﻿Additions to Hohenbuehelia (Basidiomycota, Pleurotaceae): two new species and notes on H.tristis from northern Thailand

﻿Abstract Two new species and a first geographical record of Hohenbuehelia are described from Thailand. Macroscopic and microscopic descriptions with photoplates, as well as a multigene phylogeny are provided. Hohenbueheliaflabelliformissp. nov. is recognised by large flabelliform basidiomata, densely villose yellowish-white pileus with white hairs near the point of attachment, basidiospores that mostly are ellipsoid in front view and phaseoliform in side view, the absence of cheilocystidia, and a trichoderm pileipellis. Hohenbuehelialageniformissp. nov. is characterised by fleshy basidiomata, velutinous pileus with whitish hairs near the point of attachment and the margin, elsewhere pale greyish-yellow and with only sparse white hairs, pale brown to light brown and mucilaginous context, subglobose basidiospores, lageniform cheilocystidia, an ixotrichoderm pileipellis, and the absence of pileoleptocystidia. Hohenbueheliatristis is characterised by small creamy-white, spathuliform basidiomata that are larger than the type subspecies, minutely pubescent pileus with tiny greyish hairs that disappear when mature, leaving the surface glutinous, faintly translucent and shiny, ellipsoid to sub-ellipsoid basidiospores, lecythiform to sublageniform cheilocystidia, and an ixotrichoderm pileipellis. Hohenbueheliatristis is recorded for Thailand for the first time. Based on the polymorphism observed in part of the nrLSU gene, the presence of two divergent lineages within H.tristis is discussed.


Introduction
Hohenbuehelia Schulzer belongs to the family Pleurotaceae Kühner in the order Agaricales Underw. In former studies, the asexual stages of Hohenbuehelia species were separately placed in the genus Nematoctonus Drechsler (Drechsler 1941; Thorn and Barron 1986). Following the One Fungus = One Name nomenclatural rule, both the asexual and sexual stages were placed under Hohenbuehelia (Taylor 2011;McNeill et al. 2012;Thorn 2013), with H. petaloides (Bull.) Schulzer as the type species. Currently, 126 taxon names are listed under Hohenbuehelia in Index Fungorum (http://www.indexfungorum.org/), for 50 accepted species (He et al. 2019;Wijayawardene et al. 2022). The typical characteristics of this genus are pleurotoid, gelatinous basidiomata, white basidiospores with a germ pore, lecythiform cheilocystidia (if present) and thickwalled metuloid pleurocystidia (Stevenson 1964; Thorn and Barron 1986;Corner 1994;Silva-Filho and Cortez 2017;Holec and Zehnálek 2020). Hohenbuehelia and Pleurotus (Fr.) P. Kumm. have some similar morphological characters. However, Hohenbuehelia is distinguished by the synapomorphic gelatinous layer in the context under the pileipellis, which is absent in Pleurotus (Mentrida 2016). Most Hohenbuehelia species are decomposers and widely distributed in temperate and tropical areas (Laessøe and Petersen 2019). Hohenbuehelia species have been found growing on dead branches, decayed wood, logs, and sometimes on the bark of living trees or on herbaceous stems (Holec and Zehnálek 2020).
A few Hohenbuehelia species have been reported as edible. However, they have very low culinary value (Cruz and Gándara 2005). This genus is a good source of polyphenols and polysaccharides (Li et al. 2012;Li et al. 2017;Wang et al. 2018;Wang et al. 2019). Bioactive compounds extracted from some Hohenbuehelia species were found to have antioxidant properties as well as antitumor and antiviral activities (Ji et al. 2012;Sandargo et al. 2018a). Furthermore, the new derivatives thiopleurotinic acid A and thiopleurotinic acid B from extracts of H. grisea (Peck) Singer (strain MFLUCC 12-0451) were found to exhibit cytotoxicity towards a mouse fibroblast cell line, as well as antimicrobial activities (Sandargo et al. 2018a). Another compound, 4-hydroxypleurogrisein, was shown to inhibit hepatitis C virus infectivity in mammalian liver cells (Sandargo et al. 2018b). Hohenbuehelia sp. ZW-16 has been used for bioethanol production (Liang et al. 2013). Thorn et al. (2000) also found that the mycelia of the asexual stage of some Hohenbuehelia species are able to produce adhesive knobs that can capture nematodes. The diversity of bioactive compounds from Hohenbuehelia species and their potential applications underline the importance of detailed taxonomic study of this genus (Shipley et al. 2006;Bohni et al. 2013Sandargo et al. 2018a. Thailand has a high mushroom diversity with many new species yet to be discovered (Thongbai et al. 2018;Vadthanarat et al. 2021). Only four Hohenbuehelia species have been recorded from Thailand, namely H. grisea, H. panelloides Høiland, H. petaloides and H. reniformis (G. Mever & Fr.) Sing. (Chandrasrikul et al. 2011;Sandargo et al. 2018a). However, most of those reports did not provide detailed morphological descriptions nor molecular data in order to confirm the identifications (H. grisea was identified, based on ITS sequences only, without morphological data). In this study, during the survey of pleurotoid mushrooms in northern Thailand, several collections of Hohenbuehelia were obtained and studied. Based on morphological and phylogenetic results, two new species, H. flabelliformis and H. lageniformis, and the first record of H. tristis are described herein.

Sample collection and morphological study
The mushroom specimens were collected during the rainy season in 2019 and 2020 from Chiang Mai and Chiang Rai Provinces, in northern Thailand. The fresh basidiomata were photographed in situ. Details including collecting date, locality, habitat and ecology of the surroundings, were noted. The specimens were wrapped in aluminium foil or kept in plastic boxes and brought back to the lab for morphological descriptions.
Macromorphological descriptions were done, based on the fresh specimens and colour codes were given following the colour charts of Kornerup and Wanscher (1978). The specimens were dried in a hot air dryer at 50 °C until the samples were completely dried and then kept separately in zip-lock plastic bags. The specimens were deposited in the Herbarium of Mae Fah Luang University (MFLU). Micromorphological characters were observed from the dried specimens. A razor blade was used to make thin sections of the specimens and these were mounted on slides in water, 5% potassium hydroxide (KOH) solution or 1% ammoniacal Congo red. The microcharacters were studied and photographed using a compound microscope Nikon Eclipse Ni. Freehand drawings were made for the microscopic features. Fifty basidiospores per basidioma were measured in side view. The notation [x/y/z] denotes the number of basidiospores (x) measured from the number of basidiomata (y) of the number of collections (z). At least 25 basidia, pleurocystidia, cheilocystidia, and pileipellis hyphae were observed and measured. The dimensions of microscopic structures are presented in the following format: (a-)b-c-d(−e), in which c represents the average, b the 5 th percentile, d the 95 th percentile, and minimum and maximum values a and e are shown in parentheses. Q, the length/width ratio, is presented in the same format. Facesoffungi numbers and MycoBank numbers are provided for each new species.

DNA extraction and sequencing
Genomic DNA was extracted from the dried herbarium specimens using the Biospin Fungus Genomic DNA Extraction Kit (Bioer Technology, Hangzhou, China), following the manufacturer's instructions. The ITS region and parts of the nrL-SU and tef1 genes were amplified by a polymerase chain reaction (PCR) and sequenced. The following primers were used: ITS1-F and ITS4 for ITS (White et al. 1990;Gardes and Bruns 1993), LR0R and LR5 for nrLSU (Vilgalys and Hester 1990;White et al. 1990) and EF1-983F and EF1-1567R for tef1 (Rehner and Buckley 2005). The PCR cycling for ITS and LSU was as follows: 3 min at 94 °C; 35 cycles of 30 s at 94 °C, 30 s at 52 °C, 1 min at 72 °C; 10 min at 72 °C. For tef1, the following programme was used: 5 min at 95 °C; 35 cycles of 1 min at 94 °C, 2 min at 52 °C, 1.5 min at 72 °C; 10 min at 72 °C. The PCR-amplified products were purified and sequenced in forward and reverse directions, using PCR primers by Sangon Biological Engineering Technology & Services (Shanghai).

Sequence alignment and phylogenetic analyses
Sequence reads were checked using Bioedit Sequence Alignment Editor version 7.0.9.0 and assembled using SeqMan (DNAstar, Madison, WI, USA). Each sequence was blasted using the Basic Local Alignment Search Tool (BLAST) against the National Center for Biotechnology Information (NCBI) database (http://www. ncbi.nlm.nih.gov/genbank/) to check that it was from the correct genus and not from contamination, as well as to find the closest matches. Newly-generated sequences were deposited in GenBank. All sequences (Table 1) including the outgroup were retrieved and aligned using MAFFT v.7 (Katoh et al. 2017) on the online server (http://mafft.cbrc.jp/alignment/server/). For tef1, introns were delimited by comparing with the amino acid sequence of a reference sequence and locating the GT/AG motifs of the splicing sites and removed for further analyses. The ITS and LSU alignments were trimmed separately using TrimAl to eliminate ambiguously aligned positions (Capella-Gutiérrez et al. 2009). The length of each character set was: ITS1+ITS2 = 445; LSU+5.8S = 995; TEF1 (exons) = 438. After checking for supported conflicts (BS ≥ 70%) between single-gene Maximum Likelihood (ML) phylogenies, a concatenated three-gene dataset was assembled.
All phylogenetic analyses were done on the CIPRES Science Gateway version 3.3 web server (Miller et al. 2010), accessed at https://www.phylo.org/. For both Maximum Likelihood and Bayesian analyses, a mixed-model (partitioned) scheme was used, with the alignment divided in the following three character sets: ITS1+ITS2, LSU+5.8S, tef1. Maximum Likelihood phylogenetic inference was performed using RAxML-HPC2 version 8.2.12 (Stamatakis 2006) on XSEDE. Five Pleurotus species were used as outgroup. For Bayesian analysis, the best-fit substitution models were selected from jModelTest2 version 2.1.6 (Darriba et al. 2012) on XSEDE. The best-fit models were HKY+G for ITS1+ITS2, GTR+I+G for nrLSU+5.8S and SYM+G for tef1. Bayesian analysis was performed in MrBayes version 3.2.7a (Ronquist et al. 2012). Two runs of five chains each were run for 500,000 generations and sampled every 200 generations. The average standard deviation of split frequencies was 0.008720 at the end of the runs. The burn-in phase (25%) was estimated by checking the stationarity in the generation-likelihood plot in Tracer ver.

Phylogenetic analyses
The combined dataset consisted of 39 Hohenbuehelia and five Pleurotus accessions (Table 1). The final alignment, including the gaps, was 1,878 characters long and was deposited in TreeBASE (submission ID 29653). The Bayesian and ML analyses resulted in similar tree topologies; thus, only the ML tree is shown with both Maximum Likelihood bootstrap (BS) values and Bayesian posterior probabilities (PP) (Fig. 1). In the phylogram, H. flabelliformis (MFLU22-0008 and MFLU22-0009) was closely related to H. algonquinensis (RGT 870601/12 UWO) from Canada with high support (90% BS, 1.00 PP). Hohenbuehelia lageniformis (MFLU22-0010 and MFLU22-0012) was closely re-  to the two specimens MFLUCC 12-0451 and HFJAU0029, identified as H. grisea (both only ITS) from GenBank, except for three substitution heteromorphisms in the ITS sequence of MFLU22-0016 (see Table 2). heteromorphisms and were closely related to H. tristis (RV95/214 DUKE and RV95/295 DUKE) from Australia with 92% BS, but low support in the BI analysis. Diagnosis. This species is distinguished from other Hohenbuehelia species by large flabelliform basidiomata, yellowish-white pileus that is densely villose with white hairs longer near the point of attachment, and shorter towards the margin, ellipsoid basidiospores, absence of cheilocystidia, and a trichoderm pileipellis.
Habitat and distribution. Solitary, gregarious to imbricate, on decaying branches in a tropical forests in northern Thailand.
In the phylogenic tree, the most closely-related species to H. lageniformis was H. odorata (voucher TBGT17443). However, the genetic distance between the ITS sequence of H. lageniformis and H. odorata was 4.62% (27/584), which supports the distinction of the two species. Moreover, these two species also show morphological differences (see above). MycoBank No: 332010 Figs 6, 7 Remarks. The following description is based solely on the Thai materials we collected and examined.
Habitat Notes. Two accessions identified as H. grisea, the culture MFLUCC 12-0451 from Thailand and HFJAU0029 from China (unpublished), had the same ITS sequence than H. tristis (MFLU22-0015 and MFLU22-0016) except for three substitution heteromorphisms in the ITS sequence of MFLU22-0016 (see Table 2). The two former sequences retrieved from GenBank have no corresponding morphological descriptions available as evidence. Therefore, these might have been wrongly identified, since ITS sequences are identical to sequences obtained from our collections of H. tristis (except for the heteromorphisms detailed in Table 2). Additional confirmation of the taxonomic identity of our specimens was obtained by comparing the morphology of our specimens with H. grisea which was originally described as Pleurotus atrocoeruleus var. griseus Peck from New York. The latter is distinguished by a greyish to greyish-brown, sparsely tomentose pileus, the lamellae becoming cream-coloured in age (Peck 1891). Hohenbuehelia tristis is characterised by reniform basidiomata, minutely pubescent pileus with greyish hairs that disappear when mature, leaving the surface gelatinous, faintly translucent and shiny, ellipsoid to sub-ellipsoid basidiospores, lecythiform to sublageniform cheilocystidia, and an ixotrichoderm pileipellis. Hohenbuehelia tristis described from New Zealand differs from our collections (MFLU2022-0015 and MFLU2022-0016) by having smaller basidiomata (10-20 × 10-15 mm), smaller (mostly narrower) basidiospores (7 × 3 µm), larger and pseudo-amyloid metuloids (80-90 × 15-20 µm), pileipellis as tufts of parallel larger hyphae (3-8 µm in diam.) (Stevenson 1964). In our phylogenetic analysis (Fig. 1), the Thai accessions of H. tristis formed a monophyletic group with the accessions of H. tristis from New Zealand. The morphological differences we observed between the Thai and New Zealand accessions suggested that they might not be conspecific. Additionally, the single synapomorphic position we observed in the LSU sequence (position 685 in the alignment; see Table 3) is not incompatible with two distinct species, since a genetic distance of  only one substitution can be observed between other closely-related species (e.g. H. tremula and H. longipes; Table 4). However, the LSU sequence of the Thai specimen MFLU22-0016 showed two to three heteromorphisms corresponding to the differences between the other Thai specimens and the New Zealand specimens (Table 3). This suggests either incomplete lineage sorting or that the two lineages can still interbreed. Heteromorphisms were also observed in the ITS sequence of MFLU22-0016, but we could not compare with ITS sequences of materials from New Zealand, which were unavailable. In view of all the heteromorphisms, we decided not to treat the Thai materials as a new species distinct from H. tristis.

Discussion
The Pleurotaceae belong to the Agaricales and comprise the monophyletic pleurotoid genera Pleurotus and Hohenbuehelia (Thorn et al. 2000;Koziak et al. 2007). Hohenbuehelia species have often been misidentified, in part because holotypes are missing or because types of species put in synonymy were not adequately studied (Consiglio et al. 2018b). In the past, most Hohenbuehelia species were introduced, based only on short morphological descriptions (e.g., Peck 1891;Coker 1944;Stevenson 1964). Consiglio (2016Consiglio ( , 2017a and Consiglio and Setti (2017) designated lectotypes, neotypes, and epitypes to clarify older species names or names that lack modern and molecularly-characterised holotypes. For example, the holotype of H. tristis was identified by Stevenson (1964) without any molecular evidence. Later, an nrLSU sequence of H. tristis was obtained for the first time by Moncalvo et al. (2000). The heteromorphisms we observed in the nrLSU and ITS sequences of one of the Thai specimens related to H. tristis suggested that interbreeding may occur between two divergent lineages within H. tristis. Although more data would be needed to confirm it, we hypothesise that those two lineages diverged in geographical isolation (between Southeast Asia and Oceania) and then came in contact in Southeast Asia. This kind of biogeographical history, revealed by DNA sequence variation, have been observed in other Agaricales, for example, Agaricus subrufescens Peck (Chen et al. 2016). Some of the recently-described species were still introduced, based on only single-gene molecular evidence. In this study, we provide multiple-gene sequence data and detailed descriptions supporting the introduction of two new Hohenbuehelia species and a note on H. tristis from Thailand. At present, a total of six Hohenbuehelia species have been reported from Thailand including the ones in this study and three that were previously reported, namely H. panelloides, H. petaloides, and H. reniformis (Chandrasrikul et al. 2011). The report of H. grisea by Sandargo et al. (2018a) has to be excluded since we showed that the correct identification of their material is H. tristis. More studies on Hohenbuehelia are needed to clarify their taxonomy and more new species might be discovered. Table 4. Genetic distance (number of substitutions, excluding heteromorphisms) between LSU sequences of closely-related Hohenbuehelia species.