Molecular Phylogeny and Morphology Reveal Four Novel Species of Corynespora and Kirschsteiniothelia (Dothideomycetes, Ascomycota) from China: A Checklist for Corynespora Reported Worldwide

Plant debris are habitats favoring survival and multiplication of various microbial species. During continuing mycological surveys of saprobic microfungi from plant debris in Yunnan Province, China, several Corynespora-like and Dendryphiopsis-like isolates were collected from dead branches of unidentified perennial dicotyledonous plants. Four barcodes, i.e., ITS, LSU, SSU and tef1-α, were amplified and sequenced. Morphological studies and multigene phylogenetic analyses by maximum likelihood and Bayesian inference revealed three new Corynespora species (C. mengsongensis sp. nov., C. nabanheensis sp. nov. and C. yunnanensis sp. nov.) and a new Kirschsteiniothelia species (K. nabanheensis sp. nov.) within Dothideomycetes, Ascomycota. A list of identified and accepted species of Corynespora with major morphological features, host information and locality was compiled. This work improves the knowledge of species diversity of Corynespora and Kirschsteiniothelia in Yunnan Province, China.


Introduction
Hyphomycetes are highly diverse and distributed in terrestrial and freshwater habitats. More than 1500 Hyphomycetes genera and 30,000 species have been recorded worldwide [1,2]. These fungi show distinct morphological features, which often allow for species identification, as DNA sequences have been hitherto unavailable for most genera and species. Given the large amount of hyphomycetes, it is challenging to classify their taxonomic placement based on morphology alone because some of them may belong to the same species or even to different genera. The introduction of molecular phylogenetic analyses led to a better understanding of the heterogenous genera and species and further clarified their taxonomic status. Investigating fungal diversity is an important task in assembling the fungal tree of life (AFToL) [3], which contributes to the knowledge of biological diversity and the exploration and utilization of fungal resources.
Corynespora was established by Güssow [4] with C. mazei as the type species. Wei [5] provided a historical review and considered C. mazei a synonym of the previously described Helminthosporium cassiicola Berk. & M.A. Curtis and transferred the latter species, resulting in the new combination Corynespora cassiicola (Berk. & M.A. Curtis) C.T. Wei. This genus is mainly characterized by distinct, determinate or percurrently extending conidiophores and monotretic, integrated, terminal conidiogenous cells that produce solitary or sometimes catenate, distoseptate conidia [6]. To date, 198 epithets for Corynespora have been listed in Species Fungorum [7], but many species associated with leaf spots were defined at ogy and phylogenetic analyses. Hernandez-Restrepo et al. [49] proposed the monotypic order Kirschsteiniotheliales for Kirschsteiniotheliaceae due to its distant relation to other orders in Dothideomycetes.
Yunnan Province is located in southwestern China. It lies at 21 • 09 -29 • 15 N and 97 • 32 -106 • 12 E and includes vast territory with distinct climatic characteristics and abundant natural resources. Its average annual temperatures is 12-22 • C, and the total annual precipitation is approximately 1500 mm. Such favorable conditions support more than 18,000 higher plant species (51.6% of China's total) in this province, resulting in a very wide range of habitats favoring the growth of various microbial species. However, its mycobiota, especially microfungi, is poorly understood. During our survey of the taxonomy and diversity of saprobic microfungi in Yunnan Province, a Dendryphiopsis-like fungus and three Corynespora-like fungi were collected on dead branches from terrestrial habitats. Based on multilocus phylogenetic analyses and morphological characteristics, we introduced four novel species of Corynespora and Kirschsteiniothelia in Dothideomycetes. This study broadens our understanding of the diversity of Corynespora and Kirschsteiniothelia taxa.

Sample Collection, Isolation and Morphology
Samples of dead branches were collected randomly from humid environments and river banks, where there is a deep litter layer comprising rotten softwood, dead branches and decayed leaves of various plants in the forest ecosystems of Yunnan Province. Dead branches are a rich habitat for saprobic hyphomycetes. Samples were placed in Ziploc TM bags for transport to the laboratory, where they were processed and examined as described by Ma et al. [50]. Colonies on decaying wood surface were examined and visually observed with a stereomicroscope (Motic SMZ-168, Xiamen, China) from low (0.75 times) to high (5 times) magnification. Fresh colonies were picked with sterile needles at a stereomicroscope magnification of 5 times, placed on a slide with a drop of lactic acid-phenol solution (lactic acid, phenol, glycerin, sterile water; 1:1:2:1, respectively), then placed under an Olympus BX 53 light microscope fitted with an Olympus DP 27 digital camera (Olympus Optical Co., Tokyo, Japan) for microscopic morphological characterization. The tip of a sterile toothpick dipped in sterile water was used to capture the conidia of the target colony directly from the specimen; the conidia were then streaked on the surface of potato dextrose agar (PDA; 20% potato + 2% dextrose + 2% agar, w/v) and incubated in an incubator at 25 • C overnight. The single germinated conidia were transferred to fresh PDA plates [51]. Cultures were grown on PDA and incubated in an incubator at 25 • C for 2 weeks; then, morphological characters, including color, shape and size, were recorded. All fungal strains were stored in 10% sterilized glycerin at 4 • C for further studies. The studied specimens and cultures were deposited in the Herbarium of Jiangxi Agricultural University, Plant Pathology, Nanchang, China (HJAUP). The names of the new taxa were registered in Index Fungorum [2].
The final volume of the PCR reaction was 25 µL, comprising 1 µL of DNA template, 1 µL of each forward and reward primer, 12.5 µL of 2 × Power Taq PCR MasterMix and 9.5 µL of double-distilled water (ddH 2 O). The PCR thermal cycling conditions of ITS, SSU and LSU were initialized at 94 • C for 3 min, followed by 35 cycles of denaturation at 94 • C for 30 s, annealing at 55 • C for 50 s, elongation at 72 • C for 1 min and a final extension at 72 • C for 10 min before being maintained at 4 • C; the tef1-α were initialized at 95 • C for 3 min, followed by 35 cycles of denaturation at 95 • C for 30 s, annealing at 60 • C for 30 s, elongation at 72 • C for 1 min and a final extension at 72 • C for 10 min before being maintained at 4 • C. The PCR products were checked by 1% agarose gel electrophoresis staining with ethidium bromide. Purification and DNA sequencing were carried out at Beijing Tsingke Biotechnology Co., Ltd. China. New sequences generated in this study were deposited in the NCBI GenBank (www.ncbi.nlm.nih.gov, accessed on 10 December 2022; Tables 1 and 2). Table 1. List of Corynespora species and GenBank accessions used in the phylogenetic analyses. New sequences are indicated in bold.

Taxon
Strain Number GenBank Accession Numbers

Phylogenetic Analyses
The newly generated sequences, together with other sequences obtained from Gen-Bank (four loci: ITS, LSU, SSU and tef1-α (Table 1); three loci: ITS, LSU and SSU (Table 2)), were separately aligned using the MAFFTv.7 [55] online server (http://maffTh.cbrc.jp/ alignment/server/, accessed on 23 December 2022) and manually optimized when needed. Phylogenetic analyses were first conducted individually for each locus, then for a combined dataset of these loci. The four ITS, LSU, SSU and tef1-α alignment datasets and the three ITS, LSU and SSU alignment datasets were concatenated with Phylosuite software v1.2.2 [56], and absent sequence data in the alignments were treated with a question mark as missing data. Phylosuite software v1.2.2 [56] was used to construct separate phylogenetic trees based on ITS, LSU, SSU and tef1-α sequence data, as well as ITS, LSU and SSU sequence data. The concatenated and aligned datasets were analyzed separately using maximum likelihood (ML) and Bayesian inference (BI). The maximum-likelihood phylogenies were inferred using IQ-TREE [57] under an edge-linked partition model for 10000 ultrafast bootstraps [58]. For Corynespora, the final tree was selected among suboptimal trees from each run by comparing the likelihood scores using SYM+G4 for ITS, TNe+G4 for LSU+tef1-α and K2P+I for the SSU substitution model. Bayesian inference phylogenies were inferred using MrBayes 3.2.6 [59] under a partition model (2 parallel runs, 2000000 generations), in which the initial 25% of sampled data were discarded as burn-in. The best-fit model was SYM+G4 for ITS, GTR+F+G4 for LSU+tef1-α and K2P+I for SSU. For Kirschsteiniothelia, the final tree was selected among suboptimal trees from each run by comparing the likelihood scores using TIM2e+R3 for ITS+SSU and TN+F+G4 for the LSU substitution model. Bayesian inference phylogenies were inferred using MrBayes 3.2.6 [59] under a partition model (2 parallel runs, 2000000 generations), in which the initial 25% of sampled data were discarded as burn-in. The best-fit model was SYM+G4 for ITS+LSU+SSU. ModelFinder [60] was used to select the best-fit partition model (edge-linked) using the BIC criterion. The trees were viewed in FigTree v. 1.4.4 (http://tree.bio.ed.ac.uk/software/figtree, accessed on 10 December 2022) and further edited in Adobe Illustrator 2021.

Molecular Phylogeny
The phylogenetic tree ( Figure 1) inferred from maximum-likelihood and Bayesian inference analyses based on combined ITS, LSU, SSU and tef1-α sequence data consisted of three families (Corynesporascaceae, Periconiaceae and Cyclothyriellaceae). The concatenated sequence matrix comprised 23 sequences with 3147 total characters (the combined dataset, ITS: 1-498, LSU: 499-1348, SSU: 1349-2374, tef1-α: 2375-3147), 537 distinct patterns, 375 parsimonyinformative sites, 147 singleton sites and 2625 constant sites; Cyclothyriella rubronotata (TR) and C. rubronotata (TR9) were regarded as an outgroup. Maximum-likelihood and Bayesian inference analyses of the combined dataset resulted in phylogenetic reconstructions with largely similar topologies; the best-scoring ML tree is shown in Figure 1. Bootstrap support values for maximum likelihood higher than 75% and Bayesian posterior probabilities greater than 0.90 are shown above the nodes. The best-scoring ML consensus tree (lnL = -8859.832) with ultrafast bootstrap values from ML analyses and posterior probabilities from MrBayes analysis at the nodes is shown in Figure 1. The strains of Corynespora mengsongensis form a distinct clade sister to C. nabanheensis with good statistical support (ML/BI = 85/0.95); C. yunnanensis forms a high-support clade (ML/BI = 99/1.00) with the lineage consisting of C. mengsongensis and C. nabanheensis, and they form a sister clade to C. submersa (ML/BI =85/0.75).
The phylogenetic tree ( Figure 2) inferred from maximum-likelihood and Bayesian inference analyses based on combined ITS, LSU and SSU sequence data consisted of four orders (Acrospermales, Kirschsteiniotheliales, Monoblastiales and Strigulales). The concatenated sequence matrix comprised 36 sequences with 1260 total characters (combined dataset, ITS: 1-162, LSU: 163-471, SSU: 472-1260), 413 distinct patterns, 253 parsimony-informative sites, 183 singleton sites and 824 constant sites; Pseudorobillarda eucalypti (MFLUCC 12-0422) and P. phragmitis (CBS 398.61) were regarded as an outgroup. Maximum-likelihood and Bayesian inference analyses of the combined dataset resulted in phylogenetic reconstructions with largely similar topologies; the best-scoring ML tree is shown in Figure 2. Bootstrap support values for maximum likelihood higher than 75% and Bayesian posterior probabilities greater than 0.90 are shown above the nodes. The best-scoring ML consensus tree (lnL = -6307.741) with ultrafast bootstrap values from ML analyses and posterior probabilities from MrBayes analysis at the nodes is shown in Figure 2. The strains of K. nabanheensis form a separate clade closely related to K. thailandica, K. thujina, K. tectonae and K. rostrata, with strong statistical support (ML/BI = 95/0.98).
Culture characteristics: Colony on PDA reaching 80-88 mm diam. after 2 weeks in an incubator under dark conditions at 25 • C; irregular circular, velvety surface with dense, gray-white mycelia along the entire margin; reverse brown to dark brown.
Corynespora nabanheensis Jing W. Liu & Jian Ma, sp. nov., Figure 4. Index Fungorum number: IF900077 Etymology: The name refers to Nabanhe Nature Reserve, the locality where the fungus was collected.

Discussion
In this study, we collected saprophytic hyphomycetes on dead branches from terrestrial habitats in Yunnan Province, China. Based on the morphomolecular approach, four novel taxa are introduced: Corynespora mengsongensis sp. nov., C. nabanheensis sp. nov., C. yunnanensis sp. nov. and Kirschsteiniothelia nabanheensis sp. nov.
Corynespora show high morphological similarity to Corynesporina, Corynesporopsis, Hemicorynespora and Solicorynespora in terms of their terminal, monotretic, conidiogenous cells and differ only on the basis of single conidial characteristics (e.g., single or catenate, euseptate or distoseptate, basipetal chain or acropetal chain) [63]. The weak differentiation of these similar genera should only be maintained until sufficient molecular analysis allows for a more phylogenetic classification of genera. In addition, it is challenging to separate Corynespora from Helminthosporium based on morphology alone, as four Corynespora species, C. caespitosa, C. endiandrae, C. leucadendri and C. olivacea, were transferred to Helminthosporium based on molecular phylogenetic analyses, which led to the genus Helminthosporium also meeting the criteria of Corynespora [28].
The genus Corynespora produces conspicuous morphological features, and its generic type, C. cassiicola, is a ubiquitous species, mainly in tropical and subtropical areas, and has been recorded from a wide range of plants [64]. Most Corynespora species are known as saprobes and plant pathogens from woody and herbaceous hosts [8,9,13], but occasionally, C. cassiicola is also found in nematodes, sponges and human skin [64]. To date, 132 species of Corynespora (Table 3) have been be accepted worldwide, whereas four invalid names enclosed in quotation marks are also listed in Table 3. Many species have been identified only based on morphological studies, and only 13 species, including our three new species, have been subjected to molecular phylogenetic analyses. Morphological comparison is important for species identification, but the lack of a large amount of molecular data made it difficult to evaluate previously described Corynespora species by molecular methods. Thus, we recommend supplementary sequence data for previously described Corynespora species by re-examining their type materials or collecting fresh new specimens and using molecular phylogenetic analyses to evaluate their taxonomic placement as necessary.
Hawksworth [36] established the genus Kirschsteiniothelia and regarded K. aethiops as the type species. Boonmee et al. [37] treated the genus in a new family, Kirschsteiniotheliaceae, based on evidence from morphological and phylogenetic analyses. Hernandez-Restrepo et al. [49] raised Kirschsteiniotheliaceae to the new order Kirschsteiniotheliales in Dothideomycetes, although this order does not form a well-supported clade within Dothideomycetidae as a sister clade to Asterinales; the two orders diverged approximately 221 MYA according to divergence time estimates [65].
Sun et al. [38] accepted five former Dendryphiopsis species, D. arbuscula, D. binsarensis, D. biseptata, D. fascicularis and D. goaensis, in Kirschsteiniothelia following the latest treatment of Dendryphiopsis by Wijayawardene et al. [40]. However, these five species were invalidly introduced as new combinations in Kirschsteiniothelia on the basis of Art. F.5.1 (no identifier number cited) and Art. 41.1 (lacking a full and direct basionym reference) of the International Code of Nomenclature for Algae, Fungi, and Plants [33]. In addition, Sun et al. [38] provided a checklist for 35 Kirschsteiniothelia species including the distribution, habitat, host and morphology type of each species, but K. ebriosa [66] and K. vinigena [66] are not included. Subsequently, Verma et al. [62] described a new species, K. shimlaensis, from decaying stump in India.          All conidia are smooth, except where indicated; 2"-", the number of septation is not given.
Kirschsteiniothelia is one genus of many lignicolous fungi encountered in aquatic and terrestrial habitats. Following the treatment of Sun et al. [38], the genus currently consists of 39 species including K. nabanheensis [38,62,66], but most species have been identified based on morphological studies, and to date, only 17 species are represented by DNA sequences in GenBank. Kirschsteiniothelia has mainly been reported in the USA (nine species), China (eight species) and Thailand (six species), and little published information is has been recorded in other regions [38,62,66]. Thus, it is unclear whether is closely related with geographic regions.
Studies conducted to date on Corynespora and Kirschsteiniothelia have mainly focused on their alpha taxonomy, and most knowledge of both genera is related to woody and herbaceous hosts, whereas we have a less developed understanding of many natural substrates, such as dung, insects and other fungi, including lichens. Because most species of both genera lack cultures, some of them may have received scant consideration in singlespore isolation before the advent of Sanger sequencing and even have particular substrate requirements. Similarly, little attention has been accorded to the roles of these genera in decomposition and nutrient recycling, their geographical distribution, substrate specificities and teleomorph relationships. Therefore, it is not yet possible to quantify their roles in ecosystem function. Although this study broadens our understanding of the diversity of Corynespora and Kirschsteiniothelia taxa, additional large-scale surveys of fungal resources in aquatic and terrestrial habitats within different geographic regions and with different ecological environments, host information and climatic conditions are needed, which will contribute to a comprehensive knowledge of the fungal diversity of these genera. Further collaboration will also be necessary to quantify their functional roles and strengthen our ability to conserve fungal resources.
Author Contributions: J.L., Y.H., X.L., J.X., Z.X. and J.M. designed the study and were involved in the writing of the paper. J.L. and X.S. were responsible for sample collections. J.L. and L.Z. were involved in phylogenetic analyses. R.F.C.-R., R.C. and J.M. contributed to planning and editing of the paper. All authors have read and agreed to the published version of the manuscript.