individually identification. Our phylogenetic analysis demonstrated identified as P. graminicola, P. arrhenomanes, P. aphanider- that P. arrhenomanes is phylogenetically distant from P. matum, P. myriotylum, P. torulosum, and P. vanterpoolii, graminicola and more closely related to P. vanterpoolii. respectively. The morphology of group C isolates was inter- The oogonia of P. vanterpoolii and P. torulosum were sig- mediate between P. arrhenomanes and P. graminicola. P. nificantly smaller than those of the other species. However, arrhenomanes is commonly distinguished from P. gramini- P. vanterpoolii was phylogenetically related to P. arrheno- cola by the number of antheridia per oogonium. However, manes with large oogonia, rather than to P. torulosum. the number of the antheridia per oogonium among group Furthermore, P. torulosum was more closely related to C isolates did not significantly differ from those of group P. myriotylum with large oogonia than to P. vanterpoolii. P. A and B isolates. Chen and Hoy (1993) described P. aphanidermatum, with large oogonia, was more distantly arrhenomanes isolates that had fewer antheridia per oogo- related to other species with large oogonia. The results nium. The banding patterns from restriction digests of suggest that the similarity in the morphological characteris- rDNA-ITS sequences in group C were identical. Moreover, tics do not always reflect the genetic relationship between P. group C isolates were closely related to those in group B graminicola and related species. Hendrix and Papa (1974) on the phylogenetic trees of the rDNA-ITS region and combined P. graminicola with related species into what they the COX II gene, suggesting that group C isolates are P. called the P. graminicola complex group. However, six of arrhenomanes. the species included in the complex by Hendrix and Papa The type culture ATCC11101 of P. aristosporum was could be clearly delineated using our rDNA-ITS and COX also tested in the present study. This species has been re- II gene analyses. ported as an important pathogen of turf grasses (Feng Our results indicate that P. graminicola and closely re- and Dernoeden 1999; Hodges and Coleman 1985; Nelson lated species are phylogenetically distinct and should retain and Craft 1991). Van der Plaats-Niterink (1981) distin- their individual species status. Because of their importance guished P. aristosporum from P. arrhenomanes on the basis as pathogens of damping-off and root rot diseases on grami- of fewer antheridia per oogonium, both monoclinous and neous crops and turf grasses, accurate delineation of these diclinous antheridia, and aplerotic oospores. However, species can help resolve questions of their ecology and when the morphologies of P. aristosporum and P. pathogenicity. At this time, we believe that proper identifi- arrhenomanes were compared under the same culture con- cation of these species can be accomplished only by combin- ditions, the species were identical. Based on the RFLP ing molecular analyses such as those described in this report analysis of the COX II gene, P. aristosporum was distin- with traditional morphological methods. guished as a subgroup, B-2, of group B isolates, which also included the B-1 subgroup, which were newly identified as Acknowledgments We thank Dr. J.W. Kim, Seoul National Univer- P. arrhenomanes. In our phylogenetic analyses, subgroup B- sity, Dr. M. Tojo, Osaka Prefecture University, and Dr. A. Chikuo, Japanese National Institute of Floricultural Science, for providing the 2 isolates fell into the same clades as subgroup B-1 isolates isolates. This work was supported in part by a Grant-in-Aid for on both the ITS and COX II trees. These results indicate Scientific Research (B) (15310024) from the Ministry of Education, that P. aristosporum is indeed closely related to P. Science, Sports, and Culture of Japan. arrhenomanes both morphologically and evolutionarily. Pythium graminicola and related species are associated with various turf grass diseases but vary widely in their References virulence (Aoyagi et al. 1999; Feng and Dernoeden 1999; Hodges and Campbell 1993; Hodges and Coleman 1985). Abad ZG, Shew HD, Lucas LT (1994) Characterization and patho- Nelson and Craft (1991) and Hsiang et al. (1995) reported genicity of Pythium species isolated from turfgrass with symptoms of root and crown rot in North Carolina. Phytopathology 84:913– that P. graminicola was a major cause of Pythium root and 921 crown rot in bentgrass turf. Abad et al. (1994) and Hodges Aoyagi A, Kageyama K, Hyakumachi M (1999) Isolation of Pythium and Coleman (1985) found that P. arrhenomanes caused species from zoysia grass and their effect on severity of large patch serious disease of bentgrass in high-sand-content greens disease. Plant Dis 83:171–175 Chen W, Hoy JW (1993) Molecular and morphological comparison of and at high summer temperatures. In the present study, 19 and P. graminicola. Mycol Res 97:1371– of 23 isolates of P. graminicola from turf grasses were newly 1378 182

Chen W, Hoy JW, Schneider RW (1992) Species-specific polymor- mitochondrially encoded cytochrome oxidase I and II gene. Mycolo- phisms in transcribed ribosomal DNA of five Pythium species. Exp gia 95:269–284 Mycol 16:22–34 Matthews VD (1931) Studies on the genus Pythium. University of Dick MW (1990) Keys to Pythium. University of Reading Press, Read- North Carolina Press, Chapel Hill, NC ing, UK Matumoto C, Kageyama K, Suga H, Hyakumachi M (1999) Phylo- Feng Y, Dernoeden PH (1999) Pythium species associated with root genetic relationships of Pythium species based on ITS and 5.8S dysfunction of creeping bentgrass in Maryland. Plant Dis 83:516–520 sequences of ribosomal DNA. Mycoscience 40:321–331 Hendrix FF Jr, Papa KE (1974) Taxonomy and genetics of Pythium. Matumoto C, Kageyama K, Suga H, Hyakumachi M (2000) Intras- Proc Am Phytopathol Soc 1:200–207 pecific DNA polymorphisms of . Mycol Res Hodges CF, Campbell DA (1993) Infection of adventitious roots 104:1333–1341 of palustris by Pythium species at different temperature Middleton JT (1943) The taxonomy, host range, and geographic distri- regimes. Can J Bot 72:378–383 bution of the genus Pythium. Memoirs Torrey Bot Club 20:1–171 Hodges CF, Coleman LW (1985) Pythium-induced root dysfunction of Nelson EB, Craft CM (1991) Identification and comparative pathoge- secondary roots of Agrostis palustris. Plant Dis 69:336–340 nicity of Pythium spp. from roots and crowns of turfgrasses exhibit- Hsiang T, Wu C, Yang L, Liu L (1995) Pythium root rot associated with ing symptoms of root rot. Phytopathology 81:1529–1536 cool-season dieback of turfgrass in Ontario and Quebec. Can Plant Swofford DL (2001) PAUP: phylogenetic analysis using parsimony, Dis Surv 75:191–195 V4.0. Sinauer Associates, Sunderland, MA Kageyama K, Uchino H, Hyakumachi M (1998) Characterization of Van der Plaats-Niterink AJ (1981) Monograph of the genus Pythium. hyphal swelling group of Pythium: DNA polymorphisms and cultural Stud Mycol 21:1–242 and morphological characteristics. Plant Dis 82:218–222 Wang PH, White JG (1997) Molecular characterization of Pythium Kageyama K, Suzuki M, Priyatmojo A, Oto Y, Ishiguro K, Suga H, species based on RFLP analysis of the internal transcribed spacer Aoyagi T, Fukui H (2003) Characterization and identification of region of ribosomal DNA. Physiol Mol Plant Pathol 51:129–143 asexual strains of Pythium associated with root rot of rose in Japan. Waterhouse GE (1967) Key to Pythium Pringsheim; Commonwealth J Phytopathol 151:485–491 Mycological Institute. Mycol Paper 109:1–15 Martin FN (2000) Phylogenetic relationships among some Pythium White TJ, Bruns T, Lee S, Taylor J (1990) Amplification and direct species inferred from sequence analysis of the mitochondrially sequencing of fungal ribosomal RNA genes for phylogenetics. In: encoded cytochrome oxidase II gene. Mycologia 92:711–727 Innis MA, Gelfand DH, Sninsky JJ, White TJ (eds) PCR protocols: Martin FN, Tooley PW (2003) Phylogenetic relationships among some a guide to methods and applications. Academic, San Diego, pp 315– Phytophthora species inferred from sequence analysis of the 322