Two new entomopathogenic species of Ophiocordyceps in Thailand

Abstract Ophiocordyceps is entomopathogenic and the largest studied genus in the family Ophiocordycipitaceae. Many species in this genus have been reported from Thailand. The first new species introduced in this paper, Ophiocordycepsglobiceps, differs from other species based on its smaller perithecia, shorter asci and secondary ascospores and additionally, in parasitising fly species. Phylogenetic analyses of combined LSU, SSU, ITS, TEF1α and RPB1 sequence data indicate that O.globiceps forms a distinct lineage within the genus Ophiocordyceps as a new species. The second new species, Ophiocordycepssporangifera, is distinguished from closely related species by infecting larvae of insects (Coleoptera, Elateridae) and by producing white to brown sporangia, longer secondary synnemata and shorter primary and secondary phialides. We introduce O.sporangifera based on its significant morphological differences from other similar species, even though phylogenetic distinction is not well-supported.


Introduction
The genus Ophiocordyceps was introduced by Petch (1931) to accommodate species which have different features of asci and ascospores from Cordyceps (Petch 1931). Ophiocordyceps was treated as a subgenus of Cordyceps by Kobayasi (1941Kobayasi ( , 1982 and Mains (1958). Sung et al. (2007a) established the new family Ophiocordycipitaceae in Hypocreales (Sordariomycetes) and revised Ophiocordyceps as the type genus based on phylogenetic analyses. This is followed in the Outline of Ascomycetes (Wijayawardene et al. 2018). The main characters of the sexual morph species of Ophiocordyceps are fibrous, hard, pliant-to-wiry, dark stromata with superficial to immersed perithecia (Sung et al. 2007a, Ban et al. 2015. The asexual morphs in the majority of species have hirsutella-like and hymenostilbe-like features (Kepler et al. 2013, Maharachchikumbura et al. 2015. The hosts of species in Ophiocordyceps are larval lepidopterans and coleopterans, adult hymenopterans, hemipterans, dipterans, orthopterans or dragonflies (Odonata) and, in few cases, spiders (Kobayasi 1941, Mains 1958, Sung et al. 2007a, Ban et al. 2015. Hitherto, Ophiocordyceps included 233 species (Index Fungorum, June 2018) with a worldwide diversity (Sung et al. 2007a, Ban et al. 2015, Shrestha et al. 2017. Thailand is located in the tropical areas with a rich biodiversity (Luangsa-ard et al. 2008, Aung et al. 2008, Luangsa-ard et al. 2010. A variety of entomopathogenic species (more than 400 species) (Index Fungorum, June 2018, Luangsa-ard et al. 2008, Luangsa-ard et al. 2010 were reported from Thailand after the first species recorded by Petch in 1932. In this study, we introduce two new species of Ophiocordyceps, which were found on larvae of insects (Lepidoptera, Cossidae) and adult Diptera. The descriptions of these two new species and phylogenetic evidence for the new taxa are provided. Morphological differences between two new species and their related species are also discussed.

Collection, isolation, and morphology study
Specimens were collected in The Mushroom Research Centre, Chiang Mai, Thailand, from soil and grass litter and taken to the laboratory. Fruiting bodies were examined using free hand sections under a stereomicroscope. Water-mounted slides were prepared for a microscope study and photographed under a compound microscope. Strains were isolated from single spores by using the protocol in Chomnunti et al. (2014). Cultures were incubated at 25 °C for 4-10 weeks on potato extract agar (PDA) in light-promoted sporulation.
DNA extraction, PCR amplification and determination of DNA sequences DNA was extracted from both dried specimens and cultures by using E.Z.N.A.TM Fungal DNA MiniKit (Omega Biotech, CA, USA), according to the manufacturers proto-cols. Universal known primers were used in PCR amplification; ITS4/ITS5 for internal transcribed spacer gene region (ITS), NS1/NS4 for partial small subunit ribosomal RNA gene region (SSU), LROR/LR5 for partial large subunit rDNA gene region (LSU) (Vilgalys andHester 1990, White et al. 1990), 983F/2218R for partial translation elongation factor 1-alpha gene region (TEF1α) (Sung et al. 2007b) and CRPB1A/RPB1Cr for partial RNA polymerase II largest subunit gene region (RPB1) (Castlebury et al. 2004). PCR products were sequenced by Sangon Biotech (Shanghai) Co., Ltd., Shanghai, China. Specimen was performed by using TaKaRa PMD18-T vector system (TaKaRa Biotechnology, Dalian, China), while PCR products could not be sequenced directly.

Phylogenetic analyses
Sequence data were obtained from GenBank based on previous studies as listed in Table 1. MAFFT v.7 was used to align combined datasets of ITS, SSU, LSU, TEF1α and RPB1 regions (Katoh and Standley 2013, http://mafft.cbrc.jp/alignment/server/). BioEdit (Hall 2011) was used to check alignment manually. Gaps were treated as missing data. Tolypocladium inflatum W. Gams and T. ophioglossoides (J.F. Gmel.) C.A. Quandt et al. (Kepler et al. 2012, Schoch et al. 2012 were selected as outgroup taxa. Maximum likelihood trees (ML) were estimated by using the software RAxML 7.2.8 Black Box (Stamatakis 2006, Stamatakis et al. 2008 in the CIPRES Science Gateway platform (Miller et al. 2010). MrModeltest v.2.3 (Nylander 2004) was used to determine the best-fit model of evolution for Bayesian analyses. MrBayes v.3.1.2 (Ronquist and Huelsenbeck 2003) was used to evaluate posterior probabilities (PP) (Rannala andYang 1996, Zhaxybayeva andGogarten 2002) by Markov Chain Monte Carlo sampling (BMCMC). Six simultaneous Markov chains were run for 10,000,000 generations, trees were sampled every 100 th generation and 100,001 trees were obtained. The first 25% of trees (25,000) were discarded, as they represented the burn-in phase of the analyses, while the remaining trees (75,001) were used for calculation of posterior probabilities in the majority rule consensus tree (critical values for the topological convergence diagnostic is 0.01). Trees were figured in FigTree v1.4.0 programme (Rambaut 2012). Bayesian Posterior Probabilities (BYPP) equal to or great than 0.90 were given below each node (Fig. 1).
Notes. Ophiocordyceps sporangifera is closely related to O. myrmicarum D.R. Simmons & Groden in our phylogenetic tree (Fig. 1). The morphology of O. sporangifera is different from O. myrmicarum in having longer primary and secondary synnemata, a white to brown sporangium, shorter phialides and it infects insect larvae (Lepidoptera, Cossidae), while O. myrmicarum was found on an ant (Myrmica rubra) (Simmons et al. 2015). The phylogenetic analysis does not have good support, but O. sporangifera is distinct from O. myrmicarum. In the phylogenetic tree, the relationships of O. sporangifera and O. myrmicarum are obscure because they share one clade with short branch length (100% ML/ 1 BYPP), while the two strains of O. sporangifera clustered together with a low bootstrap support (88% ML/ 0.90 BYPP). The type strain of O. sporangifera has 0 bp in nrSSU, 3 bp in TEF1α and 5 bp in RPB1 that are different from O. myrmicarum. However, the morphological features of those two species are different, thus, they should be treated as two separate species (Table 3).

Discussion
We introduce two new entomopathogenic species of Ophiocordyceps, one from Coleoptera (Elateridae) and the other from flies (Diptera). Morphological and phylogenetic analyses have provided insights to resolve generic delimitation (Sung et al. 2007a, Jeewon and. Most of the species of this genus are parasitic on insects (Sung et al. 2007a, Maharachchikumbura et al. 2015, Wijayawardene et al. 2017). The sexual morph species in this genus is characterised by fibrous, hard, pliant-to-wiry, darkcoloured stroma with superficial to immersed perithecia (Sung et al. 2007a, Ban et al. 2015, Maharachchikumbura et al. 2015, while the asexual morph species have mainly hymenostilbe-like and hirsutella-like features, branched or unbranched phialides with oval to fusiform conidia (Kepler et al. 2013, Maharachchikumbura et al. 2015.

Ophiocordyceps myrmicarum Ophiocordyceps sporangifera Host
Myrmica rubra (Hymenoptera) Elateridae, Coleoptera Primary synnemata Whitish-yellow aging to rufous brown 9-18 cm high 1-2 mm diam., brown to deep brown Secondary synnemata (μm) Hyaline aging to rufous brown, up to 350 long, narrow (25) at base, common on agar but not observed on host Brown to white, not smooth 1092-1937 × 21-34, arising from the all parts of the primary synnemata, observed on both of the host and agar Primary phialides (μm) Subulate, hyaline or pigmented at base, 39.9-86.2 long, 3.6-5.4 wide at base Slender,solitary,hyaline,unbranched,narrow,smooth, Composed of 1-4 conidia, hyaline to brown, at phialide apex 10.5-12.9 × 6.4-8.7, composed of 1-12 conidia, hyaline on host, 1-4 conidia on culture, hyaline to brown on culture Reference Simmons et al. 2015 This study fers from closely related species by producing capitate, stipitate ascostromata, vertical, narrowly ovoid to obclavate, occasionally irregular perithecia and cylindrical secondary ascospores. Both morphology and phylogenetic analyses clearly show O. globiceps as a new species within Ophiocordyceps. Ophiocordyceps sporangifera is an asexual morph species and groups with O. myrmicarum in the phylogenetic tree (Fig. 1). Ophiocordyceps sporangifera can be distinguished from O. myrmicarum by infecting and parasitising larvae of insects (Lepidoptera, Cossidae), producing white to brown sporangium, longer primary and secondary synnemata and shorter primary and secondary phialides. The new species can be defined based on the distinctive morphological characters even through the phylogenies are not well-supported . In case of intricate differences between a gene tree and a species tree and, in addition, several morphs can be under the influence of many genes which are not really being reflected in the phylogeny . In our study, morphological characters strongly support O. sporangifera as a new species within Ophiocordyceps, even through phylogenetic analysis is not well-resolved. In this case, other loci which have more phylogenetic variation than the current loci may be able to differentiate these two species.