Molecular Phylogenetic and Comparative Genomic Analysis of Pleurocordyceps fusiformispora sp. nov. and Perennicordyceps elaphomyceticola in the Family Polycephalomycetaceae

Several Pleurocordyceps species have been reported as hyperparasitic fungi. A new species, Pleurocordyceps fusiformispora, and a known species, Perennicordyceps elaphomyceticola, are described here based on morphology and phylogenetic evidence from six genes (ITS, SSU, LSU, TET1-α, RPB1, and RPB2). Pl. fusiformispora differed from the other Pleurocordyceps species by producing flaky colonies, ovoid or elliptic α-conidia, and fusiform or long fusiform β-conidia. Both full genomes of Pe. elaphomyceticola and Pl. fusiformispora were sequenced, annotated, and compared. The antiSMASH and local BLAST analyses revealed significant differences in the number and types of putative secondary metabolite biosynthetic gene clusters, i.e., NPPS, PKS, and hybrid PKS–NRPS domains, between the two species. In addition, the putative BGCs of six compounds, namely ε-poly lysine, 4-epi-15-epi-brefeldin A, Monorden D/monocillin IV/monocillin VII/pochonin M/monocillin V/monocillin II, Tolypyridone, Piperazine, and Triticone DABFC, were excavated in the present study. This study motivates the use of heterologous expression and gene knockout methods to discover novel biologically active SMs from Polycephalomycetaceae.


Introduction
The classification system of Cordyceps sensu lato is widely accepted, with Perennicordyceps and Pleurocordyceps being sister genera within the family Polycephalomycetaceae [1].Polycephalomycetaceae is a kind of fungus group with important application value and broad development and application prospects in medicine, agriculture, and other fields [1].Xiao et al. [2] noted that Pleurocordyceps sp.exhibits significant potential for the production of a diverse range of secondary metabolites.Recent studies showed that Pleurocordyceps nipponicus has antibacterial [3,4] and antitumor activities [5].Surapong et al. [6] isolated two compounds, Cordytropolone and Leptosphaerone A, from liquid cultures of Pl. nipponicus, and experimental studies showed that Cordytropolone has some antifungal activity.
Natural products, also known as secondary metabolites (SMs), have played an important role in the history of drug discovery and development [7].As more and more microbial genomes are sequenced, bioinformatic analysis has revealed a vast resource of novel SMs.Genome mining is a new strategy of SM discovery based on gene cluster sequences and biosynthetic pathways.At the same time, it can directly associate the structures of SMs with synthetic pathways and facilitate the study of SM biosynthesis and combinatorial biosynthesis [8].Some analysis tools and procedures (e.g., antibiotics and Secondary Metabolites Analysis Shell, antiSMASH; ClusterFinder; Antibiotic Resistant Target Seeker, ARTS) allow for the rapid and direct detection of biosynthetic gene clusters (BGCs), as well as the diversity of gene cluster families, such as polyketide synthases (PKSs), non-ribosomal peptide synthetases (NRPSs), and terpene synthases (TSs) [9].More SM BGCs can be discovered through gene mining [10].For instance, the sequence of amino acids of NpPKS3 in the lichenized fungus Nephrmopsis pallescens was 53% consistent with the type III PKS (CYSBs) of Beauveria bassiana [11].Sayari et al. [12] found that all 20 genomes of Ceratocysti daceae showed extremely conserved PKS-III gene clusters containing homologous genes encoding the CHS.Wang et al. [9] discovered beauveriolide BGC in the Cordyceps militaris genome using bioinformatic analysis and then produced the compound via heterologous expression.Fumosorinone is a new 2-pyridone alkaloid isolated from Cordyceps fumosorosea [10].The BGC of fumosorinone consisted of a hybrid PKS-NRPS, two cytochrome P450, a transenoyl reductase gene, and two other transcription regulatory genes [13].In recent years, there has been little work on secondary metabolite gene clusters in Perennicordyceps and Pleurocordyceps.
In this study, the species Pl. fusiformispora Hong Yu bis and Z.H. Liu, D.X.Tang, Y.L. Lu, sp.nov.was first introduced.In order to discover more potential gene clusters of SMs, the whole genomes of Pl. fusiformispora and Pe.elaphomyceticola were sequenced and annotated and were used in gene mining studies.The potential of the Polycephalomycetaceae fungi to produce SMs was further analyzed.

Culture and Morphological Observations
Sexual morph observation was performed by photographing and measuring ascomata using an Olympus SZ61 stereomicroscope (Olympus Corporation, Tokyo, Japan).Freehand or frozen sections of the fruit body structure were placed in a solution of lactophenol cotton blue for microscopic study and photomicrography.The frozen sections were prepared using a freezing Microtome HM525NX (Thermo Fisher Scientific, Waltham, MA, USA).The micro-morphological characteristics of fungi (perithecia, asci, apical caps, and ascospores) were examined using Olympus CX40 and BX53 microscopes.The PDA solid media (20 g/L potato powder, 20 g/L glucose, 18 g/L agar powder, 1 L H 2 O (all chemicals and reagents were from Kunming City, China)) in which the colonies were cultivated were kept at room temperature (25 • C) for 40 days, and then the specimens were photographed and recorded using a Canon 750D camera (Canon Inc., Tokyo City, Japan) to observe the morphological properties of the colonies.The necessary microscope slide cultures were made according to [14], leaving them at 25 • C for 10 days.The colony was photographed and measured every fourth day.To characterize the strain, microscope slide cultures were generated by inoculating a small portion of the mycelium onto a 25 mm 2 area of PDA medium (20 g/L potato powder, 20 g/L glucose, 18 g/L agar powder, 1 L H 2 O (all chemicals and reagents were from Kunming City, China)) block overlaid by a cover slip.Specimens were photographed and measured with an Olympus SZ61 stereomicroscope (Hamburg, Germany).Morphological observations and measurements were carried out using an Olympus CX40 microscope and a FEI QUANTA200 scanning electron microscope (Valley City, ND, USA).Fifty measurements were taken to collect the necessary information about the hypha, synnema, conidial mass, phialide, and conidium.

Genome Sequencing and Assembly
Pe. elaphomyceticola and Pl.fusiformispora strains were cultured on PPDA solid medium (20 g/L potato powder, 10 g/L yeast powder, 20 g/L glucose, 18 g/L agar powder, 1 L H 2 O) at 25 • C for 60 days.The mycelium was transferred to PPA liquid medium (20 g/L potato powder, 10 g/L yeast powder, 20 g/L glucose, 1 L H 2 O) at 25 • C static cultivation for 2-3 months.Appropriate amounts of Pe. elaphomyceticola and Pl.fusiformispora mycelium were scraped, and total genomic DNA was extracted using the Plant DNA Isolation Kit (Foregene Co., Ltd., Chengdu, China), and then sequenced on an automatic sequence analyzer (BGI Co., Ltd., Wuhan, China) using the same primers as used in amplification.Sequencing data contain some low-quality reads with joints, which can cause significant interference in subsequent information analysis.Illumina NovaSeq 2000 (Nanopore, Wuhan, China) high-throughput sequencing platform, used for sequencing a gene library with 400 bp insertion fragment, was used with a sequencing mode of paired-end and 2 × 150 bp.The fastp ([https://github.com/OpenGene/fastp],accessed on 14 July 2023) was used to filter the raw reads, discard low-quality reads, and obtain clean data.In order to ensure the quality of subsequent information analysis, it was necessary to further filter the raw data to generate high-quality sequences.The standards for data filtering mainly include the following points: joint contamination removal, using AdapterRemoval (v2.0) [22] to remove joint contamination from the 3 'end, and quality correction, using SOAPec (v2.0) software to perform quality correction on all reads based on Kmer frequency, with a Kmer setting of 17 used for correction.A5-MiSeq and SPAdes were used to construct contig and scaffold.Finally, the assembly effects of contigs and scaffolds were evaluated using pilon v1.18 [23] software.

Gene Prediction and Annotation
A combination of de novo gene prediction, transcript mapping, and homologous searches were used for gene prediction.Based on the existing database on gene function and metabolic pathways, the predicted genes were annotated by BLAST search, including Kyoto Encyclopedia of Genes and Genomes (KEGG), NCBI non-redundant protein sequences (NR), Gene Ontology (GO), Cluster of Orthologous Groups of eukaryotic complete genomes (KOG), Pfam, and Interpro.

Analysis of Secondary Metabolite Biosynthesis Gene Cluster
The antiSMASH (https://antismash.secondarymetabolites.org/,accessed on 1 September 2023) online program was used to perform gene cluster prediction at the level of genomic scaffolds of Pe. elaphomyceticola and Pl.fusiformispora complex.Based on antiSMASHdetected scaffolds with gene clusters, the online program FGENESH (www.softberry.com/,accessed on 26 September 2023) was used to predict gene structures using Ophiocordyceps sinensis as a parameter.To obtain the domain, the PKS/NRPS online program (https://nrps.igs.umaryland.edu/,accessed on 11 October 2023) was used to predict gene clusters in contigs where NRPS/PKS genes were located.At the same time, the online program Protein BLAST (https://blast.ncbi.nlm.nih.gov/,accessed on 25 October 2023) was used for NRPS/PKS genes of contig protein ratio analysis.

Cluster Analysis
The polygene nucleotide sequences (ITS, SSU, LSU, TEF-1a, RPB1, RPB2) were downloaded from NCBI (https://www.ncbi.nlm.nih.gov/,accessed on 25 August 2023) and compared with these sequences using the Clustal W program in the MEGA5.0software for multi-sequence comparison [24,25].Based on a six-gene dataset, the software PhyloSuite (v1.2.2 Win) was used to construct phylogenetic tree of maximum likelihood (ML) and Bayesian inference (BI).In addition, we used the method from the online program IQ-TREE web server (http://iqtree.cibiv.univie.ac.at/, accessed on 15 July 2023) since the report 1000 and the rest of the default parameter to construct an ML cluster analysis tree of NRPS or hybrid PKS-NRPS proteins of Pe. elaphomyceticola, Pl. fusiformispora, and other fungi.

Phylogenetic Analysis
Based on the joint matrix of nucleotide sequences of ITS, SSU, LSU, TEF1-α, RPB1, and RPB2, the molecular phylogenetic tree of Polycephalomycetaceae was reconstructed by ML and BI.The total length of the concatenated dataset of six genes across the 86 samples was 5303 bp, including 554 bp for ITS, 920 bp for SSU, 911 bp for LSU, 975 bp for TEF-1α, 739 bp for RPB1, and 1,204 bp for RPB2.The phylogenetic tree, represented by Tolypocladium ophioglossoides NBRC 106330 and T. ophioglossoides NBRC 100998 as the outgroup taxa, consisted of three clades, i.e., Polycephalomyces clade, Perennicordyceps clade, and Pleurocordyceps clade, with a total of 86 sequences (Table 1), of which 4 were self-detected.The matrix had 2646 distinct patterns, 1669 parsimony-informative sites, 666 singleton sites, and 13,982 constant sites.ModelFinder was used to select the best-fitting likelihood model (GTR+F+I+G4) for ML analyses and BI analyses according to the Akaike information criterion (AIC).The ML and BI tree with the best score found was −40,864.213, and the total tree length was 1.098.The generic-level relationships of ML and BI trees were topologically similar.The tree was visualized with its maximum likelihood bootstrap proportions (ML-BS) and Bayesian posterior probability (BI-BPP) in Figtree v.1.4.3 (Figure 1).
From the perspective of a six-gene phylogenetic tree, the three branches of the Polycephalomycetaceae receive high support rates.The cluster of Pl. fusiformispora collected and described in this study was in the adjacent branches of Pl. sinensis.Pl. fusiformispora and Pl.sinensis were sister species to each other and formed a separate clade (BS = 96%, BPP = 0.89).Pe. elaphomyceticola and Pe.prolifica were sister species to each other and formed a separate clade (BS = 99%, BPP = 1).This was consistent with the results [1].Etymology: The species name refers to the production of fusiform or long fusiform conidia during the asexual phase.Hyperparasitic on Ophiocordyceps sp.(Ophiocordycipitaceae) and on insects buried in soil.Sexual morph: undetermined.Asexual morph: Synnemata measure 1.2-1.5 cm long by 0.1-0.5 mm wide, clavate, capitate, crowd on the insect body and Ophiocordyceps sp., unbranched, white to yellowish, with or without fertile head at the apex.Stipes 2.1-2.8cm long, 0.7-1.0mm wide.Colonies on PDA grow slowly, attaining a diameter of 5.1-5.5 cm in 40 days at 25 • C, clustered, white-yellow, and reverse dry yellow.Synnemata emerge after 20 days, flaky, branched, 1.4-1.9cm long and 0.3-0.5 cm wide, showing radiating distributions.Phialides exist in two types: αand β-phialides.Both types of phialide often produce new phialides at their own apices or yield catenulate β-conidia, collarettes not flared, periclinal thickening not visible.The α-phialides are acropleurogenous on conidiophores and solitary on hyphae, narrow lageniform or subulate, taper abruptly from the base to the apex, 8.42-20.9µm long, 1.3-2.9µm wide at the base, and 0.6-1.6 µm wide at the apex.The β-phialides are solitary on hyphae, lanceolate, taper gradually from the base to the apex, 8.7-14.8µm long, 2.5-3.1 µm wide at the base, and 0.9-1.1 µm wide at the apex.α-conidia are ovoid or elliptic and occur in the conidial mass on the agar or on the final portion of synnema, 2.9-4.8 × 1.3-3.1 µm.β-conidia are fusiform or long fusiform and are produced on the surface mycelium of colony, usually in chains on a phialide, 2.8-4.4 × 1.7-2.8µm.

Genome Sequencing and Assembly
The Illumina sequencing produced 24,006,990 raw reads and 3,600,527,330 high-quality reads of Pl. fusiformispora (Table S1).The size of the genome Pe. elaphomyceticola was 31.51Mb, containing 48.79% GC content and 7794 protein-coding genes along with 103 tRNA, 18 rRNA, and 52 ncRNA.Among the genome size of Pl. fusiformispora was 33.02 Mb, containing GC content, protein-coding genes, tRNA, rRNA, and ncRNA, which were

Genome Annotation
The Pe. elaphomyceticola species were identified using the KEGG databases (3509 genes/ 45.02%), the EggNOG databases (7196 genes/92.32%),the NR databases (7536 genes/96.68%),the GO database (5302 genes/68.02%),and the Pfam databases (5927 genes/76.04%).However, an analysis of 6706 non-redundant Pl. fusiformispora genes in a publicly available protein sequence database produced mixed results; these were from the KEGG databases (3723 genes/42.28%),the EggNOG databases (8183 genes/92.94%),the NR databases (8566 genes/97.29%),the GO database (5112 genes/69.50%),and the Pfam databases (6726 genes/76.39%)(Table S1).The KEGG functional classification of Pe. elaphomyceticola and Pl.fusiformispora indicated protein families for genetic information processing, signaling and cellular processing, and signal transduction (respectively, Figure 5b,e).In addition, the rich genetic information and diversity of signaling proteins may facilitate more efficient information exchange and secondary metabolism.According to the EggNOG database, Pl. fusiformispora and Pe.elaphomyceticola captured more predicted genes as "Function unknown", "Posttranslational modification, protein turnover, chaperones", and "Carbohydrate transport and metabolism" (Figure 5a,d).However, a wide variety of carbohydrate metabolism and post-translational events suggested that they enhanced the bioactivity and energy conversion efficiency of regulatory proteins.GO annotation of the genomes of both species showed that translation, protein transport, and glucose metabolism were abundant genes in biological processes (Figure 5c,f).Furthermore, the GO annotation revealed integral components of the membrane, nucleus, and cytoplasm from the cellular component and ATP binding, metal ion binding, and zinc ion binding from the molecular function.The two species of Pe. elaphomyceticola and Pl.fusiformispora were wild strain, in which many metabolic genes might be involved in signal transduction.

Additional Annotation Carbohydrate-Active Enzymes (CAZy)
Enzymes that played an important role in carbohydrate modification, biosynthesis, and the degradation of fungi were carbohydrate-active enzymes (CAZy) [36], which are present in a database of carbohydrate-active enzymes and a special database of carbohydrate enzymes [37].It was shown that Pe. elaphomyceticola (Figure 6a) and Pl.fusiformispora (Figure 6b) had many glycoside hydrolases (GHs), glycosyl transferases (GTs), and auxiliary activities (AAs), leading us to speculate that Pe. elaphomyceticola and Pl.fusiformispora possibly breakdown complex carbohydrates and capture more energy.

Additional Annotation
Carbohydrate-Active Enzymes (CAZy) Enzymes that played an important role in carbohydrate modification, biosynthesis, and the degradation of fungi were carbohydrate-active enzymes (CAZy) [36], which are present in a database of carbohydrate-active enzymes and a special database of carbohydrate enzymes [37].It was shown that Pe. elaphomyceticola (Figure 6a) and Pl.fusiformispora (Figure 6b) had many glycoside hydrolases (GHs), glycosyl transferases (GTs), and auxiliary activities (AAs), leading us to speculate that Pe. elaphomyceticola and Pl.fusiformispora possibly breakdown complex carbohydrates and capture more energy.

Pathogen-Host Interactions (PHI)
The full name of PHI is Pathogen Host Interactions Database, which is a database of pathogen-host interactions, mainly derived from fungi, oomycetes, and bacterial pathogens.The infected hosts include animals, plants, fungi, and insects [38].This database plays an important role in searching for target genes for drug intervention, and it also includes antifungal compounds and corresponding target genes.The results showed that Pe. elaphomyceticola (Figure 7a) and Pl.fusiformispora (Figure 7b) had reduced virulence and the loss of pathogenicity genes.The primary annotated genes of Pe. elaphomyceticola and Pl.fusiformispora from the PHI database were used to reduce virulence and do not affect pathogenicity, indicating that Pe. elaphomyceticola and Pl.fusiformispora are not highly pathogenic strains.

Pathogen-Host Interactions (PHI)
The full name of PHI is Pathogen Host Interactions Database, which is a database of pathogen-host interactions, mainly derived from fungi, oomycetes, and bacterial pathogens.The infected hosts include animals, plants, fungi, and insects [38].This database plays an important role in searching for target genes for drug intervention, and it also includes antifungal compounds and corresponding target genes.The results showed that Pe. elaphomyceticola (Figure 7a) and Pl.fusiformispora (Figure 7b) had reduced virulence and the loss of pathogenicity genes.The primary annotated genes of Pe. elaphomyceticola and Pl.fusiformispora from the PHI database were used to reduce virulence and do not affect pathogenicity, indicating that Pe. elaphomyceticola and Pl.fusiformispora are not highly pathogenic strains.
includes antifungal compounds and corresponding target genes.The results showed that Pe. elaphomyceticola (Figure 7a) and Pl.fusiformispora (Figure 7b) had reduced virulence and the loss of pathogenicity genes.The primary annotated genes of Pe. elaphomyceticola and Pl.fusiformispora from the PHI database were used to reduce virulence and do not affect pathogenicity, indicating that Pe. elaphomyceticola and Pl.fusiformispora are not highly pathogenic strains.

Analysis of Secondary Metabolite Biosynthesis Gene Cluster
Overview of Twelve Genomic BGCs of Pe. elaphomyceticola and Pl.fusiformispora AntiSMASH and local BLAST analyses revealed that Pe. elaphomyceticola (39) and Pl.fusiformispora (50) had different putative SM BGCs (Table S3).Pl. fusiformispora and Pe.elaphomyceticola predicted that the largest number of SM BGC types were NRPS, with 16 and 15, respectively.Pl. fusiformispora had five hybrids, PKS + NRPS, and Pe.elaphomyceticola had two.Both Pl. fusiformispora and Pe.elaphomyceticola had five terpenes and two hybrids, NRPS + other genes.Pl. fusiformispora and Pe.elaphomyceticola had seven and six other genes, respectively.For the predicted PKSs, Pl. fusiformispora had fifteen PKSs, while Pe.elaphomyceticola had nine.The Pl. fusiformispora genome had fifteen PKSs, including ten HR-PKSs, three non-reducing (NR) PKSs, and two partially reducing (PR) PKSs.The genome of Pe. elaphomyceticola had nine PKSs, including four NR-PKSs, four HR-PKSs, and one PR-NRPS.These results indicated that the number and type of SM BGCs obtained in different species vary.
Several BGCs of Pe. elaphomyceticola and Pl.fusiformispora were 100% similar to MIBiG sequences.Purev et al. [39] found that ε-poly-lysine had antifungal activity.The predicted Pe. elaphomyceticola Region 13.1 and Pl.fusiformispora Region 21.4 were responsible for ε-poly-lysine (Figure 8).The local BLAST comparison demonstrated that ε-poly-lysinehas requires enzymes with the A-P-T structural domain, PKc-like super family, GAT_1 super family, and MFS.Both species have the potential to synthesize ε-polylysinehas.Zhang et al. [40] studied genome mining of in-house strains using hybrid PKS-NRPS as a query and identified an endophyte Tolypocladium sp.49Y, which possesses a potential 4-hydroxy pyridone biosynthetic gene cluster.Moreover, heterologous expression in Aspergillus oryzae NSAR1 revealed that this gene cluster was functional and able to produce a rare type of 4-hydroxy pyridones called Tolypyridone.The predicted Region 19.2 and Region 19.3 of Pl. fusiformispora might be responsible for Tolypyridone biosynthesis (Figure 9).Qin et al. [41] proved that Monorden D/monocillin IV/monocillin VII/pochonin M/monocillin V/monocillin II compounds have modest antibacterial activities.By comparison, Pl. fusiformispora Region 21.2 and Region 21.3 had highly homologous regions, being highly similar to the synthetic Monorden D/monocillin IV/monocillin VII/pochonin M/monocillin V/monocillin II gene cluster (Figure 10).These results indicated that the catalytic synthesis of NR-PKS (SAT-KS-AT-PT-ACP-Te), HR-PKS (KS-AT-DH-ER-KR-ACP), FixC super family, MFS_Tpol_MDR_like, and CYP60B-like were necessary genes.The results showed that (Figure 11) Pe. elaphomyceticola Region 61.1 was like the gene cluster DS231617.1 that catalyzed the synthesis of Triticone ABFC.However, Pl. fusiformispora did not have the cluster of Triticone ABFC.Pe. elaphomyceticola Region 9.1 was highly similar to the known gene cluster that catalyzed the synthesis of piperazine (Figure 12).The slight difference was that one gene in Pe. elaphomyceticola Region 9.1 had dehydrogenase and CYP503A1-like functions.Moreover, KJ728786.1 synthesized 4-epi-15-epi-brefeldin A (Figure 13).The core gene structural domain (KS-AT-ER-KR-ACP) in Pl. fusiformispora Region 68.1 was similar to KJ728786.1.However, Pl. fusiformispora lost some of its modified genes during evolution.It was speculated that the position direction of key enzymes and modified genes was different when different genera were in the compound.At the same time, there will be gene fusion and gene loss in the process of evolution.core gene structural domain (KS-AT-ER-KR-ACP) in Pl. fusiformispora Region 68.1 was similar to KJ728786.1.However, Pl. fusiformispora lost some of its modified genes during evolution.It was speculated that the position direction of key enzymes and modified genes was different when different genera were in the compound.At the same time, there will be gene fusion and gene loss in the process of evolution.core gene structural domain (KS-AT-ER-KR-ACP) in Pl. fusiformispora Region 68.1 was similar to KJ728786.1.However, Pl. fusiformispora lost some of its modified genes during evolution.It was speculated that the position direction of key enzymes and modified genes was different when different genera were in the compound.At the same time, there will be gene fusion and gene loss in the process of evolution.

Cluster Analysis
The protein clustering results of NRPS and hybrid PKS-NRPS from Pe. elaphomyceticola and Pl.fusiformispora were compared to other fungal NRPS and hybrid PKS-NRPS.It was showed that Pe. elaphomyceticola Region 13.1 and Pl.fusiformispora Region 21.4 aggregated with Epichloe festucae (BBU42014.1),with being catalyzed the biosynthesis of ε-polylysine or its analogs (Figure 14

Cluster Analysis
The protein clustering results of NRPS and hybrid PKS-NRPS from Pe. elaphomyceticola and Pl.fusiformispora were compared to other fungal NRPS and hybrid PKS-NRPS.It was showed that Pe. elaphomyceticola Region 13.1 and Pl.fusiformispora Region 21.4 aggregated with Epichloe festucae (BBU42014.1),with being catalyzed the biosynthesis of ε-polylysine or its analogs (Figure 14

Synteny Analysis
The scaffolds containing the SM BGC in the genomes of Pe. elaphomyceticola (39) and Pl.fusiformispora (50) were subjected to synteny analysis.The scaffolds where the SM BGC are located are divided into more than 10 collinear blocks, and there may be rearrangement (Figure 16).From bottom to top, they are Pl.fusiformispora and Pe.elaphomyceticola, respectively.

Synteny Analysis
The scaffolds containing the SM BGC in the genomes of Pe. elaphomyceticola (39) and Pl.fusiformispora (50) were subjected to synteny analysis.The scaffolds where the SM BGC are located are divided into more than 10 collinear blocks, and there may be rearrangement (Figure 16).From bottom to top, they are Pl.fusiformispora and Pe.elaphomyceticola, respectively.

Discussion
Herein, we describe a new species of Pl. fusiformispora using a combination of morphology and phylogeny.The newly established species distinctly form independent clades in the phylogenetic tree (Figure 1).Morphologically, the new species Pl. fusiformispora is similar to Pl. aurantiacus, Pl. agarica, Pl. heilongtanensis, Pl.lanceolatus, Pl. marginaliradians, Pl. nutansis, Pl. vitellina, and Pl.yunnanensis in that they have two types of phialides and conidia.However, the species Pl. fusiformispora has colonies that are flaky, α-conidia that are ovoid or elliptic, and β-conidia that are fusiform or long fusiform and differ from those of other species of Pleurocordyceps [1,24,25,42].The discovery of the new species of Pl. fusiformispora adds to the diversity of the genus Pleurocordyceps.
Xiao et al. [1] noted five new samples (Pl.sinensis GACP 19-2301, Pl. sinensis MFLU 21-0269, Pl. sinensis MFLU 21-0268, Pl. sinensis GACP 20-0865, and Pl.sinensis GACP 20-2304), each of which is parasitic on a different host clustered with Pl. sinensis.These hosts were clavicipitoid fungi and constituted new hosts for Pl.sinensis.In addition, Xiao et al. [1] introduced Pl. nutansis as a new species under Pleurocordyceps and identified the host as Ophiocordyceps nutans.However, in the phylogenetic tree, there was a disorder in the systematic position of these two different hosts, which might be due to genetic differences caused by different strains.
Some species of Polycephalomycetaceae have also been reported as hyperparasitic fungi; these species were Cordyceps, Elaphomyces, Hirsutella, Myxomycetes, and Ophiocordyceps [24,25,32,43,44].Through this investigation, we have observed that Pl. fusiformispora can parasitize both Lepidoptera larvae and Ophiocordyceps sp.We speculate that Pl. fusiformispora may also exhibit a hyperparasitism phenomenon.In addition, Xiao et al. [1] identified that Pe. elaphomyceticola can parasitize Ophiocordyceps sp., suggesting that Pe. elaphomyceticola probably also exhibits a hyperparasitism phenomenon.Most species in the genus Pleurocordyceps and Perennicordyceps exhibit a hyperparasitism phenomenon [1].At present, little is known about the hyperparasitic mechanisms of the Pleurocordyceps and Perennicordyceps groups.Further research is needed on the relationship between parasitic species and their hosts.
In this paper, we present the basic genomic characteristics of Pe. elaphomyceticola and Pl.fusiformispora.The results showed that there was a certain difference in genome size, GC content, N50, and total number of genes.The BGCs of SMs that might be associated with hyperparasitism were analyzed through a gene mining analysis of Pe. elaphomyceticola and Pl.fusiformispora complexes.The estimated number of SM BGCs in the genomes of Pe. elaphomyceticola and Pl.fusiformispora were 39 and 50, respectively.However, we

Discussion
Herein, we describe a new species of Pl. fusiformispora using a combination of morphology and phylogeny.The newly established species distinctly form independent clades in the phylogenetic tree (Figure 1).Morphologically, the new species Pl. fusiformispora is similar to Pl. aurantiacus, Pl. agarica, Pl. heilongtanensis, Pl.lanceolatus, Pl. marginaliradians, Pl. nutansis, Pl. vitellina, and Pl.yunnanensis in that they have two types of phialides and conidia.However, the species Pl. fusiformispora has colonies that are flaky, α-conidia that are ovoid or elliptic, and β-conidia that are fusiform or long fusiform and differ from those of other species of Pleurocordyceps [1,24,25,42].The discovery of the new species of Pl. fusiformispora adds to the diversity of the genus Pleurocordyceps.
Xiao et al. [1] noted five new samples (Pl.sinensis GACP 19-2301, Pl. sinensis MFLU 21-0269, Pl. sinensis MFLU 21-0268, Pl. sinensis GACP 20-0865, and Pl.sinensis GACP 20-2304), each of which is parasitic on a different host clustered with Pl. sinensis.These hosts were clavicipitoid fungi and constituted new hosts for Pl.sinensis.In addition, Xiao et al. [1] introduced Pl. nutansis as a new species under Pleurocordyceps and identified the host as Ophiocordyceps nutans.However, in the phylogenetic tree, there was a disorder in the systematic position of these two different hosts, which might be due to genetic differences caused by different strains.
Some species of Polycephalomycetaceae have also been reported as hyperparasitic fungi; these species were Cordyceps, Elaphomyces, Hirsutella, Myxomycetes, and Ophiocordyceps [24,25,32,43,44].Through this investigation, we have observed that Pl. fusiformispora can parasitize both Lepidoptera larvae and Ophiocordyceps sp.We speculate that Pl. fusiformispora may also exhibit a hyperparasitism phenomenon.In addition, Xiao et al. [1] identified that Pe. elaphomyceticola can parasitize Ophiocordyceps sp., suggesting that Pe. elaphomyceticola probably also exhibits a hyperparasitism phenomenon.Most species in the genus Pleurocordyceps and Perennicordyceps exhibit a hyperparasitism phenomenon [1].At present, little is known about the hyperparasitic mechanisms of the Pleurocordyceps and Perennicordyceps groups.Further research is needed on the relationship between parasitic species and their hosts.
In this paper, we present the basic genomic characteristics of Pe. elaphomyceticola and Pl.fusiformispora.The results showed that there was a certain difference in genome size, GC content, N50, and total number of genes.The BGCs of SMs that might be associated with hyperparasitism were analyzed through a gene mining analysis of Pe. elaphomyceticola and Pl.fusiformispora complexes.The estimated number of SM BGCs in the genomes of Pe. elaphomyceticola and Pl.fusiformispora were 39 and 50, respectively.However, we found that Pl. fusiformispora has a large number of PKS and HR-PKS.The number of NRPS, other and PKS, and other in each species was small or even absent.It was also unknown which functions the existing Pe.elaphomyceticola and Pl.fusiformispora might perform in their life activities.
Through genome mining, Cordyceps militaris was found likely to have existing NRPS and PKS [45].In this study, many putative NRPS, PKS, and hybrid NRPS-PKS have been obtained.Many of the compounds were more similar to known gene clusters, such as Ochrindole A, Tubulysin A, Viriditixin, Cryptosporioptide BAC, Pyripyropene A, Leucinostatin A/Leucinostatin B, Emericellamide AB, Squalestatin S1, Sirodesmin PL, Pyripyropene, Ansaseomycin AB, Fusaric acid, Choline, Cyclopiazonic acid, Ilicicolin H, Lucilactaene, Ascochlorin, YWA1, Fusarin C, Betaenone C probetaenone I stemphyloxin II, AKML BD AC, Patulin, UNLL-YC2Q1O94PT, Scytophycin, Mangicol A, Ergotamine, and Oxaleimide C, could not be analyzed further because there were no modified genes, known gene clusters, key enzymes, indoles, or terpenes.The unique NR-PKS domains of Pl. fusiformispora include SAT-KS-AT-PT-ACP-Te, KS-AT-DH-MT, and PT-ACP-Te.At the same time, Pe. elaphomyceticola also had the unique NR-PKS domain of KS-AT-DH.They also had a common domain, NR-PKS (PKS-AT).The unique and common domains play an important role in the life cycle of each species.To exert biological functions, corresponding compounds need to be catalyzed, reflecting genetic differences between species.
Purev et al. [39] isolated the fungal gene "epls" encoding ε-poly lysine synthetase and confirmed that overexpression of epls in the different strain Epichloë festucae Fl1 resulted in the production of shorter ε-PL with 8-20 lysine, which exhibited a comparable antifungal activity to the longer one.In this study, the whole genome sequences of two species (Pe.elaphomyceticola Region 13.1 and Pl.fusiformispora Region 21.4) were present in homologous regions and were as high as 100% similar to the ε-PL gene sequence produced in MIBiG database.Therefore, the two species of Pe. elaphomyceticola and Pl.fusiformispora might produce ε-PL synthetase compounds, and their antifungal activities affect the growth of fungal hosts.
Zhang et al. [40] enacted genome mining of in-house strains using polyketide synthasenonribosomal peptide synthase as a query and identified an endophyte, Tolypocladium sp.49Y, which possessed a potential Tolypyridone biosynthetic gene cluster.Heterologous expression in Aspergillus oryzae NSAR1 revealed that this gene cluster is functional and able to produce a rare type of 4-hydroxy pyridones called Tolypyridone.And it was also found that Tolypyridone had antifungal activity.The whole genome sequences of Pl. fusiformispora were present in homologous Regions and were as high as 100% similar to the Tolypyridone gene sequence produced in the MIBiG database.Pl. fusiformispora may produce Tolypyridones compounds during infection with a parasite (Ophiocordyceps sp.), thereby inhibiting the normal development of Ophiocordyceps sp.Qin et al. [41] discovered that the compounds monocillin VI/monocillin VII/monocillin II/monorden D/monocillin IV/monocillin V/pochonin M had moderate antibacterial activity.It had also been proven that these compounds had the potential to control bacterial diseases.The whole genome sequences of Pl. fusiformispora were present in homologous regions and were as high as 100% similar to the monocillin VI/monocillin VII/monocillin II/monorden D/monocillin IV/monocillin V/pochonin M gene sequence produced in MIBiG database.As a result, Pl. fusiformispora might produce monocillin VI/monocillin VII/monocillin II/monorden D/monocillin IV/monocillin V/pochonin M. Further investigation of the compounds presented in this study is needed to understand the functionality of Pl. fusiformispora.
The triticones, also known as spirostaphylotrichins, were characterized by a spirocyclic γ-lactam core structure.The triticones A to F were first purified from Ptr culture filtrates but were then found, along with an additional 18 other triticone compounds, in five other ascomycete fungi, Staphylotrichum coccosporum, Curvularia pallescens, Pyrenophora seminiperda, Cochliobolus lunatus, and Bipolaris spp.[46][47][48][49][50][51][52].Triticone A and B are known to be phytotoxic, producing yellowish-brown lesions following leaf puncture assays on a range of hosts, including wheat.Triticones C and D were described as weakly active,

Figure 1 .
Figure 1.Phylogenetic tree of Polycephalomycetaceae was constructed, based on the concatenation of ITS, SSU, LSU, TEF1-α, RPB1, and RPB2 sequence data.The phylogeny was inferred using the IQ-tree.The maximum likelihood bootstrap valued greater than 60% (on the left) and the Bayesian posterior probabilities over 0.6 (on the right) were indicated above the nodes.The new species and known ones were indicated in back bold font.

Figure 1 .
Figure 1.Phylogenetic tree of Polycephalomycetaceae was constructed, based on the concatenation of ITS, SSU, LSU, TEF1-α, RPB1, and RPB2 sequence data.The phylogeny was inferred using the IQ-tree.The maximum likelihood bootstrap valued greater than 60% (on the left) and the Bayesian posterior probabilities over 0.6 (on the right) were indicated above the nodes.The new species and known ones were indicated in back bold font.

J
. Fungi 2024, 10, x FOR PEER REVIEW 19 of 28 core gene structural domain (KS-AT-ER-KR-ACP) in Pl. fusiformispora Region 68.1 was similar to KJ728786.1.However, Pl. fusiformispora lost some of its modified genes during evolution.It was speculated that the position direction of key enzymes and modified genes was different when different genera were in the compound.At the same time, there will be gene fusion and gene loss in the process of evolution.

Figure 8 .
Figure 8.Comparison of putative BGC of ε-poly lysine in Pe. elaphomyceticola and Pl.fusiformispora.The number after the region and the number before the decimal point represent the scaffold, and the number after the decimal point represents the gene cluster.

Figure 9 .
Figure 9.Comparison of biosynthesis of putative Tolypyridone in Pl. fusiformispora.The number after the region and the number before the decimal point represent the scaffold, and the number after the decimal point represents the gene cluster.

Figure 10 .
Figure 10.Comparison of biosynthesis of putative Monorden D/monocillin IV/monocillin VII/pochonin M/monocillin V/monocillin II in Pl. fusiformispora.The number after the region and the number before the decimal point represent the scaffold, and the number after the decimal point represents the gene cluster.

Figure 8 .
Figure 8.Comparison of putative BGC of ε-poly lysine in Pe. elaphomyceticola and Pl.fusiformispora.The number after the region and the number before the decimal point represent the scaffold, and the number after the decimal point represents the gene cluster.

Figure 8 .
Figure 8.Comparison of putative BGC of ε-poly lysine in Pe. elaphomyceticola and Pl.fusiformispora.The number after the region and the number before the decimal point represent the scaffold, and the number after the decimal point represents the gene cluster.

Figure 9 .
Figure 9.Comparison of biosynthesis of putative Tolypyridone in Pl. fusiformispora.The number after the region and the number before the decimal point represent the scaffold, and the number after the decimal point represents the gene cluster.

Figure 10 .
Figure 10.Comparison of biosynthesis of putative Monorden D/monocillin IV/monocillin VII/pochonin M/monocillin V/monocillin II in Pl. fusiformispora.The number after the region and the number before the decimal point represent the scaffold, and the number after the decimal point represents the gene cluster.

Figure 9 .
Figure 9.Comparison of biosynthesis of putative Tolypyridone in Pl. fusiformispora.The number after the region and the number before the decimal point represent the scaffold, and the number after the decimal point represents the gene cluster.

Figure 8 .
Figure 8.Comparison of putative BGC of ε-poly lysine in Pe. elaphomyceticola and Pl.fusiformispora.The number after the region and the number before the decimal point represent the scaffold, and the number after the decimal point represents the gene cluster.

Figure 9 .
Figure 9.Comparison of biosynthesis of putative Tolypyridone in Pl. fusiformispora.The number after the region and the number before the decimal point represent the scaffold, and the number after the decimal point represents the gene cluster.

Figure 10 .
Figure 10.Comparison of biosynthesis of putative Monorden D/monocillin IV/monocillin VII/pochonin M/monocillin V/monocillin II in Pl. fusiformispora.The number after the region and the number before the decimal point represent the scaffold, and the number after the decimal point represents the gene cluster.

Figure 10 .
Figure 10.Comparison of biosynthesis of putative Monorden D/monocillin IV/monocillin VII/pochonin M/monocillin V/monocillin II in Pl. fusiformispora.The number after the region and the number before the decimal point represent the scaffold, and the number after the decimal point represents the gene cluster.

Figure 10 .
Figure 10.Comparison of biosynthesis of putative Monorden D/monocillin IV/monocillin VII/pochonin M/monocillin V/monocillin II in Pl. fusiformispora.The number after the region and the number before the decimal point represent the scaffold, and the number after the decimal point represents the gene cluster.

Figure 11 .
Figure 11.Comparison of biosynthesis of putative Triticone DABFC in Pe. elaphomyceticola.The number after the region and the number before the decimal point represent the scaffold, and the number after the decimal point represents the gene cluster.

Figure 11 . 28 Figure 12 .
Figure 11.Comparison of biosynthesis of putative Triticone DABFC in Pe. elaphomyceticola.The number after the region and the number before the decimal point represent the scaffold, and the number after the decimal point represents the gene cluster.J. Fungi 2024, 10, x FOR PEER REVIEW 20 of 28

Figure 13 .
Figure 13.Comparison of biosynthesis of putative 4-epi-15-epi-brefeldin A in Pl. fusiformispora.The number after the region and the number before the decimal point represent the scaffold, and the number after the decimal point represents the gene cluster.

Figure 12 .
Figure 12.Comparison of biosynthesis of putative piperazine compound 1 and piperazine compound 2 in Pe. elaphomyceticola.The number after the region and the number before the decimal point represent the scaffold, and the number after the decimal point represents the gene cluster.

J 28 Figure 12 .
Figure 12.Comparison of biosynthesis of putative piperazine compound 1 and piperazine compound 2 in Pe. elaphomyceticola.The number after the region and the number before the decimal point represent the scaffold, and the number after the decimal point represents the gene cluster.

Figure 13 .
Figure 13.Comparison of biosynthesis of putative 4-epi-15-epi-brefeldin A in Pl. fusiformispora.The number after the region and the number before the decimal point represent the scaffold, and the number after the decimal point represents the gene cluster.

Figure 13 .
Figure 13.Comparison of biosynthesis of putative 4-epi-15-epi-brefeldin A in Pl. fusiformispora.The number after the region and the number before the decimal point represent the scaffold, and the number after the decimal point represents the gene cluster.

Figure 14 .
Figure 14.Clustering tree of NRPS proteins between Pe. elaphomyceticola and Pl.fusiformispora species and other fungi.Values at the nodes represent bootstrap values.Bold lines are shown at the nodes for 100 support.The scale bar 2.0 indicates the number of expected mutations per site.Bold indicates their clustering situation.

Figure 14 .
Figure 14.Clustering tree of NRPS proteins between Pe. elaphomyceticola and Pl.fusiformispora species and other fungi.Values at the nodes represent bootstrap values.Bold lines are shown at the nodes for 100 support.The scale bar 2.0 indicates the number of expected mutations per site.Bold indicates their clustering situation.

Figure 15 .
Figure 15.Clustering tree of hybrid PKS-NRPS proteins between Pe. elaphomyceticola and Pl.fusiformispora species and other fungi.Values at the nodes represent bootstrap values.Bold lines are shown at the nodes for 100 support.The scale bar 2.0 indicates the number of expected mutations per site.Bold indicates their clustering situation.

Figure 15 .
Figure 15.Clustering tree of hybrid PKS-NRPS proteins between Pe. elaphomyceticola and Pl.fusiformispora species and other fungi.Values at the nodes represent bootstrap values.Bold lines are shown at the nodes for 100 support.The scale bar 2.0 indicates the number of expected mutations per site.Bold indicates their clustering situation.

Figure 16 .
Figure 16.Scaffold synteny analysis of biosynthetic gene clusters containing secondary metabolites in the genomes of Pe. elaphomyceticola and Pl.fusiformispora.Numerical representation of gene length (bp).Different colored lines represent collinear regions between different genomes.

Figure 16 .
Figure 16.Scaffold synteny analysis of biosynthetic gene clusters containing secondary metabolites in the genomes of Pe. elaphomyceticola and Pl.fusiformispora.Numerical representation of gene length (bp).Different colored lines represent collinear regions between different genomes.

Table 1 .
Sources of selected isolates and GenBank accession number for ITS and five genes of three genera within Polycephalomycetaceae in this study.