Introducing a new pleosporalean family Sublophiostomataceae fam. nov. to accommodate Sublophiostoma gen. nov.

Collections of microfungi on bamboo and grasses in Thailand revealed an interesting species morphologically resembling Lophiostoma, but which can be distinguished from the latter based on multi-locus phylogeny. In this paper, a new genus, Sublophiostoma is introduced to accommodate the taxon, S. thailandica sp. nov. Phylogenetic analyses using combined ITS, LSU, RPB2, SSU, and TEF sequences demonstrate that six strains of the new species form a distinct clade within Pleosporales, but cannot be assigned to any existing family. Therefore, a new family Sublophiostomataceae (Pleosporales) is introduced to accommodate the new genus. The sexual morph of Sublophiostomataceae is characterized by subglobose to hemisphaerical, ostiolate ascomata, with crest-like openings, a peridium with cells of textura angularis to textura epidermoidea, cylindric-clavate asci with a bulbous or foot-like narrow pedicel and a well-developed ocular chamber, and hyaline, fusiform, 1-septate ascospores surrounded by a large mucilaginous sheath. The asexual morph (coelomycetous) of the species are observed on culture media.

Colonies on PDA medium showing slow growth, 30-35 mm diam. after 4 weeks at 25-30 °C, white to pale yellow at the magins, white to grey in the center, with dark grey turfing in the middle of the colony; reverse white at the outer margin, more grey towards the inner area, becoming yellowish-grey in the center, dense, irregular, slightly raised to umbonate, dull with undulate edge, velvety, not produced pigmentation on agar.

Discussion
Phylogenetic analyses clearly indicate that six strains of Sublophiostoma thailandica form a distinct clade outside the suborders Massainae and Pleosporinae, but within Pleosporales (Fig. 1). The clade of Sublophiostoma is phylogenetically closely related to Neomassarinaceae and Sporormiaceae (Fig. 1). However, Sublophiostoma differs from Neomassarinaceae in having black, subglobose, hemisphaerical to lenticular, glabrous ascomata, a peridium of cells of textura angularis to textura epidemoidea at the inner layers and cylindric-clavate asci with a long pedicel. Genera in Neomassarinaceae have light brown to brown, coriaceous ascomata, a peridium with cells of textura angularis, cylindrical to cylindric-clavate asci, with a short pedicel, and a narrow sheath or surrounded by hyaline gelatinous sheath 1,24 . Sporormiaceae differs from Sublophiostoma in having globose to pyriform, perithecioid or cleistothecioid, ascolocular pseudothecia, membraneous or coriaceous ascomata, with or without an ostiolar canal, a peridium of darkly pigmented cells of textura angularis, clavate, globose or cylindrical asci and dark brown, usually septate and poly-celled, muriform ascospores, with oblong, suboboviod, hyaline to brown, 1-septate conidia in its asexual morph 1 . Sublophiostoma shares similar characteristics with     25 ) should be considered when discussing new families in Dothideomycetes. In Sublophiostomataceae the pseudoparaphyses are clearly trabeculate (Fig. 1), in Neomassarinaceae they are cylindrical to filiform, septate, branching, cellular pseudoparaphyses and in Lophiostomataceae they are septate, long, hyaline, anastomosing and branched, cellular pseudoparaphyses, embedded in gelatinous matrix between and above the asci.
All the strains of Sublophiostoma thailandica were found in Chiang Mai and Chiang Rai Provinces which are located in the north of Thailand and therefore we cannot conclude whether members of Sublophiostomataceae are distributed in other parts of Thailand or worldwide. More taxa of this genus from diverse localities should be studied to understand the distribution of Sublophiostomataceae.
Considering the MP tree in Fig. 2 and number of base pair differences, it can be postulated that the six strains of Sublophiostoma thailandica could represent a species complex. Within the ITS regions, MFLUCC 11-0174 strain and MFLUCC 11-0185 strain have one nucleotide difference (< 1%) from MFLUCC 11-0207, MFLUCC 11-0165 and MFLUCC 11-0172. Across the TEF region, there are less than ten base pair differences (< 1%). Thus, for the time being, we considered the six strains as belonging to the same species, Sublophiostoma thailandica following the guidelines in Jeewon & Hyde 26 . Even the phylogenetic tree (Fig. 1) does support establishment of one species as all strains cluster together. However, it is worth mentioning that these six strains have more than ten base pair differences in RPB2. The latter gene region is well known for its DNA sequence variability and one can speculate that the above six strains could represent more than one species. Our MP tree (Fig. 2) depicts that these six strains can be segregated into 2 subclades but there is no support. These six strains are also slightly different in ascus shape and size and the sheath surrounding the ascospores but these could also be minor phenotypic variation within one species, a common phenomenon which happens across many fungi which are subjected to different in vitro environmental conditions. In order to consider them as distinct species or genetic variants of the same species, further in-depth analyses are needed.
The order Pleosporales presently comprises two suborders which are Massarineae and Pleosporinae, as well as many unresolved clades 1,2,8,13,19,23 . Many of these unresolved or unclear lineages have low bootstrap support, which is probably due to limited taxon-sampling 9,27 . An assessment of the morphological characters which is the traditional way of classifying fungi and DNA based phylogenetic analyses are important in determining the status of taxa and could help to resolve some of the taxonomic confusions that presently exist in Pleosporales.

Materials and methods
Collections, morphology and isolation. Specimens, morphologically similar to Lophiostoma, were collected in Chiang Mai and Chiang Rai, Thailand, and brought back to the laboratory in paper bags. Gross morphology was observed using a Motic SMZ 168 dissecting microscope. Vertical sections of ascomata were made by freehand and mounted in water on a slide to observe their microscopic features. Indian ink was used to observe the presence or lack of a mucilaginous sheath surrounding the ascospores. Congo red and cotton blue were used as stains to observe the internal structures and septations. Prepared slides were observed under a Nikon ECLIPSE 80i compound microscope, photographed using a Canon EOS 450D digital camera fitted to the microscope, and preserved in lacto glycerol after photographing. The photographed images were processed using Adobe Photoshop CS5 Extended version 12.0 software (Adobe Systems Inc., The United States). Measurements were determined using Tarosoft Image Frame Work program v. 0.9.7. Single spore isolation was carried to obtain pure culture, using the methods mentioned in Chomnunti et al. 28 and Senanayake et al. 29 in a Petri-dish containing 2% water agar (WA) and incubated overnight at 20-25 °C. Germinating ascospores were then transferred aseptically onto potato dextrose agar (PDA; 39.0 g/L sterile distilled water, Difco potato dextrose). The production of asexual morphs was facilitated as described in Phookamsak et al. 30  DNA extraction, amplification, sequencing. Fresh mycelium grown on PDA was scraped and used to extract DNA, following the manufacturer's standard protocol as described in the Biospin Fungus Genomic DNA extraction kit (BioFlux, Hangzhou, P.R. China). Polymerase chain reactions (PCR) were carried out using the following primers: LR0R and LR5 to amplify the 28s subunit rDNA (LSU) 14,32 , NS1 and NS4 to amplify the 18s subunit rDNA (SSU), ITS4 and ITS5 to amplify the internal transcribed spacers (ITS), EF1-983F and EF1-2218R to amplify the translation elongation factor 1 (TEF), and fRPB2-5F and fRPB2-7cR for RNA polymerase II subunit 2 (RPB2) 33,34 . The amplification reaction was performed in a 25 μl reaction volume containing 2 µl DNA, 12.5 μl 2 × Easy Taq PCR SuperMix polymerase and PCR buffer mix, 8.5 μl Distilled-Deionized-Water (ddH 2 O) and 1 µl of each primer. The thermal cycling program for ITS, LSU, SSU and TEF were as follows: an initialization step of 94 °C for 3 min, followed by 40 amplification cycles of 94 °C for 30 s, annealing step at 55 °C for 50 s, elongation step at 72 °C for 1 min and a final extension step of 72 °C for 10 min. The thermal cycling program for RPB2 amplification was performed following the protocols described in Phookamsak et al. 35 . The quality of the PCR products were checked on 1% agarose gel, stained with ethidium bromide. Sequencing was carried out at Shanghai Sangon Biological Engineering Technology & Services Co., Ltd (Shanghai, P. R. China). Sequence data generated from this study are deposited in GenBank. www.nature.com/scientificreports/ sequences in this study, strains with closest identity in GenBank and related strains in recent publications were selected and included in the sequence dataset (Table S1). Phylogenetic tree (Fig. 1), the single gene sequence data were compiled using BioEdit v. 7.0.9.0 36 , including the new collections. Capnodium coffeae (CBS 147.52) was used as outgroup taxon. Each single gene data set was aligned using MAFFT (http:// mafft. cbrc. jp/ align ment/ server/ large. html) 36 , and checked manually using BioEdit 37 . As the topology was similar in each evaluated gene region, a combined alignment of the five loci was compiled into one dataset. Maximum likelihood (ML) and Bayesian Inference (BI) analyses, a partitioned analysis was performed with the following five partitions: ITS, LSU, SSU, RPB2 and TEF sequence data. Maximum-likelihood (ML) analysis was carried out in the CIPRES Science Gateway web server (RAxML-HPC2 on XSEDE 38 ), and 1000 rapid bootstrap replicates were run with GTRGAMMA model of nucleotide evolution. Maximum likelihood bootstrap values (MLBS) equal or greater than 60% are presented at each node in the resulting phylogenetic trees. The model of evolution was performed by using MrModeltest 2.3 39 . GTR + I + G model was selected as a best-fit model for the all gene regions used in our analyses. Bayesian inference (BI) analysis was performed in the CIPRES Science Gateway web server (MrBayes on XSEDE v. 3.2.7a 38 ). The Markov Chain Monte Carlo sampling (MCMC) analyses, with four chains starting from random tree topology, were run between 10,000,000 generations for each combined dataset. Trees were sampled every 100 generations using a relative burn-in discarding the first 25% of sampled trees 40 . Bayesian posterior probabilities equal or greater than 0.90 are given at each node (Fig. 1).

Phylogenetic analyses.
In Fig. 2, maximum parsimony analysis (MP) was performed for the species of Sublophiostoma using PAUP (Phylogenetic Analysis Using Parsimony) v. 4.0b10 41 using a combined dataset of the LSU, SSU, ITS, TEF and RPB2. The heuristic search option with 1,000 random sequences addition and tree-bisection reconnection (TBR) of branch-swapping algorithm were performed. Maxtrees were setup to 1000, branches of zero length were collapsed. Gaps were treated as missing data and all characters were unordered and of equal weight. All multiple and equally parsimonious trees were saved. Descriptive tree statistics for parsimony consistency index (CI), retention index (RI), rescaled consistency index (RC) and homoplasy index (HI) were calculated. The robustness of the most parsimonious tree was evaluated by 1000 bootstrap replications resulting from maximum parsimony analysis, each with ten replicates of random stepwise addition of taxa 42 . The resulting phylogram (the first of 1,000 trees) is presented in Fig. 2. In addition, Bayesian inference (BI) analysis was performed for Fig. 2 following the method mentioned above but with 1,000,000 generations.