﻿Morphology, phylogeny, mitogenomics and metagenomics reveal a new entomopathogenic fungus Ophiocordycepsnujiangensis (Hypocreales, Ophiocordycipitaceae) from Southwestern China

﻿Abstract Ophiocordyceps contains the largest number of Cordycepssensu lato, various species of which are of great medicinal value. In this study, a new entomopathogenic fungus, Ophiocordycepsnujiangensis, from Yunnan in southwestern China, was described using morphological, phylogenetic, and mitogenomic evidence, and its fungal community composition was identified. It was morphologically characterized by a solitary, woody, and dark brown stromata, smooth-walled and septate hyphae, solitary and gradually tapering conidiogenous cells with plenty of warty protrusions, and oval or fusiform conidia (6.4–11.2 × 3.7–6.4 µm) with mucinous sheath. The phylogenetic location of O.nujiangensis was determined based on the Bayesian inference (BI) and the maximum likelihood (ML) analyses by concatenating nrSSU, nrLSU, tef-1a, rpb1, and rpb2 datasets, and ten mitochondrial protein-coding genes (PCGs) datasets (atp6, atp9, cob, cox2, nad1, nad2, nad3, nad4, nad4L, and nad5). Phylogenetic analyses revealed that O.nujiangensis belonged to the Hirsutellasinensis subclade within the Hirsutella clade of Ophiocordyceps. And O.nujiangensis was phylogenetically clustered with O.karstii, O.liangshanensis, and O.sinensis. Simultaneously, five fungal phyla and 151 fungal genera were recognized in the analysis of the fungal community of O.nujiangensis. The fungal community composition differed from that of O.sinensis, and differences in the microbial community composition of closely related species might be appropriate as further evidence for taxonomy.


Introduction
The genus Ophiocordyceps was introduced by Petch (1931), with O. blattae Petch as the type. This genus accommodated species with features of head-cover asci, septate and non-disarticulating ascospores (Petch 1931). Then, the genus was regarded as a subgenus of Cordyceps (Kobayasi 1941(Kobayasi , 1982Mains 1958). Until 2007, Sung et al. (2007 erected a new family Ophiocordycipitaceae based on phylogenetic analysis and the characteristics of darkly pigmented stromata, which were pliant to wiry or fibrous to tough in texture. And they revised the classification of Ophiocordyceps, treating it as the type genus of Ophiocordycipitaceae. Ophiocordyceps has the largest number of species in Ophiocordycipitaceae, with 307 species named in Ophiocordyceps to date. (http:// www.indexfungorum.org/, retrieval on November 3, 2022).
The methods of morphology and phylogeny were utilized for species identification, and the phylogenetic analyses based on concatenating nrSSU, nrLSU, tef-1α, rpb1, and rpb2 datasets became the popular means (Sung et al. 2007;Quandt et al. 2014;Sanjuan et al. 2015;Wang et al. 2020a). Moreover, the mitochondrial genome had been an effective instrument for studying species' origin, classification, and evolution due to its advantages of high copy number, low mutation rate, and fast evolution rate (Alexeyev et al. 2013;Aguileta et al. 2014;Williams et al. 2014). The significant difference in the mitochondrial genome of fungi could be distinguished (Nie et al. 2019). The biogenetic analyses of the fungal mitochondrial genome could verify the genetically related species. NCBI has published the mitochondrial genomes of more than 680 fungi, including approximately 60 species of Hypocreales Zhao et al. 2021).
Some species in Ophiocordyceps have enormous medicinal and commercial value, such as O. sinensis, traditional in Chinese medicine. Owing to their extraordinary efficacy, wild sources were widely sold as commodities and gradually became scarce. (Han et al. 2019;Dai et al. 2020). Therefore, seeking additional new resources would defuse the tense situation. For example, O. lanpingensis and O. xuefengensis had been authenticated as possessing ingredients that were beneficial for health and considered to be desirable alternatives for O. sinensis (Zou et al. 2017;Zhang et al. 2017). Ophiocordyceps is widely distributed in China, and of particular note are some recent reports of new species from southwestern China (Wang et al. 2018;Wang et al. 2020b;Chen et al. 2021).
The companion fungi were essential for the growth and development of the host. For example, Tuber-associated microbial communities played a potentially important role in mycelial growth, ascocarp development, and mycorrhizal synthesis of Tuber (Li et al. 2018). And adding Grifola sp. in the cultivation process of G. umbellate could promote sclerotia formation (Guo et al. 2002). Thus, the composition and diversity of companion fungi should be analyzed to gain insight into new species and their microbial resources.
In this study, a new species of Ophiocordyceps, which parasitized on the larvae of Hepialidae, was collected from Yunnan in southwestern China. The phylogenetic location was elucidated based on the Bayesian inference (BI) and the maximum likelihood (ML) analyses by concatenating nrSSU, nrLSU, tef-1a, rpb1, and rpb2 datasets, and mitochondrial protein-coding genes (PCGs) datasets. Morphological characteristics were observed and recorded. The composition and diversity of the fungal communities hosting the new species were identified.

Sample collection and isolation
Samples were collected on Hepialidae larvae in the soil in Yajiaoluo (27°07'48"N, 98°52'12"E), Fugong County, Nujiang Prefecture, Yunnan Province, China. Specimens were photographed in the fields with a Canon 750D digital camera. The fresh specimens were placed into the sterile culture dish, then transferred to the laboratory and deposited in the Yunnan Herbal Herbarium (YHH), Yunnan University.
Specimens were isolated and cultured using the tissue isolating method (Yin and Zhang 2015;Wang et al. 2020b) as follows. Specimens were dipped into 75% alcohol for 2 min to sterilize the surface and then washed with sterile water. The 2-3 mm sclerotium was ripped by tweezers and put on the culture medium (200 g potato, 20 g dextrose 20, 15-20 g agar, 10 g yeast extract, 5 g peptone in 1 L sterile water) (Xu et al. 2019), with three replications. Then they were transferred to the room at 25 °C for culturing. The cultures were deposited in the Yunnan Fungal Culture Collection (YFCC), at Yunnan University.

Morphological observations
A moderate quantity of pure cultures was picked by an inoculating needle onto the center of the culture medium and maintained at 25 °C. After 6-10 weeks, shape, size, texture, and color were photographed with a Canon 750D camera. The superficial pure cultures were lightly stuck on transparent adhesive tapes, then the tapes were patched on slides, and the slides were placed on the Olympus BX53 microscope for micro-morphological observations and measurements (Wang et al. 2020a;Wang et al. 2020b).

DNA extraction, PCR amplification, and sequencing of nuclear genes
The genomic DNA of the samples (containing specimens and pure cultures) was isolated using the ZR Fungal DNA kit (Zymo, California, USA), then the DNA extract was checked on 1% agarose gel, and DNA concentration and purity were determined with NanoDrop ND-2000 spectrophotometer (Thermo Scientific, Wilmington, USA). The nrSSU and nrLSU (nuclear ribosomal small and large subunits), rpb1 and rpb2 (the largest and second-largest subunit sequences of RNA polymerase II), and tef-1α (the translation elongation factor 1α) regions were amplified with the primer pairs used by Wang et al. (2020b). The PCR mixtures contained 2 × Taq PCR Master Mix (Tiangen, Beijing, China) 25 µL, forward primer (10 µM) 0.5 µL, reverse primer (10 µM) 0.5 µL, template DNA (1 ng/µL) 1 µL, and finally added sterile ddH 2 O up to 50 µL. Finally, the PCR amplification and sequencing were performed as described by Wang et al. (2015).

Sequencing, assembly, and annotation of mitogenome
The genomic DNA of the pure cultures was isolated through the above-mentioned method, the extracted DNA was transported to BGI genomics Co., Ltd (Wuhan, China) for sequencing. The sequencing library was built by the IlluminaTruseq DNA Sample Preparation Kit (BGI, Shenzhen, China), and the Illumina HiSeq 4000 Platform was applied to the PE2 × 150 bp sequencing. After data quality control, the unpaired, short, and low-quality reads were removed, and the clean reads were obtained (Zhao et al. 2021). Next, the reads of the mitogenome were collected from the clean data employing GetOrganelle v.1.6.2e, and the mitogenome was assembled using BLAST 2.2.30 and SPAdes. V.3.13.0. The mitogenome was initially annotated by MFannot (https://megasun.bch.umontreal.ca/RNAweasel/, accessed on 10 December 2020) and MITOS (http://mitos2.bioinf.uni-leipzig.de/index.py, accessed on 10 December 2020) (Valach et al. 2014;Jin et al. 2020;Chen et al. 2021).
Purified amplicons were pooled in equimolar amounts and paired-end sequenced on an Illumina MiSeq PE300 platform (Illumina, San Diego, USA), following the standard protocols by Majorbio Bio-Pharm Technology Co. Ltd. (Shanghai, China). The raw reads were deposited into the NCBI Sequence Read Archive (SRA) database (Sequence Read Archive (SRA) Accession Number: SAMN28950406-SAMN28950409).
Raw FASTQ files were de-multiplexed using an in-house Perl script, and then quality-filtered by fastp version 0.19.6 ) and merged by FLASH version 1.2.7 (Magoč and Salzberg 2011). Then the optimized sequences were clustered into operational taxonomic units (OTUs) employing UPARSE 7.1 (Edgar 2013) with the 97% sequence similarity level. Chimeric sequences, chloroplast sequences, mitochondrial sequences, and the OTUs identified as Plantae, Rhizaria, Chromista, and those with no rank and unclassified kingdom were removed from samples.

Composition and phylogenetic analysis of microbial communities
Bioinformatic analysis was carried out by the Majorbio Cloud platform (https://cloud. majorbio.com). The taxonomy of each OTU representative sequence was analyzed by RDP Classifier version 2.2 (Wang et al. 2007) against the ITS gene database (Unite V7.2) through a confidence threshold of 0.7. A phylogenetic tree was constructed to illustrate the relationships between the fungi at the family level, employing FastTree version 2.1.3 (http://www.microbesonline.org/fasttree/) and the ML algorithm .

0.02
Fusarium fujikuroi JX910420 Notes. Ophiocordyceps nujiangensis was closely phylogenetically related to O. karstii and O. liangshanensis. The formation of stromata on the head of the host was a feature common to all three species. However, the length of the stromata varies between the three species. O. nujiangensis had a stromata length longer than O. karstii, but shorter than O. liangshanensis (Table 1). O. nujiangensis, on the other hand, had slightly longer conidiophores and slightly smaller conidia than O. liangshanensis (Table 1). In total, 135,048 effective sequences were obtained. Based on the minimum number of reads in the sample, 33,762 reads were randomly selected for each sample to avoid bias in the sequencing depth. The rarefaction curve (the Shannon-Wiener curve) showed that the sequencing depth was very reasonable for representing the diversity of the fungal community (Suppl. material 3). At the phylum level, a total of five phyla were identified, including Ascomycota, Basidiomycota, Mortierellomycota, Rozellomycota, and Glomeromycota. Of these, Ascomycota was dominant, with an average of 99.66%. The rest averaged no more than 1 percent. And the unclassified was dominant in the 151 identified genera, the average proportion was 29.56%, followed by Trichothecium (27.16%) and Microdochium (26.81%) (Fig. 4). Namely, numerous companion fungi were verified in the fruiting body of O. nujiangensis. The results also indirectly suggested that O. nujiangensis might be a new species as its ITS sequence could not be aligned in the database.

Phylogenetic analyses of the fungi at the family level
The top 50 families were classified into four phyla (Suppl. material 4), comprising Ascomycota, Basidiomycota, Mortierellomycota, and Rozellomycota; however, none in Glomeromycota. There were 41 families subordinated to Ascomycota, including the three families (Clavicipitaceae, Ophiocordycipitaceae, and Cordycipitaceae), which distributed Cordyceps sensu lato. And the phylogenetic locations of the three families  Liang et al. (2007) were essentially the same as previously reported in the study by Sung et al. (2007) and Wang et al. (2020a). The results implied that O. nujiangensis might have many companion fungi, which belongs to Cordyceps sensu lato.

Discussion
Ophiocordyceps nujiangensis was morphologically characterized by solitary, woody, and dark brown stromata, smooth-walled and septate hyphae, solitary and gradually  A total of five fungal phyla and 151 fungal genera were identified in this study. Among them, Ascomycota and the unclassified were the dominant phylum and genus. Except for the dominant, Trichothecium and Microdochium also had high proportions at the genus level. The genus, Trichothecium, was a heterogonous group of filamentous fungi; some species were pathogenic fungi (Summerbell et al. 2011;Han et al. 2021). Microdochium was a common cereal pathogen fungus that adapted nicely to the cool (Parry et al. 1995;Gagkaeva et al. 2020). Some companion fungi had been confirmed that had vital functions (Guo et al. 2002;Li et al. 2018). The growth and development of the host were mostly due to the combined effect of the microbial adding peptone and yeast community (Han et al. 2019;Xie et al. 2021). Thus, the genera might have had an essential influence on the growth and development of O. nujiangensis. Furthermore, a comparison of the fungal communities of O. sinensis and O. nujiangensis showed that they had different community compositions. However, Trichothecium and Microdochium could not be found among the top 19 genera in fungal communities of O. sinensis reported (Xia et al. 2016). Consequently, the differences in the microbial community composition of closely related species might be suitable as further evidence for identifying species.
The phylogenetic analysis of mitochondrial genes became an adequate means to delimit fungal species, except for morphological observation and the five-gene phylogenetic tree (Nie et al. 2019;Meng et al. 2020). Similar topologies were obtained by utilizing 14 PCGs, PCGs + rRNA, or mitochondrial whole genomes (Hu et al. 2021). It was illustrated that the stable phylogenetic trees could be reconstructed using the phylogenetic analysis of mitochondrial genes. In the present research, the phylogenetic tree of Hypocreales was rebuilt, which was similar to the report by Chen et al. (2021). It had been shown that the phylogenetic trees with mitochondrial genes were reliable.
The characteristic differences between the new species and other species could be distinguished through the morphology data, and the phylogenetic location of the new species could be determined by the phylogeny and mitogenomics data. It was attempted to further study the companion fungi of the new species, but the available data on the species and their phylogenetic relationship were considerably lacking. Metagenomics provided more comprehensive genetic information about microorganisms and the microorganisms with which they associated (Venter et al. 2004;Truong et al. 2017;Huang and Wang 2020). Therefore, the method might be an efficient avenue for reconstructing the "Tree of Life".