Molecular and morphological characterization of Phyllactinia cassiae-fistulae ( Erysiphaceae ; Ascomycota ) from Thailand

Phyllactinia cassiae-fistulae and its Ovulariopsis anamorph, a causal agent of powdery mildew on Cassia fistula, have been found in Thailand for the first time. Phylogenetic analysis using the 28S ribosomal DNA sequences clearly demonstrated that P. cassiae-fistulae distinctly formed a unique clade at the basal part of Phyllactinia with 100% bootstrap support. This phylogenetic analysis supports the unique morphology of P. cassiae-fistulae anamorph having cylindrical-ellipsoil conidia and short conidiophores similar to Oidium species.


INTRODUCTION
During the survey of powdery mildews from 2008 to 2011 in Northern Thailand, several interesting powdery mildews were discovered.One of them has been found on Cassia fistula and Senna siamea (Caesalpinioideae; Fabaceae) and was identified as Phyllactinia cassiaefistulae.This species was first described by Paul and Thakur (2006) in India as a new variety, P. bauhiniae var.cassia, and later revised as P. cassiae-fistulae by Braun and Paul (2009).Kirschner and Chen (2008) demonstrated first record of this species on C. fistula in Taiwan (without teleomorphic stage) and reported detailed morphological characteristics of anamorphic stage.Anamorph of this fungus has a unique characteristic that is conspicuously distinct from all other species of Phyllactinia, but produced Phyllactinia teleomorph.
In this study, molecular analysis combined with morphological analysis was performed to clarify taxonomy of P. cassiae-fistulae.This study is the first report of P. cassiae-fistulae from Thailand, and also the first report of this species on S. siamea in the world.

Sample sources
Specimens were collected in the northern Thailand (Chiang Mai Province) from December to March during 2009.All herbarium specimens were deposited at the mycological herbarium in Department of Entomology and Plant Pathology, Faculty of Agriculture, Chiang Mai University, Thailand and Mie University Mycological Herbarium (MUMH), Japan.

Morphological observation
Fresh specimens of powdery mildew on C. fistula leaves were examined by using a light microscope with 20 and 40x objective phase contras t lenses.Morphological observation on fungal colonies of anamorphic stage was stripped off from the leaf surfaces with clear adhesive tape or with a clean needle on teleomorphic stage, mounted on a microscope slide in distilled water.Morphological characteristics were measured in 30 replicates for each structure on anamorph: size and shape of conidia, conidiophore; position of the basal septum; shape and position of hyphal appressoria and presence or absence of fibrosin bodies (Toanun et al., 2005); on teleomorph: size and shape of chasmothecia, appendages, asci, ascospores (To-anun et al., 2003).Observation of conidial and ascospore germ tubes were carried out using the method of Hirata (1942).

Molecular phylogenetic analysis
Whole-cell DNA was extracted from mycelia or conidia using the chelex method (Walsh et al., 1991;Hirata and Takamatsu, 1996).The 28S ribosomal DNA (rDNA) including the domains D1 and D2, and ITS region including the 5.8S rDNA were amplified by the polymerase chain reaction (PCR) using nested primer sets.PCR reactions were conducted with TaKaRa Taq DNA polymerase (TaKaRa, Tokyo) under the following thermal cycling conditions in a PCR thermal cycler SP (Takara, Kyoto, Japan): an initial step for denaturing at 95°C for 2 min; thermocycling for 30 cycles that each cycle consisted of 30 s at 95°C followed by 30 s at 52°C for annealing, and 30 s at 72°C for extension; and a final extension cycle at 72°C for 7 min.
For amplification of the ITS regions, primer sets of ITS1, ITS4, ITS5, p3, PM6 and Ph7 were used for amplification.A Phyllactinia and Leveillula specific primer Ph7 (TGTTGCTTTGGYAGGCCG) was designed in this study.Primers ITS5 (White et al., 1990) and p3 (Kusaba and Tsuge, 1995) were used for the first amplification.Nested primer sets ITS5/PM6 and Ph7/ITS4 were used for the second amplification.
The nucleotide fragments of the second PCR products were sent to SolGent Co. (Daejeon, South Korea) for sequencing by using NL1 and NLP2 as sequence primers of 28S rDNA, and using ITS1 and ITS4 (White et al., 1990) as sequence primers of ITS regions.
The nucleotide sequences of rDNA were aligned with MUSCLE program (Edgar, 2004).Maximum parsimony trees were constructed from the alignment data matrix using parsimony ratchet method (Nixon, 1999) in PAUP 4.0b8 (Swofford, 2001) and PAUPRat ver. 1 (Sikes and Lewis, 2001).The strength of the internal branches of the resulting trees was tested by bootstrap analysis (Felsenstein, 1985) using 1,000 replications.Lack of bootstrap value indicates less than 50% support at that node.A tree with the highest likelihood value among the equal parsimonious trees was determined by PAUP 4.0 (Swofford, 2001).

Symptoms
The symptoms appeared on the lower side of the leaves by produce effuse, thin to dense white colonies from the end of November.Chasmothecia production (perfect stage) was seen from mid-January.Chasmothecia have not been found every year in the same tree.The symptoms were only found on leaves.The severe infected leaves caused early leaf defoliation.

Phylogenetic analysis
The 28S rDNA sequences consisted of two sequences from C. fistula and one sequence from S. siamea were aligned with 24 sequences of Leveillula, Phyllactinia and Pleochaeta retrieved from DNA database (Takamatsu et al., 2008).Pleochaeta shiraiana was used as an outgroup taxon based on Takamatsu et al. (2008).The alignment data matrix consisted of 27 sequences and 610 total characters.Of these, 518 characters were constant, 26 characters were variable and parsimony-uninformative, and 66 characters were informative for parsimony analysis.A total of 201 equally parsimonious trees (CI = 0.6628, RI = 0.8366, RC = 0.5545) with 172 steps were constructed by the parsimony ratchet analysis.A tree with the highest likelihood value among the 201 trees is shown in Figure 3. P. cassiae-fistulae sequences deposited in DDBJ under the accession number AB691226 and AB691227 including Ovulariopsis anamorph on S. siamea AB691228 distinctly formed an We also determined the rDNA ITS sequences for five samples of P. cassiae-fistulae on C. fistula and conducted FASTA search at the EMBL DNA database (http://www.ebi.ac.uk/embl/) using the sequences as queries.The highest similarities were obtained with P. angulata AB080566 (76.9%) and next with P. chubutiana AB243690 (75.8%).This result indicates that P. cassiaefistulae is genetically isolated among Phyllactinia species.Because we could not obtain unambiguous alignment of P. cassiae-fistulae with other Phyllactinia species in ITS sequences, we did not conduct phylogenetic analysis of ITS sequences.Sequences analysis of ITS region further support the isolated phylogenetic situation of P. cassiaefistulae among Phyllactinia species shown in the 28S rDNA analysis (Figure 3).

DISCUSSION
Several powdery mildew species have been reported on Cassia (Sattar and Hussain, 1976;Thaung, 2007;Zhao et al., 2010) in the world.However, there is no record of powdery mildew on Cassia in Thailand.This is the first report of powdery mildew on Cassia in Thailand.The morphological observations of anamorph of P. cassiae- fistulae demonstrated that the cylindrical-ellipsoid conidia are quite distinct from other known Phyllactinia species having lanceolate conidia (Braun, 1987;Paul and Thakur, 2006;Braun and Paul, 2009).However, this fungus produced chasmothecia having acicular appendages with a bulbous swelling at the base that is a typical character of Phyllactinia (Braun, 1987).
Oidium cassiae-siameae has been recorded as a powdery mildew on Cassia (Amano, 1986;Braun, 1987).Kirschner and Chen (2008) described and illustrated a powdery mildew on C. fistula and compared it with O. cassiae-siameae specimen.The result revealed that the powdery mildew on C. fistula has morphology similar to Oidium species, but quite differs from Oidium species by produced endophytic hyphae.This endophytic behavior is typical appearance of the tribe Phyllactinieae and its host species is the same with that reported as a host of P. cassiae-fistulae (Braun and Paul, 2009).
The present study is the first report of phylogenetic analysis of P. cassiae-fistulae.The result indicated that the 28S rDNA sequences from three P. cassiae-fistulae isolates on C. fistula and S. siamea formed an independent clade at the basal part of Phyllactinia/Leveillula clade with bootstrap support of 100%, and are sister to all other Phyllactinia and Leveillula sequences.This result may indicate that Phyllactinia is paraphyletic group as described by Takamatsu et al. (2008).Additionally, this phylogenetic clade showed the closest related between P. cassiae-fistulae on C. fistula and S. siamea.Therefore, molecular phylogenetic analysis based on the 28S rDNA sequences supported the unique anamorphic morphology of P. cassiae-fistulae.The isolated phylogenetic placement of P. cassiae-fistulae was also supported by the ITS sequence analysis.
A recent molecular phylogenetic study of the genera in the subtribe Cassiinae (Acharya et al., 2011) demonstrated that C. fistula and S. siamea are classified into Cassia sensu lato.The present study showing that both C. fistula and S. siamea are commonly infected by P. cassiae-fistulae supports the close relationship of the host plants.

Figure 3 .
Figure 3.A parsimony ratchet tree based on the 28S rDNA sequences for 27 taxa, consisting of three sequences from P. cassiae-fistulae and 24 sequences of Phyllactinia, Leveillula and Pleochaeta retrieved from DNA database.Horizontal branch lengths are proportional to the number of nucleotide substitutions that were inferred to have occurred along a particular branch of the tree.Bootstrap values (≥50%) are shown above branches (CI = 0.6628, RI = 0.8366, RC = 0.5545).