Research article | Open | Published:
Estimated divergence times of Hirsutella (asexual morphs) in Ophiocordyceps provides insight into evolution of phialide structure
BMC Evolutionary Biologyvolume 18, Article number: 111 (2018)
Hirsutella Pat genus, the asexual morphs of the Ophiocordyceps Sung, is globally distributed entomopathogenic fungi, which infect a variety of arthropods, mites and nematodes. The fungal species also have shown potential application in the field of biological control, bio-medicine and food development. Although these fungi are synonymized under Ophiocordyceps, formal taxonomic assignments remain necessary for classification of species in Hirsutella. However, due to the heterogeneity and complexity of Hirsutella genus, more detailed taxonomic and phylogenetic analyses are required to address the following subjects: (1) the relationships between the phialide morphological characteristics and phylogenetic information of Hirsutella with asexual morphs, (2) the origin and evolution of the phialide structure, and (3) host specificity and fungal pathogenicity.
Five typical phialide structures are summarized, in which the variation in phialide characteristics overlaps well with phylogenetic information. A new member of the special twisted neck clade in the Hirsutella-like group, Ophiocordyceps retorta, was reported based on these analyses. The molecular clock calibration analysis based on one fossil record revealed that Hirsutella (asexual morph) species originated from a common ancestor approximately 102 million years ago (Mya) (Early Cretaceous, Lower Albian) and then resolved into two major lineages. One lineage was typically phialidic, which was a larger shape, including H. guyana, H. nodulosa and H. sinensis clades (86.9 Mya, 95% highest posterior density (HPD): 69.1–101.4 Mya). Another main lineage of the phialides was more diversified and smaller than the former, which included H. citriformis and H. thompsonii clades (71.9 Mya, 95% HPD: 41.8–99.6 Mya).
Our results showed that certain phialide characteristics of Hirsutella were phylogenetically informative for two groups of taxa. The differentiation of the phialides structures in the major clades demonstrated a clear evolutionary path of Hirsutella (asexual morph) species, which exhibited two trends depending on the host size. Fungi in one of the groups displayed elongated conidiogenous cells with increased complexity of auxiliary structures from the mycelia. The species in another group reduced the volume of phialides and spores, which might be due to an energy-efficient strategy. These results suggested that a common origin allowed for diversification of given clades into separate niches. The distinct parallel evolutionary path combined with the specific phialides structure might result in the host specificity of Hirsutella (asexual morphs). A direct relationship between Hirsutella (asexual morphs) and the Cretaceous–Tertiary extinction was not found, which suggested that the diversity of phialides is more likely to be caused by long-term environmental adaptation and evolution rather than dramatic extinction events. This evolutionary result might correspond to the background of important biological and geological events in the late Cretaceous occurring near the divergence times of Hirsutella (asexual morphs).
The genus Hirsutella Pat., established in 1892 by Patouillard with H. entomophila as the type, includes the asexual morphs of specialized parasites belonging to the family Ophiocordycipitaceae [1, 2]. It is an important fungal genus due to various features, including its potential for use in the biological control of invertebrate pests, nematodes and mites [3, 4]. The fungal species are also a popular traditional medicine and nutritious food in many Asian countries due to the production of several bioactive secondary metabolites [5,6,7,8,9].
To conform the International Code of Botanical Nomenclature (Article 59), many fungal taxa typified by asexual morphs has been proposed suppression of generic names [2, 7, 10, 11]. Currently, as the oldest asexual generic name associated with the ‘Ophiocordyceps sphecocephala clade’, Hirsutella was synonymized under Ophiocordyceps, most of which sporulate on cadaver of the host adult insects [12,13,14]. However, the genus Hirsutella was suggested to be separately- treated due to the complex interrelationship in their anamorphs, which has been applied to a heterogenous group of fungi across several families of Hypocreales . Therefore, additional sampling, more detailed taxonomic and phylogenetic analyses of these species are required to confirm their placement in Ophiocordyceps instead of suppression of the generic name continuous due to a lack of knowledge or importance of these fungi. These efforts maybe promote the impending taxonomic revisions of the family Ophiocordycipitaceae [11, 14, 16].
The main taxonomic features of Hirsutella (asexual morphs) that distinguish from other asexually typified genera, are the basally inflated phialide in a discontinuous hymenial layer and the conidia embedded in mucous sheaths . Since Hirsutella was first described, the morphological features of this genus have expanded to including the possibility of polyphialides lacking significant basal inflation and conidia lacking a mucous coat [11, 15]. According to Hodge , the morphological features of Hirsutella species are insufficient to determine the evolutionary relationships between species. Therefore, further works on the phylogeny of Hirsutella species have been conducted [19, 20]. Simmons et al.  attempted to define certain clades and correlate them with morphology. They recognized the genus as five distinct groups: H. citriformis, H. guyana, H. nodulosa, H. sinensis and H. thompsonii, and tried to lay the ground work for new nomenclature by identifying characters and supported clades for future partitioning of Ophiocordyceps. However, no robust pattern was found that based on the host, nor did Hodge .
In addition to the trends in host taxa, one morphological character, the structure of phialides, nearly overlaps with phylogenetic information from the phylogenetic tree established in our previous work . Phialides of Hirsutella species vary in size, shape and branching. In terms of size, the phialides could be divided into two groups. The phialide in one of the groups is shorter than 25 μm, such as H. citriformis  and H. thompsonii , and is long, with an average length of more than 40 μm (even reaching 100 μm in some species) in another group, such as H. aphidis, H. guyana, H. illustris and H. sinensis. As for the shape, Hirsutella species also display much diversity in phialides. H. sinensis has a typical bottle shape with a cylindrical base . H. nodulosa exhibits helical twisting at the apex of phialides and warts on phialides and hyphae . H. citriformis is mainly represented by a dumpy ovoid base and a single slender neck . H. thompsonii, the most widely studied Hirsutella species and an important biocontrol agent for mite pests in agriculture, has a cylindrical or round phialide that is small in shape. If the morphological characteristics of these phialides could be consistently associated with the molecular phylogeny, the data would be beneficial and provide evidence for support of revisions to the phylogeny and taxonomic transitions which move forward under the new rules.
As the host-specific entomopathogenic fungi, Hirsutella (asexual morphs) species are much different from Beauveria and Metarhizium, which have more than 700 hosts due to morphological structure and physiological characteristics . Despite the characteristics of single sporulation, low sporulation, slow growth and a unique host insect, Hirsutella have been preserved and evolved in the natural environment for a long time, which cause epidemics in their host population, and show remarkably environmental adaptability. It is demonstrated that conidia mucilage, like natural film coatings, contribute to spore surface hydrophobicity and play a positive role in conidial viability of Hirsutella satumaensis under adverse environments . The appearance of conidiogenous cells (phialides), such as the position, shape, length and size, are the main identifying characteristics in taxonomy and have been extensively studied. However, it is little understand whether there is a connection between the diversity of phialides and ecological adaptation in host specificity and entomopathogenicity. In addition, since it is difficult to amplify some universal gene fragments for the species, and only partial sequences are available for genetic analysis, few study has been performed to comprehensively analyse the evolution of Hirsutella (asexual morphs) at the molecular level [15, 28, 29]. To address these problems, more efforts should be made to supplement more available taxonomic information, to perform accurate phylogenetic research, to estimation the evolutionary timescales using the molecular clock, and to improve understanding of evolutionary processes across all taxonomic levels.
To better understand the internal phylogeny of the Hirsutella (asexual morphs) species and the relationship with phialide structure, we conducted a phylogenetic analysis with three loci from 40 species representing a broad diversity of conidiogenous cell structures in this study. On the basis of phylogenetic analysis, one fossil record of Paleoophiocordyceps coccophagus G. H. Sung, Poinar & Spatafora , which is phylogenetically close to Hymenostilbe and Hirsutella, was used as a calibration point for a molecular clock analysis to estimate the divergence times of the different lineages of the Hirsutella (asexual morphs) species, with particular regard to the phialide structure. We also attempt to reveal the origin, evolution history and relationship with pathogenicity and host specificity of the fungi. In addition, a new member of the twisted phialide neck group in Hirsutella (asexual morphs) was added from an epidemiological study on invertebrate insects in China.
The new Hirsutella (asexual morphs) species was isolated from cadaver of Cochlidiidae larvae (Lepidoptera) collected from Leigongshan Nature Reserve, Guizhou province, China (26° 15’ N, 108° 09′ E), on 12 August 2010 by X. Zou. The specimen and an isolated strain of asexual stage were deposited as GZUIFR-lgs2 and GZUIFR-hir100812, respectively.
Fungal isolation and culture
The isolated strain GZUIFR-hir100812 was determined with sterile culture as described previously [30, 31]. Briefly, the surface of specimen was rinsed with sterile water, followed by surface sterilization with 75% ethanol. A piece of mycelia from the haemocoel was used to inoculate on potato dextrose agar (PDA) plate that was allowed to grow for 14 days at 16 °C under 12-h light/12-h dark conditions.
OM and SEM observations
For the asexual conidiogenous structure and conidia observations, fresh hyphae were observed and imaged using an optical microscope (OM, BK5000, OPTEC, USA) after staining with lactic acid phenol cotton blue solution. The captured images were edited and digitally contrasted with Paint Shop Pro v. 5.0.1 (Corel, Ottawa, Canada) .
Scanning electron microscopy (SEM) was conducted according to Qu et al. [27, 30]. Briefly, the fresh fungal hyphae were collected from PDA cultures and fixed with 4% glutaraldehyde at 4 °C overnight. Then the sample was washed three times with phosphate buffer solution (PBS) for 10 min/time. The fixed hyphae and conidia were dehydrated using 50, 70, 90 and 100% alcohol for 10 min/each level, and with supercritical carbon dioxide. After sprayed with gold, the samples were examined via SEM (S-3400 N, HITACHI, Japan) and imaged.
DNA extraction, PCR amplification and sequencing
Mycelia (0.05–0.1 g) were harvested from PDA plates and used for genomic DNA extraction with a Fungi DNA isolation Kit according to the manufacturer’s instructions (OMEGA Bio-Tek, USA). Sequences from two nuclear loci, the large subunits of the rDNA (LSU) and the internal transcribed spacers (ITS1–5.8S rDNA–ITS2 region, ITS), and the transcription elongation factor-1α (TEF), were used to phylogenetic analyses. The ITS ribosomal and LSU region were amplified using PCR with primers ITS1/ITS4 and LRor/LR5 respectively [11, 32,33,34]. The TEF gene was amplified with the primers EF1T and 1567R . PCR product was sequenced and new sequences were submitted in GenBank as shown in Additional file 1: Table S1 (the GenBank access numbers in bold).
According to our previous phylogenetic studies, the representative taxa of Hirsutella members from the major clades were chosen to construct a phylogeny [21, 32]. A total of 40 taxa were selected to represent the morphological and ecological diversity of Hirsutella (asexual morphs), including the outgroup taxa Purpureocillium (Cordyceps cylindrica) and Drechmeria (Drechmeria gunnii), which are also classified into Ophiocordycipitaceae . The GenBank accession numbers and strain voucher information are shown in Additional file 1: Table S1. DNA sequences were aligned with the Clustal X ver. 2.0 software as described previously . A combined dataset with the three regions, consisting of LSU (572 bp), TEF (581 bp) and ITS (577 bp) were analysed. The Akaike information criterion (AIC) in jModelTest 2.1.7 was used to select nucleotide substitution models for each three partition [36, 37]. Maximum likelihood (ML) phylogenetic analyses of the combined dataset with partition model parameters in RAxML ver. 8.0 was conducted to determine the best tree topology and bootstrap support values from 500 search replicates , which are summarized in FigTree v 1.4.2. Bayesian posterior probabilities (BPPs) were estimated with the same partition parameters in an analysis conducted in MrBayes v 3.1.2 , in which two runs of four chains each were simultaneously executed for 5,000,000 generations, with sampling every 500 generations . TreeGraph ver. 2 was used to compute BPP from a summary of 7500 trees retained after a burn-in of the first 2500 trees collected.
Divergence time estimation of phialide structures in Hirsutella (asexual morphs)
A tree was generated with the aligned sequence data using the MCMC (Markov chain Monte Carlo) process to infer divergence times with BEAST ver. 2.2.1 . One internal calibration point corresponding to the fossil record, P. coccophagus, a fungal parasite of a scale insect from the Early Cretaceous (Upper Albian) (99–105 Mya), was used in the analysis as soft constraints following a uniform limitation [29, 41]. The root node, which represented the rise of Ophiocordyceps, was set at 99–105 Mya . In order to estimate nucleotide substitutions, which were used to count branch lengths and variance-covariance structure of the branch length with the estbranches program, the baseml program of PAML ver. 3.14 with GTR + G was applied to independently calculate each gene as recommended [29, 42].
The gamma-distributed rate variation and a proportion of invariant sites of heterogeneity model were applied with base frequencies estimated during the analysis . A relaxed molecular clock using the uncorrelated log-normal model (mean of root age: 102 with a standard deviation (SD) of 1.5, 0.001 for root rate with an SD of 0.0011; 0.005 for rate autocorrelation with an SD of 0.005) was applied with a Yule process speciation prior for branching rates . The Metropolis-coupled Markov Chain Monte Carlo (MCMC) was run for 10 million generations following a burn-in of 1 million generations and was sampled every 1000 generations. Three replicate runs were performed to check the convergence of the MCMC. The measures of effective sample sizes (ESS) were used to determine the Bayesian statistical significance of each parameter (ESS > 200). Posterior ages were nearly identical among the three analyses, which was consistent with the convergence of the Bayesian dating analyses. Mean parameter estimates and 95% highest posterior densities were determined by analyzing the combined BEAST tree files using Tracer v.1.7 .
A tree that was constructed using maximum likelihood and Bayesian posterior probabilities with Drechmeria gunnii and Cordyceps cylindrica as outgroups (Fig. 1) could be broadly divided into five taxa, which correspond to those in our previous studies [11, 21, 30]. The conidiogenous cell structures with asexual morph morphology were also demonstrated to be consistent with phylogenetic divisions. Our results indicated that the Hirsutella (asexual morphs) were restricted to two major groups, group A containing the clades H. guyana, H. nodulosa and H. sinensis, which occupied a long branch with strong support and displayed larger phialides, and group B with smaller phialide structures, including the clades H. citriformis and H. thompsonii.
The group A possess larger conidiogenous cells, verruculose apex, and orange segment-like conidiophores, including clade H. guyana, H. nodulosa, H. sinensis and the new species. The H. nodulosa clade contained all Hirsutella (asexual morph) species with a twisted neck and a bootstrap support of 99% ML/100% BPP. O. acicularis, which has been placed in the H. sinensis clade according to Simmons et al. , was shifted in clade H. nodulosa in this work, likely due to the inclusion of the ITS rDNA data. The H. sinensis clade included isolates originating from a variety of insect hosts, represented with H. sinensis, H. strigosa and H. rhossiliensis. The phylogenetic analysis placed the H. guyana clade as sister with H. sinensis clade with a bootstrap support of 74% ML/85% BPP.
The group B included H. citriformis and H. thompsonii clad with inflated, ovoid bases forming cymbiform or ellipsoid conidia. H. citriformis was the main represented species in H. citriformis clade, which is a widely distributed species infecting the Asian citrus psyllid Diaphorina citri . The H. citriformis clade received bootstrap support values of 80% ML/95% BPP in the phylogenetic tree. The H. thompsonii clade, which was composed of five H. thompsonii species, H. necatrix and H. cryptosclerotium, and had a bootstrap support of 100% ML/100% BPP.
In addition, the new species O. retorta was clustered with H. leigongshanensis and acquired a strong bootstrap support (97% ML/100% BPP) in the H. nodulosa clade, which suggested that O. retorta is one member of the special group possessing helical twisting at the apex of phialides. However, despite the high bootstrap support, some differences in morphological characteristics were observed between O. retorta and H. leigongshanensis. O. retorta has two types of conidiogenous cells with large conidia (8.5–9.6 × 4.8–6 μm). More comparisons of morphological features with other related species was shown in Additional file 1: Table S2. Molecular phylogenetic analysis further confirmed the differences between the new species and other related species. Based on the molecular phylogeny together with morphological, biological and ecological characteristics, a distinct new Ophiocordyceps species is proposed.
Phialide characteristics for the major Hirsutella asexual morphs clades reported in our phylogeny are summarized in Table 1.
H. thompsonii clade
H. thompsonii forms the core of a group of morphologically related species, including H. cryptosclerotium, H. gregis, H. necatrix, H. sphaerospora and H. tydeicola, pathogens of mites or nematodes, which display short bottle-shaped or basally inflated, globose phialides (generally no more than 20 μm), and subglobose conidia.
H. citriformis clade
The distinguished characteristics of this clade are inflated bases that narrow abruptly into a narrow long neck (20–40 μm), which mainly contains H. gigantea, H. guignardii, H. fusiformis, H. longicolla, H. radiate, H. saussurei and Ophiocordyceps elongata.
H. nodulosa clade
Clade H. nodulosa is a much special group of Hirsutella (asexual morphs) due to the distinctive helical twisting at the apex of phialides with warts on the base, which contains 11 members so far. However, H. subulata and several other Ophiocordyceps species without the twist structure are always clustered in this clade from phylogenetic topology analysis [11, 20]. One possible explanation is that the anamorphs of those Ophiocordyceps species might have a similar structure which has not been reported, or that lack of a twisted structure might be due to degeneration or loss during evolution.
H. sinensis clade
The H. sinensis clade have a complex range of hosts, including nematodes, mites and both hemi- (Hemiptera) and holometabolists (Coleoptera, Lepidoptera). The phialides of this group possess cylindric, slender or subulate bases, gradually tapering to a warted neck, usually more than 30 μm, but those of some species are exceptionally long and verruculose, even more than 100 μm, such as H. strigosa and H. illustris.
H. guyana clade
The H. guyana clade is represented by three Hirsutella isolates characterized in our study, which grouped with the H. sinensis clade and share similarly large conidiogenous cells and typical phialides, subulate with a basal cylindric portion.
Dating and evolution of phialide structures in Hirsutella (asexual morphs)
We selected 40 main species with representative types of conidiogenous cells in Hirsutella (asexual morph) according to previous phylogenetic studies . The data of nuclear genes individually provided limited information and could not perform a robust analysis. A Bayesian inference phylogenetic tree, which was constructed with a dataset consisting of 1703 bp from three genes and with a partitioning strategy, was established and yielded a clear hypothesis regarding the evolution of the different phialide structures in Hirsutella (asexual morphs) species (Fig. 1).
In Fig. 2, the cluster of taxa with same phialide structure nearly coincided with the topology of the phylogenetic tree (Fig. 1), which provided an insight into the evolution of the internal phylogeny, origin, and evolutionary history of different types of phialide. Dating analyses support the supposition that the ancestor of the Hirsutella (asexual morphs) fungi was at least present in clade 1, that is, in the Early Cretaceous (Lower Albian) (101.9 Mya, 95% HPD: 99.0–104.9 Mya). The other clades distributed and diversified in the Upper Cretaceous (Fig. 2 and Table 2). Then, the new tree resolved the Hirsutella (asexual morphs) lineage into two major clades, 2 and 3, with topology corresponding to Fig. 1. The clade 2 included the H. thompsonii and H. citriformis clades, and clade 3 represented the H. guyana, H. nodulosa and H. sinensis clades. Since originated from a common ancestor, clade 2 and clade 3 were derived and evolved independently in parallel (Fig. 2). According to the clear divergence age estimates, clade 4 as well as clade 5 originated earlier as an independent group in the Early Maastrichtian (86.9 Mya, 95% HPD: 69.1–101.4). The age estimate of the crown node of clade 4 is at least 70.5 Mya (95% HPD: 48.3–90.0) in the Later Maastrichtian; and clade 5 diverged approximately 66.8 Mya (95% HPD: 42.2–91.8). In addition, clade 8 and clade 9 were originated later, at 71.9 Mya (95% HPD: 41.8–99.6 Mya), all sharing smaller phialide structures. The two clades centred at 47.9 Mya (95% HPD: 21.0–80.0) and 40.1 Mya (95% HPD: 11.9–71.6), respectively, around the Paleogene of Cenozoic (Fig. 2). Additionally, no major divergence of Hirsutella (asexual morphs) lineages was correlated with the Cretaceous–Tertiary extinction event (approximately 65.5 Mya). Therefore, a causal relationship between mass extinction events and the evolution of the phialides of the Hirsutella asexual morphs is not validated.
Ophiocordyceps retorta X. Zou, J.J. Qu & Z.Q. Liang, sp. nov. Fig. 3.
MycoBank no.: MB 820078; GenBank no.: KY415594, KY415595, KY415597.
Type: China, Guizhou Province, Leigongshan Nature Reserve, hyphae emerging from a larval cadaver of a species of Cochlidiidae (Fig. 3a, 12 August, 2010, X. Zou. Holotype: GZUIFR-lgs2 and an isolated strain of its asexual stage GZUIFR-hir100812 were deposited at the Institute of Fungal Resources of Guizhou University (GZAC).
Ophiocordyceps retorta differs from related species mainly in its two types of phialides, verruculose, twisted or undulate neck.
Host was approximately 2.5 cm long × 0.9 cm wide. Synnemata were approximately 60–100 mm long × 0.3 mm wide, several, cylindrical, tapering at the end, arising from the thorax of the insect, sometimes simple-branched, brown, fawn yellow near the apex. The fungus hardly grows above 30 °C, while slowly spreads, with 15–23 mm dim. After 20 d on PDA agar under 20–22 °C (Fig. 3b&c). Colony was circular, dense and flat, centre of surface with light brown synnemata, pale yellow filiform aerial hyphae on colonies margins; light brown pigment secreting into the media making the back of the colonies show brown, 8–15 mm dim. Mycelium was hyaline, smooth verruculose and septate, with 2.1–3.5 μm wide. Conidiogenous cells with two types: A-phialides (Fig. 3d), 24–28.8 × 1 μm long, subulate with barely inflated base, 3–3.5 μm broad, tapering to a slender, verruculose and twisted neck, 9.6–10 μm long, 1 μm wide; and B-phialides (Fig. 3f), slender gourd-shaped or awl-shaped and tapered, width of base 3.5 × 1.5–2 μm, length 40.8–52.6 μm, width of neck 1.8–2.0 μm; both phialides types formed directly from the mycelium end, occasional polyphialidic. Conidia shaped like orange segments or oval, 8.5–9.6 × 4.8–6 μm, enveloped in hyaline mucus, thickness 1.8–2.7 μm (Fig. 3g).
Host: Larva of a species of Cochlidiidae, Lepidoptera.
Etymology: refers to the structure of the phialide neck, twisted (Lat. “retorta”).
Distribution: Leigongshan, Guizhou Province, China.
Previous studies with sequence data from additional genes demonstrated that Hirsutella (asexual morphs) are a monophyletic group with good statistical support [2, 29, 30], which was also supported by our phylogeny. Combining these results, it is apparent that phialides structure is both phylogenetically and taxonomically informative. Because the ITS sequence has a high resolution in the Hirsutella species distinction, it is necessary to add it into this phylogenetic analysis. However, the differences in topology of phylogenetic tree largely attributed to whether the ITS rDNA sequences are included or not. Further study is needed to confirm appropriate genes and polygenes of this genus. For example, our phylogenetic analysis removed Ophiocordyceps acicularis from the H. sinensis clade and placed it on the polytomy alongside O. cf. acicularis within the H. nodulosa clade. Simmons et al. describe that this species are not monophyletic, with representatives in both the H. nodulosa and H. sinensis clades . However, we believed that this species has a close relationship with the twisted neck group. Further information is clearly required to resolve the relationships of these asexual morphs.
Sung et al. described an astonishing ascomycetous fossil, Paleoophiocordyceps coccophagus, in Burmese amber from 100 Mya, which is the oldest evidence of animal parasitism by fungi . They firstly used the fossil and provided a minimum date for the age of the crown members of the Ophiocordycipitaceae, and then estimated that the Hypocreales as a whole were Jurassic in origin (150–200 Mya). Berbee et al. subsequently confirmed its accuracy and estimate an age of Ophiocordycipitaceae was 122 Mya (95% HPD: 109–138 Mya) . Some additional examples of fungal pathogens in Dominican amber (30–45 Mya based on coccoliths) are synnemata produced by a Hirsutella-like fungus on a Troctopsocopsis bark louse . The fossil was used for a crown calibration point for Hirsutella (asexual morphs) in Bayesian relaxed clock analyses and in age estimates of its crown group, dating it at least to the Early Cretaceous (101.9 Mya). Thus, we considered that the divergence times of Hirsutella asexual morphs are reliable.
The differentiation between phialides structures in major clades demonstrated a clear evolutionary path of Hirsutella species. The Hirsutella (asexual morphs) originated at the earliest, approximately 71.9 Mya in the Later Maastrichtian. The evolutionary trajectories of phialides followed two basic trends to host size with some degree of relationship. One trend, including lineages 8 and 9, has small phialides, irrespective of the presence of the host taxa used, usually being less than 25 μm and rarely with warts on the phialides. Conversely, another trend differs from the above in proliferation of phialide type, size range, and host diversity, which might increase the adhesive surface of conidia and sporulation efficiency. The nearly simultaneous origins that we observed for these two major lineages may be linked to the diversification of each clade into separate niches. In the current study, we did not find a direct relationship between Hirsutella (asexual morphs) and the Cretaceous-Tertiary extinction event . This suggested that the diversity of phialides is more likely to be caused by long-term environmental adaptation and coevolution with insects, rather than dramatic extinction events. These origin and evolution results might be associated with the background of important biological and geological events in the Upper Cretaceous (approximately 60.0–100.5 Mya). Around the divergence times of Hirsutella (asexual morphs), the diversification of phialide events might accompanied with the abiotic stress of sea level and climate changes leading to high biotic stress conditions largely due to high nutrient levels, coevolving with increased plants and arthropods in late Cretaceous [47, 48].
The differences in the pattern of fungal evolution in regional communities may result from species differences in the relative importance of biotic interactions and coevolution [49,50,51]. With the angiosperm significant diversification between 115 and 90 Mya [52,53,54,55,56], a radiation of diverse monocot and eudicot lineages during this period began to modify the terrestrial ecosystems and initiate interactions with other organisms . Cretaceous diversification of the pollinating insects, phytophagous insects and entomogenous fungi are consistent with the diversification of the angiosperms . Similar to angiosperms, the exopterygotes and endopterygotes also radiated dramatically in the Cretaceous [58, 59]. Based on the functions of temporal and geographical distributions, all newly isolated Hirsutella species populations probably were forced into new coevolutionary pathways, with different types of conidiogenous cell structures evolving along new trajectories.
Coevolution between the fungus and its host insects might have originated a long time. It has been proposed that most groups of parasitic fungi probably evolved from saprophytic species developing on dead arthropods and plants [60, 61]. It has been demonstrated that plants are the potential hosts of O. sinensis, which seems to be able to jump from plants to insects and back onto a fungal host in accordance with the origins and evolution of entomopathogenicity [62, 63]. Such an evolutionary hypothesis might be used to explain the host specificity of Hirsutella (asexual morphs). It is reasonable to hypothesize that the host specificity of Hirsutella (asexual morphs) might relevant to the parallel evolutionary path and phialides diversification. Coevolution between the Hirsutella (asexual morphs) results in special physiological adaptations, which allow them to attack different insects or plants and lead to formation of a more suitable or advanced morphology of conidiogenous cells from the original structure. Such evolution in turn caused the special phialides structural differences that infect a group of related hosts. From an evolutionary point of view, the H. thompsonii and H. citriformis clades, sharing small hosts and simple phialides, originated earliest in Hirsutella, which might evolve an energy-efficient strategy to adapt niched of the host habitat. The subsequent evolution of lineages might result in producing more complex phialide structures, including the H. nodulosa and H. sinensis lineages. These results suggested that the diversity and large size of hosts possibly provided the basis for the formation of phialide complexity. For instance, the H. nodulosa clade, which originated in 70.5 Mya, might have increased the length of the neck to enhance the spread of spores by the dynamics of their ejection with twisted or undulated necks. It has been demonstrated that the subterminal part of phialide often display a helical or zigzag twisting in older cultures of fungi. Undulate and wavy tips are more pronounced in dead phialides, which reveals that the ecological functions of this structure are related to the spread of spores [64,65,66]. Therefore, we hypothesized that the twisted phialide neck might evolved independently under physiological adaptability to induce nutritional growth (access to nutrients, sunlight, water or space expansion better) or asexual reproduction (spore maturity and spread).
In Hirsutella (asexual morphs), another noted phenomenon is the polyphialidic characteristic, which is prevalent in various lineages, irrespective of the shape and size of phialides. This suggests that the evolution of the phialide structure is consistent with the progression from a simple to a more advanced form. It might also be an energy-efficient strategy to have one phialide with more than one locus of conidial development in a habitat, in which is a nutrition deficiency. In addition, a finer and more complex structure will always appear in the lineages with a large host, such as the helical neck, warted phialides, and whorled phialides, which are always present in the H. nodulosa clade or H. sinensis clade. More evidence, including additional fossil records, is needed to reveal the physiological function, ecological function, and phylogenetic position of this special phialide structure.
At present, the special group with a twisted neck of conidiogenous cells in Hirsutella (asexual morphs) species contains 11 members: H. aphidis Petch , H. brownorum Minter & B.L. Brady , H. leizhouensis H.M. Fang & S.M. Tan , H. liboensis Zou et al. , H. nodulosa Petch , H. dendritica Samson & H.C. Evans , H. parasitica Samson & H.C. Evans , H. satumaensis Aoki , H. vermicola M.C. Xiang & X.Z. Liu , H. tortricicola Zou et al.  and H. leigongshanensis Zou et al. . In this study, we added a new member of the twisted phialide neck group, Ophiocordyceps retorta sp. nov. X. Zou, J.J. Qu & Z.Q. Liang. This species differs from related species mainly in its two types of phialides, verruculose, and twisted neck and its occurrence on larvae of Cochlidiidae (more morphological comparisons of the new species with other related species are shown in Additional file 1: Table S2).
We have studied the pivotal internal phylogeny, morphological characteristics of phialide, origin and evolutionary history of the phialide structure in Hirsutella (asexual morphs). Based on phylogenetic and morphological analyses, five representative types of phialides, including the H. citriformis clade (representing by a dumpy ovoid base and a single slender neck), the H. thompsonii clade (with a cylindrical or round phialide that is small in shape, usually less than 25 μm), the H. sinensis clade (sharing a cylindrical base and an average phialide length of more than 40 μm), the H. guyana clade (typical phialides, subulate with a basal cylindric portion) and the H. nodulosa lineage (possessing phialides with apical helical twists) were firstly summarized that had a strong connection with the phylogenetic topology. The morphological characteristics and molecular genetic information could be productively combined for further generic classification studies, with great significance to the taxonomic study of Hirsutella (asexual morphs). The divergence of phialides structures in major clades demonstrated a clear evolutionary path of Hirsutella (asexual morphs), with a possible connection to host size that lead to two basic trends. One is elongation of the conidiogenous cells and increased complexity on auxiliary structures from the mycelium, and another is reducing the volume of phialides and spores, which might be linked to an energy-efficiency strategy. No major lineages was noted to be correlated with the nearby Cretaceous–Tertiary extinction event, which means that the diversity of phialides is more likely to be caused by long-term environmental adaptation and coevolution with insects rather than dramatic extinction events. These results suggested that the host specificity of Hirsutella asexual morphs might be relevant to the evolutionary pathways of phialides diversification.
The Akaike information criterion
Bayesian posterior probabilities
The measures of effective sample sizes
The Institute of Fungal Resources of Guizhou University
The highest posterior density
ITS1–5.8S rDNA–ITS2 region, the first and the internal transcribed spacers
The large subunits of the rDNA
The Markov Chain Monte Carlo
Million years ago
Phosphate buffer solution
Polymerase chain reaction
The Potato Dextrose Agar
Scanning Electron Microscope
The transcription elongation factor-1α
Patouillard N. Une Clavariée entomogène. Rev Mycol. 1892;14:67–70.
Sung G, Sung J, Hywel-Jones NL, Spatafora JW. A multi-gene phylogeny of Clavicipitaceae (Ascomycota, Fungi): identification of localized incongruence using a combinational bootstrap approach. Mol Phylogenet Evol. 2007b;44(3):1204–23.
Jaffee BA. Population biology and biological control of nematodes. Can J Microbiol. 1992;38(5):359–64.
Van der Geest LPS. IPM potentials of microbial pathogens and diseases of mites. In: Ciancio A, Mukerji KG, editors. Integrated management of arthropod pests and insect borne diseases, vol. 5: Springer, Dordrecht; 2010. p. 249–309.
Evans HC. Natural control of arthropods, with special reference to ants (Formicidae), by fungi in the tropical high forest of Ghana. J Appl Ecol 1974; 11: 37–49.
Zhang Y, Li E, Wang C, Li Y, Liu X. Ophiocordyceps sinensis, the flagship fungus of China: terminology, life strategy and ecology. Mycology. 2012;3:2–10.
Quandt CA, Kepler RM, Gams W, Araújo JPM, Ban S, et al. Phylogenetic-based nomenclatural proposals for Ophiocordycipitaceae (Hypocreales) with new combinations in Tolypocladium. IMA Fung. 2014;5:121–34.
Mazet I, Vey A. Hirsutellin a, a toxic protein produced in vitro by Hirsutella thompsonii. Microbiology. 1995;141:1343–8.
Lang G, Blunt JW, Cummings NJ, Cole ALJ, Munro MHG. Hirsutide, a cyclic tetrapeptide from a spider-derived entomopathogenic fungus, Hirsutella sp. J Nat Prod. 2005;68:1303–5.
McNeill J, Barrie FR, Buck WR, Demoulin V, Greuter W, Hawksworth D, Herendeen P, Knapp S, Marhold K, Prado J, Prud'homme Van Reine W, Smith G, Wiersema J, Turland N. International Code of Nomenclature for algae, fungi and plants. Regn Veget. 2012;154
Simmons DR, Kepler RM, Rehner SA, Groden E. Phylogeny of Hirsutella species (Ophiocordycipitaceae) from the UsA: remedying the paucity of Hirsutella sequence data. IMA Fung. 2015b;6(2):345–56.
Maharachchikumbura SS, Hyde KD, Jones EG, McKenzie EH, Huang SK, Abdel-Wahab MA, et al. Towards a natural classification and backbone tree for Sordariomycetes. Fungal Divers. 2015;72(1):199–301.
Maharachchikumbura SS, Hyde KD, Jones EG, McKenzie EHC, Bhat JD, Dayarathne MC, et al. Families of Sordariomycetes. Fungal Divers. 2016;79:1–317.
Wijayawardene NN, Hyde KD, Rajeshkumar KC, Hawksworth DL, Madrid H, Kirk PM, et al. Notes for genera: Ascomycota. Fungal Divers. 2017;86(1):1–−594.
Spatafora JW, Quandt CA, Kepler RM, Sung GH, Shrestha B, Hywel-Jones NL, Luangsa-ard JJ. New 1F1N species combinations in Ophiocordycipitaceae (Hypocreales). IMA Fung. 2015;6(2):357–62.
Simmons DR, Lund J, Levitsky T, Groden E. Ophiocordyceps myrmicarum, a new species infecting invasive Myrmica rubra in Maine. J Invertebr Pathol. 2015a;125:23–30.
Liang ZQ. Studies on classification of the genus Hirsutella Pat. I. Advances and the characteristics of taxonomy. J Guizhou Agricul College. 1990a;9(1):58–68.
Hodge KT. Revisionary studies in Hirsutella (Hyphomycetes), PhD Dissertation. Ithaca: Cornell University; 1998. p. 25.
Evans HC, Groden E, Bischoff JF. New fungal pathogens of the red ant, Myrmica rubra, from the UK and implications for ant invasions in the USA. Fung Biol. 2010;114:451–66.
Ciancio A, Colagiero M, Rosso LC, Murga Gutierrez SN, Grasso G. Phylogeny and morphology of Hirsutella tunicata sp. nov. (Ophiocordycipitaceae), a novel mite parasite from Peru. Mycoscience. 2013;54:378–86.
Qu JJ, Zou X, Yu JP, Zhou JX, Zhou YM, Liang ZQ. Re-description of two entomopathogenic fungi in Hirsutella pat. Microbiol China. 2017;44(11):2615–23.
Mains EB. Entomogenous species of Hirsutella, Tilachlidium and Synnematium. Mycologia 1951; 43(6): 691–718.
Fisher FE. Two new species of Hirsutella Patouillard. Mycologia. 1950;42:290–7.
Liu XJ, Guo YL, Yu YX, Zeng W. Isolation and indentification of the anamorphic state of Cordyceps sinensis (Berk.) sacc. Mycosystema. 1989;8(1):35–40.
Petch T. Entomogenous fungi: additions and corrections, II. Trans Br Mycol Soc. 1926;11(3–4):258–66.
Speare AT. On certain entomogenous fungi. Mycologia. 1920;12(2):62–76.
Qu JJ, Zou X, Yu JP, Zhou YM. The conidial mucilage, natural film coatings, is involved in environmental adaptability and pathogenicity of Hirsutella satumaensis Aoki. Sci Rep. 2017;7(1):1301.
Tigano MS, Adams B, Maimala S, Boucias D. Genetic diversity of Hirsutella thompsonii isolates from Thailand based on AFLP analysis and partial b-tubulin gene sequences. Genet Mol Biol. 2006;29:715–21.
Sung GH, Poinar GO, Spatafora JW. The oldest fossil evidence of animal parasitism by fungi supports a cretaceous diversification of fungal–arthropod symbioses. Mol Phylogenet Evol. 2008;49(2):495–502.
Qu JJ, Yu LQ, Zhang J, Han YF, Zou X. A new entomopathogenic fungus, Ophiocordyceps ponerus sp. nov., from China. Phytotaxa. 2018;343(2):116–26.
Zou X, Liu AY, Liang ZQ, Han YF, Yang MF. Hirsutella liboensis, a new entomopathogenic species affecting Cossidae (Lepidoptera) in China. Mycotaxon 2010; 111(1): 39–44.
Luangsa-ard JJ, Mongkolsamrit S, Noisripoom W, Thanakitpipattana D, Khonsanit A, Wutikhun T. Helicocollum, a new clavicipitalean genus pathogenic to scale insects (Hemiptera) in Thailand. Mycol Prog. 2017;16(4):1–13.
Rehner SA, Samuels GJ. Taxonomy and phylogeny of Gliocladium analysed from nuclear large subunit ribosomal DNA sequences. Mycol Res. 1994;98:625–34.
White TJ, Bruns TD, Lee SJ, Taylor JW. Amplification and direct sequencing of fungal ribosomal RNA genes for phylogenetics. In: Innis MA, Gelfand DH, Sninsky JJ, White TJ, editors. PCR protocols: a guide to methods and applications. San Diego: Academic Press; 1990. p. 315–22.
Rehner SA, Buckley EA. Beauveria phylogeny inferred from nuclear ITS and EF1-α sequences: evidence for cryptic diversification and links to Cordyceps teleomorphs. Mycologia. 2005;97:84–98.
Guindon S, Gascuel O. A simple, fast, and accurate algorithm to estimate large phylogenies by maximum likelihood. Syst Biol. 2003;52(5):696–704.
Posada D. jModeltest: phylogenetic model averaging. Mol Biol Evol. 2008;25:1253–6.
Stamatakis A, Hoover P, Rougemont J. A rapid bootstrap algorithm for the RAxML web-servers. Syst Biol. 2008;75(5):758–71.
Ronquist F, Huelsenbeck JP. MrBayes 3: Bayesian phylogenetic inference under mixed models. Bioinformatics. 2003;19:1572–4.
Bouckaert R, Heled J, Kühnert D, Vaughan T, Wu CH, Xie D, et al. BEAST 2: a software platform for Bayesian evolutionary analysis. PLoS Comput Biol. 2014;10:e1003537.
Yang E, Xu L, Yang Y, Zhang X, Xiang M, Wang C, et al. Origin and evolution of carnivorism in the Ascomycota (fungi). Proc Natl Acad Sci. 2012;109(27):10960–5.
Berbee ML, Taylor JW. Dating the molecular clock in fungi-how close are we? Fungal Biol Rev. 2010;24:1–16.
Lanfear R, Calcott B, Ho SY, Guindon S. Partitionfinder: combined selection of partitioning schemes and substitution models for phylogenetic analyses. Mol Biol Evol. 2012;29:1695–701.
Li M, Tian Y, Zhao Y, Bu W. Higher level phylogeny and the first divergence time estimation of Heteroptera (Insecta: Hemiptera) based on multiple genes. PLoS One. 2012;7(2):e32152.
Poinar G. Evolutionary history of terrestrial pathogens and endoparasites as revealed in fossils and subfossils. Adv Biol. 2014;2014:1–29.
Schulte P, Alegret L, Arenillas I, Arz JA, Barton PJ, Bown PR, et al. The Chicxulub asteroid impact and mass extinction at the cretaceous-Paleogene boundary. Science. 2010;327:1214–8.
Coyne JA, Orr HA. Speciation. Sunderland. MA: Sinauer Associates; 2004.
DeConto RM, Brady EC, Bergengren J, Hay WW. Late cretaceous climate, vegetation and ocean interactions. In: Huber BT, KG ML, Wing SL, editors. Warm climates in earth history. New-York: Cambridge University Press; 2000. p. 275–96.
Dobzhansky T. Evolution in the tropics. Am Sci. 1950;38(2):209–21.
Schemske D. Tropical diversity: patterns and processes. In: Chazdon R, Whitmore T, editors. Ecological and evolutionary perspectives on the origins of tropical diversity: key papers and commentaries. Chicago: University of Chicago Press; 2002. p. 163–73.
Giraud T, Refrégier G, Le Gac M, de Vienne DM, Hood ME. Speciation in fungi. Fungal Genet Biol. 2008;45(6):791–802.
Lidgard S, Crane PR. Quantitative analysis of early angiosperm radiation. Nature. 1988;331:344–6.
Wing SL, Boucher LD. Ecological aspects of the cretaceous flowering plant radiation. Annu Rev Earth Planet Sci. 1998;26:379–421.
Soltis PS, Soltis DE. The origin and diversification of angiosperms. Am J Bot. 2004;91:1614–26.
Crepet WL. The fossil record of angiosperms: requiem or renaissance? Ann Mo Bot Gard. 2008;95:30–3.
Farrell BD. “Inordinate fondness” explained: why are there so many beetles? Science. 1998;281:555–9.
Gaunt MW, Miles MA. An insect molecular clock dates the origin of the insects and accords with palaeontological and biogeographic landmarks. Mol Biol Evol. 2002;19:748–61.
Labandeira CC, Sepkoski JJ Jr. Insect diversity in the fossil record. Science. 1993;261:310–5.
Brusca RC, Brusca GJ. Invertebrates. Sunderland: Sinauer Associates; 1990.
Visscher H, Brinkhuis H, Dilcher DL, Elsik WC, Eshet Y, Looy C. The terminal Paleozoic fungal event: evidence of terrestrial ecosystem destabilization and collapse. Proc Natl Acad Sci. 1996;93(5):2155–8.
Barron GL. Predatory fungi, wood decay, and the carbon cycle. Biodiversity (Ottawa). 2003;4(1):3–9.
Lei W, Zhang G, Peng Q, Liu X. Development of Ophiocordyceps sinensis through plant-mediated interkingdom host colonization. Int J Mol Sci. 2015;16(8):17482–93.
Humber RA. Evolution of entomopathogenicity in fungi. J Invertebr Pathol. 2008;98:262–6.
Deighton FC, Pirozynski KA. Microfungi V. More hyperparasitic hyphomycetes. Mycol Pap. 1972;128:103–10.
Seifert K. Plectosporium tabacinum. Fungi Canadensis 333. Can J Plant Pathol. 1996;18:309–11.
Petch T. Notes on entomogenous fungi. Trans Br Mycol Soc. 1942;25(3):250–65.
Balazy S. Notes on Hirsutella aphidis. Trans Br Mycol Soc. 1985;85(4):752–6.
Minter DW, Brady BL. Mononematous species of Hirsutella. Trans Br Mycol Soc. 1980;74(2):271–82.
Fang HM, Tan SM. A new species of Hirsutella, H. leizhouensis fang & tan. Acta Mycol Sin. 1992;11(1):28–31.
Samson RA, Evans HC. Taxonomic status of Didymobotryopsis (Hyphomycetes) and description of a new Hirsutella species. Mycol Res. 1991;95(7):887–8.
Aoki K, Nakasato Y, Fujimoto I, Suzuki H. Studies on the new fungous parasites of silkworms, Bombyx mori L. IV. Hirsutella satumaensis Aoki sp. nov. Bull Sericult Exper Stat (Suginami-ku, Tokyo). 1957;14:471–86.
Xiang MC, Yang EC, Xiao QM, Liu XZ, Chen SY. Hirsutella vermicola sp. nov., a new species parasitizing bacteria-feeding nematodes. Fungal Divers. 2006;22:255–66.
Zou X, Zhou YM, Liang ZQ, Xu FL. A new species of the genus Hirsutella with helical twist neck of phialides parasitized on Tortricidae. Mycosystema. 2016;35(7):807–13.
Yu LQ, Zou X, Zhang J, Han YF, Liang ZQ. A new Hirsutella species parasitic on grub larvae (Coleoptera). Microbiol China. 2017;44(06):1370–6.
We acknowledge Bing Yang for assisting in drawing.
This study was supported by the National Natural Science Foundation of China (No. 31360031, 31660150 and 31460010) for design of the study and specimens collection; the Major Project of Guizhou Province, China (Qian Ke He Major Project  3022–07) for analysis data; the Youth Science and Technology Talent Project of Guizhou Province  5617 for writing the manuscript.
Availability of data and materials
The holotype GZUIFR-lgs2 and asexual strain GZUIFR-hir100812 of Ophiocordyceps retorta belong to the Institute of Fungal Resources of Guizhou University (GZAC). For the use and treatment of specimens in this paper, we have been approved by GZAC and complied with its guidelines. All data generated or analyzed during this study are included in this published article and its supplementary information files. The datasets used and/or analysed during the current study are available from the corresponding author on reasonable request.
Ethics approval and consent to participate
Consent for publication
The authors declare that they have no competing interests.
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Table S1. GenBank accession numbers for sequences used in the phylogenetic analysis of Hirsutella (asexual morph). Table S2. Morphological comparison among Ophiocordyceps retorta and its related species. (DOC 86 kb)