Biodiversity of Basidiomycetous Yeasts Associated with Cladonia rei Lichen in Japan, with a Description of Microsporomyces cladoniophilus sp. nov

For more than a century, lichens have been used as an example of dual-partner symbiosis. Recently, this has been challenged by the discovery of various basidiomycetous yeasts that coexist in multiple lichen species, among which Cladonia lichens from Europe and the United States were discovered to be highly specifically associated with the basidiomycetous yeast of the family Microsporomycetaceae. To verify this highly specific relationship, we investigated the diversity of basidiomycetous yeasts associated with Cladonia rei, a widely distributed lichen in Japan, by applying two approaches: yeast isolation from the lichen thalli and meta-barcoding analysis. We obtained 42 cultures of Cystobasidiomycetous yeast which were grouped into six lineages within the family Microsporomycetaceae. Unexpectedly, although the cystobasidiomycetes-specific primer was used, not only the cystobasidiomycetous yeasts but species from other classes were also detected via the meta-barcoding dataset; in particular, pucciniomycetous yeasts were found at a high frequency in some samples. Further, Halobasidium xiangyangense, which was detected in every sample with high abundance, is highly likely a generalist epiphytic fungus that has the ability to associate with C. rei. In the pucciniomycetous group, most of the detected species belong to the scale insect-associated yeast Septobasidium genus. In conclusion, even though Microsporomyces species are not the only yeast group associated with Cladonia lichen, our study demonstrated that the thalli of Cladonia rei lichen could be a suitable habit for them.


Introduction
Symbioses are widespread and essential to the diversity of our ecosystems. Lichen represents a successful model of the iconic symbiosis of stable dual mutualism of a single fungus and a photosynthesizing symbiont. With their long macro-evolutionary history, lichens can inhabit any environment or habitat and currently dominate nearly 10% of the Earth's terrestrial ecosystem [1,2]. Among the lichen-forming fungi, Lecanoromycetes, as one of the largest clades, includes more than 80% of the lichen-forming species [3]. Belonging to this class, Cladonia is one of the most species-rich and morphologically distinctive genera of lichen-forming fungi, with a worldwide distribution [4].
Due to recent discoveries, lichens have been considered not only relevant to dualpartner symbiosis but also part of the complex ecosystems that are inhabited by other microscopic organisms, such as prokaryotes and fungi [5]. As lichens appeared 600 million years ago [6], the existence of secondary fungal species in Early Devonian (~415 million years ago) J. Fungi 2023, 9,473 2 of 17 lichen fossils [7] indicates that this association might have existed throughout the evolutionary history of lichen. These coexistences have been detected in various lichen species [8][9][10]. As they have been mostly considered parasites or lichenicolous fungi [11], and a few as endolichenic fungi [12], and due to their rare isolation, the functions, influence, and roles on communities of lichens have been neglected or underestimated. The recent discovery by Spribille et al. [13] shed new light on this issue by proving the presence of basidiomycetous yeasts in the lichen cortex and their contribution to different phenotypes of two genetically inseparable Bryoria lichens. In the same study, the yeasts of the class Cystobasidiomycetes were detected in the cortex of various macro-lichen taxa, suggesting that the yeasts may represent the third essential constituent of lichen symbiosis. Multiple studies conducted to infer the diversity of lichen-associated basidiomycetous yeasts have delivered conflicting results. Although some supported the association between lichens and basidiomycetous yeasts [13][14][15], others argued against this association [16,17]. Spribille et al. [13], upon screening the lichen-associated yeasts from a wide range of lichen species, detected a specific group of basidiomycetous yeasts, Cyphobasidiales (Pucciniomycotina, Basidiomycota), associated with multiple species, but the yeasts detected from Cladonia lichens were grouped into one lineage within the Microsporomycetaceae family (Cystobasidiomycetes, Pucciniomycotina, Basidiomycota). Later, these unknown yeasts were discovered in a widespread species of this genus, despite their reproductive strategies [14,18]. This showed the high specificity of association with basidiomycetous yeasts of Cladonia lichen, while the reproductive and dispersal strategies were shown to be the key factors shaping photobiont diversity in this lichen group [18]. Surveys of these Cladonia-associated yeasts have been lacking in most of Asia, where several species from the same family have been discovered recently [19].
Most studies that have attempted to discover the biodiversity of lichen-associated basidiomycetous yeasts have utilized one or two approaches that might lead to bias. In some studies, Sanger sequencing, which provides high confidence for detected taxa from sequences, was utilized to determine the lichen-associated yeasts [16], but this approach is not sensitive enough for low abundance symbionts, and the probability of detection of certain species significantly depends on the biodiversity of the community within the lichens [20]. Although isolation from the thallus fragments method seems to have provided decent evidence of the existence of lichen-associated yeasts, it is highly limited by such factors as the type of culture media used for isolation and the skills of technicians [21,22]. Further, the contamination rate of cultures from Cladonia lichens due to their natural habitat [22] increases the difficulty of revealing the true diversity of the yeast community within the thalli of these lichens. Given advances in sequencing technologies, the metabarcoding approach with new emerging high-throughput sequencing techniques seems to be a powerful tool for detecting the true diversity of the basidiomycetous yeast community within lichen thalli [23][24][25][26]. However, the meta-barcoding method also suffers numerous biases, so it requires affirmation from other methods [27][28][29].
Here, we focused on the biodiversity of basidiomycetous yeasts associated with Cladonia rei, which is widely distributed in Japan, using multiple approaches to infer the existence of lichen-associated yeast species and determine whether there is a high specificity in the third symbionts of Cladonia lichen in the region. The yeasts were isolated from multiple lichen specimens collected from various locations. The basidiomycetous yeasts were further screened via meta-barcoding analysis of conserved genomic sequences, amplified with the basidiomycete-specific primers designed by Spribille et al. [13].

Sampling
Fifteen specimens of Cladonia rei thallus were collected from two locations in Chiba prefecture, Japan, from March 2017 to November 2018 ( Table 1). The species was confirmed via morphological and molecular methods. The collected thallus was frozen and stored at −20 • C. Strains were isolated from the thallus using the method described by Yamamoto et al. [22]. The thalli were inspected for external signs of discoloration and parasitism. For each isolation, around 10 podetia were selected and washed with running water for at least 1 h and then with 7 mL 1% sodium hypochlorite. The washed podetia were rinsed with sterilized distilled water thoroughly several times and then ground into small pieces using a mortar and pestle. Finally, the thallus fragments with sizes between 150 and 500 µm were selected using two nylon sieve meshes and then placed onto Malt/Yeast extract (YM) medium [30].
Yeasts grown from the thallus fragments were isolated into axenic cultures and were then maintained on YM medium at 25 • C. Morphological, physiological, and biochemical characteristics were examined following the methods of Kurtzman et al. [31]. We tested for ballistoconidia using the inverted plate method [32]. The glass slide containing the discharged spores was removed for examination under the microscope after 3-10 days. The potential sexual cycles were investigated using the Dalmau plate method on corn meal agar (CMA), YM agar (YMA), and potato-dextrose agar (PDA), according to Kurtzman et al. [31]. A loopful of cells of each strain was streaked on an agar plate and incubated at 17 • C for 1 month, and the cultures were examined microscopically every 2 weeks.

Sequencing and Phylogenetic Analyses
The total DNA of C. rei thallus and the yeast cultures was extracted by the method of Nakada et al. [33]. The ITS rDNA sequences, which were amplified with ITS4 and ITS5 primer, were used to confirm the species of lichen thallus [34]. For the isolated yeasts, the ITS region and partial LSU rDNA were amplified with ITS4 and ITS5; LR0 and LR5 primers were used, respectively [34,35]. PCR amplification for ITS began with denaturation at 95 • C for 5 min, followed by 35 cycles of denaturation at 95 • C for 30 s, annealing at 55 • C for 1 min, and elongation at 72 • C for 1 min, finished with extension at 72 • C for 5 min. PCR amplification was carried out for LSU denaturation at 95 • C for 5 min, which was followed by 35 cycles of denaturation at 95 • C for 30 s, annealing at 55 • C for 1 min, and elongation at 68 • C for 2 min, and finished with extension at 72 • C for 5 min. The PCR products were purified using a FastGene TM Gel/PCR extract kit (Nippon Genetics, Tokyo, Japan). The forward and reverse strands of the purified PCR products were amplified using the same primers with a BigDye TM Terminator v3.1 Cycle Sequencing Kit (Applied Biosystems, Waltham, MA, USA), following purification with Sephadex G-50. Sanger sequencing was conducted with a 3130xl Genetic Analyzer and a 3500 Genetic Analyzer (Applied Biosystems). The final sequences were assembled using Genetyx v. 13 (GENETYX, Tokyo, Japan).
To infer the phylogenetic position of our strains within the class Cystobasidiomycetes, we conducted a phylogenetic analysis on the basis of the phylogenetic results from Li et al. [19] andČernajová and Škaloud [14]. Sequences of the ITS region and LSU rDNA of species in the Cystobasidiomycetes class retrieved from GenBank, together with the sequences of the obtained cultures (Table S1), were aligned using MAFFT using the Q-INS-I method; each region was aligned separately. Gblocks v. 0.91b was used to remove the ambiguously aligned region [36]. Substitution models were estimated with Akaike Information Criterion (AIC) using JModelTest v.2.1.4 [37]. The model GTR+ I + G was selected for both LSU and ITS. Our cultures all clustered within the family Microsporomycetaceae (see Results). Thus, we reconstructed the phylogeny of this family with the currently accepted species [19] and our cultures. Bannoa bischofiae and three Erythobasidium species (Erythrobasidiaceae, Cystobasidiomycetes, Pucciniomycotina, Basidiomycota) were selected as outgroups. The sequences were processed as described above. We used RAxML-HPC v.8 to construct the maximum likelihood (ML) phylogeny [38] and perform 1000 rapid bootstrap replicates. The output trees were visualized using FigTree v.1.4.4 [39]. The final presentation was undertaken in free RStudio software [40,41] with the packages ggtree [42] and treeio [43].

Meta-Barcoding Sequencing and OTU Identity Analysis
Seven specimens were selected for meta-barcoding ( Table 1). The yeast ITS rDNA from the DNA of C. rei was amplified using the Cystobasidiomycetes-specific primer ITS_symrho_2F and LR0_symrho_R designed by Spribille et al. [13]. PCR products were checked on 2% agarose gel and quantified using Qubit 2.0. The libraries were generated with NEBNext ® Ultra™ DNA Library Prep Kit for Illumina ® (San Diego, CA, USA). Sequencing was completed with the paired-end protocol 2 × 300 bp on a PE300 MiSeq instrument at Novogene, Tokyo, Japan. All output reads are available in the Short Read Archive under accession PRJNA765160. Raw paired-end reads were merged using FLASH v.1.2.7 [44]. The adapters and barcodes were trimmed from sequences using fastx_trimmer, implemented in FASTX-Toolkit [45]. The operational taxonomic units (OTUs) were identified by clustering the trimmed sequences based on the sequence similarity threshold (97%) using USEARCH v.11 [46]. Each OTU's taxonomy assignment was performed by blasting their reference sequences [47] against the full "UNITE + INSD" dataset for fungi version 2021-05-10 [48] with a 10 × 10 −50 e-value threshold. The OTUs that were identified as cystobasidiomycetous yeasts and showed a relative abundance in any sample higher than 1% were further included in the phylogenetic analysis, with ITS region data to determine approximately their position within the class using the same method as above. To visualize the yeast community composition, we used the free RStudio software [40,41] with the ggplot2 pack-age [49]. To avoid an over-detailed graph, the OTUs with a relative abundance of less than 1% were grouped.

Isolated Cultures
Forty-two strains were cultured successfully from eight C. rei specimens and identified by sequencing the ITS and LSU rDNA ( Table 2). Our phylogenetic analyses showed a consistent topology and were generally congruent with most of the major groups from [19] ( Figure 1). Furthermore, the phylogenetic analyses also grouped Lichenozyma species with Microsporomyces species and our cultures. Hence, we agree with Li et al. (2020) that the Lichenozyma genus should be considered synonymous with Microsporomyces. Although the phylogenetic analysis of the Cystobasidiomycetes class was slightly incongruent in topology with the analysis on the family Microsporomycetaceae, both indicated that all of the isolated cultures belonged to Microsporomycetaceae, with high statistical support. Furthermore, the phylogenetic analysis of Microsporomycetaceae showed that our cultures could be grouped into six distinct lineages ( Figure 2). The first lineage (Microsporomyces cladoniophilus) comprised most of the strains (24 strains) that appeared to be a sister of Microsporomyces bloemfonteinensis and the second lineage (Sp.2), with two isolated cultures 211_15 and 216_16 as their sister, with high support. The rest of the cultures were divided into two sister pairs, in which one was a sister of Microsporomyces hainanensis, but their relationships were not well supported. Furthermore, our phylogenetic results were congruent with Li et al. [19] in terms of the placement of the newly described Lichenozyma genus [14] (Microsporomyces pisutianus in Figures 1 and 2) inside the Microsporomyces genus, with high statistical support, so we proposed Lichenozyma as a synonym of Microsporomyces.

Meta-Barcoding Results
A total of 3,809,740 reads were generated and clustered into 348 OTUs that mostly belonged to the Cystobasidiomycetes and Pucciniomycetes classes ( Figure 3A). Within the Cystobasidiomycetes class, most of the OTUs with high relative abundance belonged to the two families Cystobasidiaceae and Microsporomycetaceae ( Figure 3B). Based on the relative abundance of meta-barcoding reads, the OTU1, which occurred most abundantly in every sample, belonged to the Cystobasidiaceae family, along with two other OTUs (OTU9 and OTU81) ( Figure 3B). We detected a high diversity of Microsporomycetaceae from our samples (six out of seven samples) since various OTUs were classified as belonging to the Microsporomycetaceae family. There were only three families from the Pucciniomycetes class detected from our data, and most of the OTUs were classified as Septobasidiaceae ( Figure 3C). To infer the detailed phylogenetic positions of these OTUs, we conducted phylogenetic analyses on the OTUs that were identified as Cystobasidiomycetes and Pucciniomycetes fungi using the same method as described above. From the phylogenetic tree ( Figure S1), OTU 14, 40, 9, and 1 were identified as Buckleyzyma aurantiaca, Erythrobasidium hasegawianum, Occultifur mephitis, and Halobasidium xiangyangense. This was consistent with the results from blasting against the "UNITE + INSD" dataset. Further, thirteen OTUs were grouped with the Microsporomyces species or our cultures, which suggested they belonged to these genera, while the remaining eight OTUs (19,20,21,24,25,27,33,42) were unable to be classified. For the OTUs that belonged to the Pucciniomycetes class, most were identified as Septobasidum species, which are known for their association with scale insects (Figure S2), except OTU 41, which was classified as Eocronartium species ( Figure 3C). to be classified. For the OTUs that belonged to the Pucciniomycetes class, most were identified as Septobasidum species, which are known for their association with scale insects ( Figure S2), except OTU 41, which was classified as Eocronartium species ( Figure 3C).
MycoBank Number: MB846812. With the evidence from both the culture of thallus fragments and DNA meta-barcoding, we suggest a novel Microsporomyces species which is associated with C. rei lichen.
Holotype: The strain 211_1 was isolated by H. Otake from the thallus of Cladonia rei collected in Chiba, Japan (N35.18731, E140.11192) in October 2017. It was chosen as the holotype. This culture is permanently preserved in a metabolically inactive state under the number NBRC 115437 and CBS 17989. Strains 109_1 and 226_20 were selected as the paratype. These cultures are permanently preserved in a metabolically inactive state under the number NBRC 115348 and NBRC 115439, respectively.
MycoBank Number: MB846812. With the evidence from both the culture of thallus fragments and DNA meta-barcoding, we suggest a novel Microsporomyces species which is associated with C. rei lichen.
Holotype: The strain 211_1 was isolated by H. Otake from the thallus of Cladonia rei collected in Chiba, Japan (N35.18731, E140.11192) in October 2017. It was chosen as the holotype. This culture is permanently preserved in a metabolically inactive state under the number NBRC 115437 and CBS 17989. Strains 109_1 and 226_20 were selected as the paratype. These cultures are permanently preserved in a metabolically inactive state under the number NBRC 115348 and NBRC 115439, respectively.
Diagnosis: Phylogenetic analysis indicated the closest species of M. cladoniophilus was M. bloemfonteinensis. M. cladoniophilus differed from M. bloemfonteinensis by 22-25 nucleotides (~5%) and 33-36 nucleotides (~10%) mismatched in the LSU and ITS regions, respectively. Physiologically, M. cladoniophilus differs from its closely related species, M. bloemfonteninensis, in its ability to assimilate sucrose, glycerol, D + galactose, potassium nitrate, and L-lysine, and its inability to assimilate melibiose, D-melezitose, inulin, D-ribose, DL-lactic acid, succinic acid salt, and inositol, as well as its inability to grow in vitamin-free medium, and its ability to produce urease. medium, and its ability to produce urease.
Host range and distribution: Currently only reported from Cladonia rei lichen from Chiba Prefecture, Japan.
Culture characteristics: On YMA, after 10 days at 25 °C, colonies were around 1 cm in diameter, the surface was smooth or rough, orange to dark orange colored, depressed in the center, and the margin was undulating or entire ( Figure 4A,B). The cells were ovoid and ellipsoidal, 2.1-5.4 µm × 3.0-10.3 µm (measured from 10 random cells, length/width ratio 1.45-2.54), and single, and budding was polar ( Figure 4C). In Dalmau plate culture on YM, CM, PDA, and YMA, pseudohyphae were not formed, and sexual structures were not observed. Ballistoconidia were not produced.  Host range and distribution: Currently only reported from Cladonia rei lichen from Chiba Prefecture, Japan.
Culture characteristics: On YMA, after 10 days at 25 • C, colonies were around 1 cm in diameter, the surface was smooth or rough, orange to dark orange colored, depressed in the center, and the margin was undulating or entire ( Figure 4A,B). The cells were ovoid and ellipsoidal, 2.1-5.4 µm × 3.0-10.3 µm (measured from 10 random cells, length/width ratio 1.45-2.54), and single, and budding was polar ( Figure 4C). In Dalmau plate culture on YM, CM, PDA, and YMA, pseudohyphae were not formed, and sexual structures were not observed. Ballistoconidia were not produced.

Microsporomyces pseudomagnisporus
D-Glucuronic acid + + n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a 2,3-Butanediol − +(l/w)/− n/a n/a n/a n/a n/a n/a n/a D + Glucono-1,5-lactone + + n/a n/a n/a n/a n/a n/a n/a Xylitol + +(w) n/a n/a n/a n/a n/a n/a n/a Ribitol + +(w) w + − − + − n/a L-Arabitol − − n/a n/a n/a n/a n/a n/a n/a 1,2-Propanediol − − n/a n/a n/a n/a n/a n/a n/a 2-keto-D-gluconate (2-Oxoglutaric acid) − + n/a n/a n/a n/a n/a n/a n/a Growth at 22 • C + + − + + n/a n/a n/a n/a Growth at 25 • C + + − + + + + + + n/a n/a n/a n/a Growth at 32 • C − − − w − n/a n/a n/a n/a Diazonium blue B reaction + + + + + n/a n/a n/a + (2) Microsporomyces pisutianus (Černajová and Škaloud) N.H. Nguyen, P.T. Nguyen, Imanishi and K. Shimizu, comb. nov.

Discussion
With a focus on investigating the diversity of the lichen-associated Cystobasidiomycete yeasts in C. rei lichen, we discovered multiple species belonging to this class from the thalli of this lichen. With multiple isolated strains, a new Cystobasidiomycete species inhabiting the thallus of C. rei was discovered and confirmed by meta-barcoding analysis.
The high level of specificity of Cystobasidiomycetes yeasts was speculated by previous studies [13,14,52]. For instance, it was thought the Microsporomycetaceae family might be specifically associated with the Cladonia lichens and Cystobasidiales with the lichens of the Parmeliaceae family. Our results were in agreement with the specific partnership of Cladonia lichen, but the relationship is not strictly specialized. Alongside Microsporomyces cladoniophilus sp. nov., which was detected in most of the samples, we further isolated five unknown species that belong to the Microsporomyces genus. These species were also detected by the meta-barcoding analyses, which confirmed the diversity of this yeast genus within the thallus of Cladonia lichen [15,16]. Despite the new yeast species not being isolated from the meta-barcode sequenced specimens, the sequences of these species were still detected in most of the samples. As the meta-barcoding specimens were collected in early winter, whereas the others were collected in early summer, the difference in climate may have contributed to this result [53,54].
From the meta-barcoding reads, we found that another Cystobasidiomycete yeast, Halobasidium xiangyangense, existed in every specimen, with the highest frequency of reads among the other OTUs and our isolated strains. While Halobasidium xiangyangense was first described from an isolated culture taken from pickling sauce [55], this species is likely an epiphyte fungus, as it has since been isolated from plant materials worldwide [56][57][58][59]. As generalist species are usually not constrained by the environment and can occupy a wide range of habitats [60,61], this epiphyte fungus is highly likely to be a generalist with the ability to associate with Cladonia lichen. Unexpectedly, at least one Septobasidium species, a scale insect-associated fungi [62,63], was detected in each specimen. There have not been many studies on this fungal group, but their existence in the thalli of lichen might indicate that they are also a generalist species.
Wang et al. [64] first described the family Microsporomycetaceae on the basis of the molecular data of five species. To accommodate the yeasts that have been reported as inhabitants of Cladonia lichens in Europe and the United States [13,14], the genus Lichenozyma has been proposed. Based on multiple loci phylogeny, Li et al. [19] objected to this proposal and suggested this genus is a synonym for Microsporomyces. Our phylogenetic results agreed with theirs and further indicated that the diversity of the genus Microsporomyces remains underestimated, with several unknown isolated strains grouped in this genus.
In our study, the effort to directly detect Cystobasidiomycete yeasts from the specimens with the Cystobasidiomycetes-specific primer was not successful; this was probably influenced by the biodiversity of the co-existing cystobasidiomycetous yeasts within the thallus [20]. Our results from the meta-barcoding analysis showed more than two Cystobasidiomycete species associated with C. rei lichen, and neither of them had an absolute abundance (>70% of total reads) in all samples determined. In conclusion, our study re-confirmed the existence of yeast from the Microsporomycetaceae family within the thalli of Cladonia rei lichen, and the detected H. xiangyangense further highlighted the diversity and complexity of the yeast community associated with this lichen.
Supplementary Materials: The following supporting information can be downloaded at: https:// www.mdpi.com/article/10.3390/jof9040473/s1, Table S1: List of sequences retrieved from GenBank used for phylogenetic analyses; Table S2: The OTUs list with taxonomic identification and their read counts in each sample.; Figure S1. Phylogeny tree of Cystobasidiomycetes obtained via the maximum likelihood analysis of the combined sequences of the LSU rDNA and ITS regions. The OTUs from the meta-barcoding sequences were marked with red; Figure S2 Phylogeny tree of Septobasidium obtained by maximum likelihood analysis of the ITS sequences. The OTUs from the meta-barcoding sequences were marked with red color.   (Table S1). The meta-barcoding raw reads were deposited on the Sequence Read Archive (SRA) data in NCBI under accession number SRR16004934.