Species Diversity, Distribution, and Phylogeny of Exophiala with the Addition of Four New Species from Thailand

The genus Exophiala is an anamorphic ascomycete fungus in the family Herpotrichiellaceae of the order Chaetothyriales. Exophiala species have been classified as polymorphic black yeast-like fungi. Prior to this study, 63 species had been validated, published, and accepted into this genus. Exophiala species are known to be distributed worldwide and have been isolated in various habitats around the world. Several Exophiala species have been identified as potential agents of human and animal mycoses. However, in some studies, Exophiala species have been used in agriculture and biotechnological applications. Here, we provide a brief review of the diversity, distribution, and taxonomy of Exophiala through an overview of the recently published literature. Moreover, four new Exophiala species were isolated from rocks that were collected from natural forests located in northern Thailand. Herein, we introduce these species as E. lamphunensis, E. lapidea, E. saxicola, and E. siamensis. The identification of these species was based on a combination of morphological characteristics and molecular analyses. Multi-gene phylogenetic analyses of a combination of the internal transcribed spacer (ITS) and small subunit (nrSSU) of ribosomal DNA, along with the translation elongation factor (tef), partial β-tubulin (tub), and actin (act) genes support that these four new species are distinct from previously known species of Exophiala. A full description, illustrations, and a phylogenetic tree showing the position of four new species are provided.


Species Isolation Resources Location Reference
Exophiala abietophila Silver fir (Abies alba) Norway [21] Exophiala alcalophila Soil, soap container, washing machine, bathwater from households, and human skin Brazil, Denmark, Germany, Japan, and Ukraine [3,75,76] Exophiala angulospora Exophiala spartinae Spartina alterniflora root tissue in saltwater marsh the USA [110] A search involving the keyword "Exophiala" retrieved 481 titles of research articles that had been published over the last 30 years (1992 to 2021) in the Scopus database [112]. The current upward trend associated with the research of Exophiala is expected to continue in the future ( Figure 3A). It has been determined that the majority of applications for Exophiala have been reported in the medical field, accounting for 43.8%, followed by the fields of immunology and microbiology (18.7%), biochemistry and molecular biology (11.4%), agricultural and biological science (10.3%), veterinary medicine (5.7%), and pharmacology and toxicology (2.4%) ( Figure 3B). A search involving the keyword "Exophiala" retrieved 481 titles of research articles that had been published over the last 30 years (1992 to 2021) in the Scopus database [112]. The current upward trend associated with the research of Exophiala is expected to continue in the future ( Figure 3A). It has been determined that the majority of applications for Exophiala have been reported in the medical field, accounting for 43.8%, followed by the fields of immunology and microbiology (18.7%), biochemistry and molecular biology (11.4%), agricultural and biological science (10.3%), veterinary medicine (5.7%), and pharmacology and toxicology (2.4%) ( Figure 3B).
There are 26 Exophiala species (41.3%) that have been reported as causal agents of human diseases . In addition, seven species of Exophiala (11.1%), namely E. angulospora, E. aquamarina, E. cancerae, E. equina, E. pisciphila, E. psychrophila, and E. salmonis, were identified as pathogens of sea creatures. However, the remaining 34 Exophiala species (54.0%) have not been associated with pathogenicity in humans or animals   (Table 1). However, in some previous studies, some Exophiala species have been effectively used in agricultural and biotechnological applications. Examples of these include E. pisciphila, which was able to promote the plant growth of maize by increasing phosphorus absorption, photosynthesis, and tolerance of cadmium [113,114]. Furthermore, by effectively suppressing Fusarium-wilt disease in strawberries, E. pisciphila could be considered a biocontrol agent [109]. In terms of drug discovery, exophillic acid and its derivative compounds derived from Exophiala species have exhibited activity against HIV-1 integrase [115,116]. Importantly, the antimicrobial property of chlorohydroaspyrones and exophilin A produced from Exophiala species has been reported [117,118]. Interestingly, Exophiala has demonstrated the ability to degrade hydrocarbons (e.g., benzene, toluene, and xylene) that can be employed in bioremediation applications [25,119]. Although Exophiala species have been researched in a variety of applications, certain risks still remain. Therefore, further research should be conducted in the future, particularly with regard to the aspects of management and safety. There are 26 Exophiala species (41.3%) that have been reported as causal agents of human diseases . In addition, seven species of Exophiala (11.1%), namely E. angulospora, E. aquamarina, E. cancerae, E. equina, E. pisciphila, E. psychrophila, and E. salmonis, were identified as pathogens of sea creatures. However, the remaining 34 Exophiala species (54.0%) have not been associated with pathogenicity in humans or animals   (Table  1). However, in some previous studies, some Exophiala species have been effectively used in agricultural and biotechnological applications. Examples of these include E. pisciphila, which was able to promote the plant growth of maize by increasing phosphorus absorption, photosynthesis, and tolerance of cadmium [113,114]. Furthermore, by effectively suppressing Fusarium-wilt disease in strawberries, E. pisciphila could be considered a biocontrol agent [109]. In terms of drug discovery, exophillic acid and its derivative compounds derived from Exophiala species have exhibited activity against HIV-1 integrase [115,116]. Importantly, the antimicrobial property of chlorohydroaspyrones and exophilin A produced from Exophiala species has been reported [117,118]. Interestingly, Exophiala has demonstrated the ability to degrade hydrocarbons (e.g., benzene, toluene, and xylene) that can be employed in bioremediation applications [25,119]. Although Exophiala species have been researched in a variety of applications, certain risks still remain. Therefore, further research should be conducted in the future, particularly with regard to the aspects of management and safety. Currently, only three Exophiala species have been identified in Thailand, namely E. dermatitidis, E. jeanselmei, and E. spinifera [43,46,67]. Accordingly, many studies have proposed that Thailand has proven to be a hot spot for novel microfungal species discovery [120][121][122]. During investigations of rock-inhabiting fungi in northern Thailand during the period of 2020 to 2021, we obtained fifteen Exophiala strains that are potentially representative of new species. In the present study, we describe four new species, namely E. lamphunensis, E. lapidea, E. saxicola, and E. siamensis. These four new species were identified based on morphological and molecular data. To confirm their taxonomic status, phylogenetic relationships were determined by analysis of the combined sequence dataset of ITS, nrSSU, tef, tub, and act genes.

Sample Collection and Fungal Isolation
Rock samples were collected from four natural forests located in Lamphun (three sites; 18 •  collected with a sterile chisel, kept in plastic bags, and carried to the laboratory in an ice box. All collected rock samples were processed for the isolation of fungi immediately after reaching the laboratory. Fungi were isolated using the method described by Selbmann et al. [123] with some modifications. Rock samples were washed in 1% sodium hypochlorite for 10 min and rinsed 5 times in sterile water. Fungal isolation was performed by pulverizing the rock samples and sprinkling rock powder onto 2% malt extract agar (MEA; Difco, Le Pont de Claix, France) and dichloran-rose bengal agar (DRBC; Difco, Le Pont de Claix, France) supplemented with chloramphenicol 100 ppm. Plates were incubated at 25 • C for 4 weeks. Plates were then inspected every day. Fungal colonies with dark pigments were transferred to fresh MEA. Pure fungal strains were kept in 20% glycerol and deposited in the Culture Collection of Sustainable Development of Biological Resources Laboratory (SDBR), Faculty of Science, Chiang Mai University, Chiang Mai, Thailand.

Morphological and Growth Observations
Agar plugs (5 mm in diameter) from the edges of each fungal strain were transferred onto plates containing potato dextrose agar (PDA; Condalab, Madrid, Spain), MEA, and oatmeal agar (OA; Difco, Le Pont de Claix, France) and then kept at 25 • C in the dark. After four weeks of incubation, relevant colony features, including aerial mycelium and pigment production, were recorded and the colony diameter was measured. Cardinal growth temperatures were studied on MEA for 4 weeks in the dark at 4, 10, 15, 20, 25, 28, 30, 35, 37, and 40 • C using the method described by de Hoog et al. [3] with some modifications. A light microscope (Nikon Eclipse Ni-U, Tokyo, Japan) was used to study the micromorphological features of each fungal strain. The anatomical structure related to size data (e.g., hyphae, budding cells, conidia, and chlamydospore) was based on at least 50 measurements of each structure using the Tarosoft (R) Image.

DNA Extraction, Amplification, and Sequencing
A Fungal DNA Extraction Kit (FAVORGEN, Ping-Tung, Taiwan) was used to extract genomic DNA from the 3-week-old fungal culture of each strain that grew on MEA at 25 • C. Ribosomal DNA (ITS and nrSSU regions) and protein-coding (tef, tub, and act) genes were amplified by polymerase chain reaction (PCR) using suitable primers ( Table 2). PCR amplifications were performed using 20-µL reaction mixtures containing 1 µL of genomic DNA, 1 µL of 10 µM forward and reverse primers, 10 µL of Quick TaqTM HS DyeMix (TOYOBO, Osaka, Japan), and 7 µL of deionized water. PCR amplification conditions consisted of an initial denaturation step conducted at 95 • C for 5 min, followed by 35 cycles of denaturation at 95 • C for 30 s, an annealing step for 30 s, at appropriate temperatures (Table 2), and an elongation step at 72 • C for 1 min on a peqSTAR thermal cycler (PEQLAB Ltd., Fareham, UK). PCR products were checked on 1% agarose gel electrophoresis and were purified using a PCR clean up Gel Extraction NucleoSpin ® Gel and a PCR Clean-up Kit (Macherey-Nagel, Düren, Germany). Purified PCR products were then sequenced by 1st Base Company (Kembangan, Malaysia).

Sequence Alignment
The resulting ITS, nrSSU, tef, tub, and act sequences were assessed for similarity analysis in the GenBank database via BLAST searching. The sequences from this study, and those of closely related fungi, were obtained from the nucleotide GenBank database and previous studies as listed in Table 3. Multiple sequence alignment was carried out using MUSCLE in MEGA v. 6 [128] and the results were enhanced, when necessary, using BioEdit v.6.0.7 [129].
Exophiala angulospora Exophiala aquamarina Exophiala asiatica Exophiala attenuata   Note: species obtained in this study are in bold. Superscript "T" indicates type species and "-" represents the absence of sequence data in GenBank.

Phylogenetic Analyses
Phylogenetic analyses were performed using combination datasets of ITS, nrSSU, tef, tub, and act genes. Cyphellophora eucalypti CBS 124764 and C. fusarioides MUCL 44033 were used as the outgroup. Maximum likelihood (ML) and Bayesian inference (BI) methods were used to generate a phylogenetic tree. For ML analysis, 25 categories and 1000 bootstrap (BS) replications under the GTRCAT model [145] were performed on RAxML-HPC2 version 8.2.12 [146] on the CIPRES web portal [147]. The evolutionary model of nucleotide substitution for BI analysis was selected independently for each gene using MrModeltest v. 2.1 [148]. The GTR + I + G substitution model was the best fit for the ITS and nrSSU genes while the HKY + I + G substitution model was the best fit for the tef and tub genes, and the HKY + G substitution model was the best fit for the act gene. MrBayes v.3.2.6 was used for BI analysis [149]. In total, 6 simultaneous Markov chains were run for 5 million generations with random initial trees, wherein every 1000 generations were sampled. A burn-in phase was used to eliminate the first 2000 trees while the remaining trees were utilized to create a phylogram with a 50% majority-rule consensus. The Bayesian posterior probability (PP) was then calculated. Branches with BS and PP values of more than or equal to 70% and 0.95, respectively, were deemed to have been substantially supported. The tree topologies were visualized in FigTree v1.4.0 [150].

Fungal Isolation and Morphological Observations
A total of fifteen fungal strains were obtained in this study. Thirteen strains were isolated from rock samples collected from Lamphun Province and two strains were isolated from rock samples collected from Sukhothai Province. All fungal strains were cultivated on MEA at various temperatures (4-40 • C) and the diameter of the colonies was measured after 4 weeks of incubation. The results indicated that temperature had a significant effect on fungal growth. The average colony diameter of each fungal strain is shown in Table 4. It was found that that all fungal strains could not grow at 4 and 40 • C. However, all fungal strains grew well in temperatures ranging from 25-30 • C, with the exception of the strains SDBR-CMU417 and SDBR-CMU418. Five fungal strains (SDBR-CMU404, SDBR-CMU405, SDBR-CMU406, SDBR-CMU407, and SDBR-CMU408) showed the highest average value of the colony diameter at 28 • C while eight fungal strains (SDBR-CMU409, SDBR-CMU410, SDBR-CMU411, SDBR-CMU412, SDBR-CMU413, SDBR-CMU414, SDBR-CMU415, and SDBR-CMU416) showed the highest average value of the colony diameter at 30 • C. The results indicate that the highest average value of the colony diameter of two fungal strains, namely SDBR-CMU417 and SDBR-CMU418, was found at 20 • C; however, they did not grow at 35 and 37 • C. Based on the morphological characteristics, all fungal isolates were initially identified as belonging to the genus Exophiala. The identification was then further confirmed by the multi-gene phylogenetic analysis of the ITS, nrSSU, tub, tef, and act sequences.

Phylogenetic Results
A phylogenetic tree was constructed using a combination of the ITS, nrSSU, tub, tef, and act genes containing 3616 characters, including gaps (ITS: 1-739, nrSSU: 740-1829, tef : 1830-2454, tub: 2455-3045, and act: 3046-3616). The phylogram was constructed, consisting of 105 specimens of Exophiala and 2 specimens of the outgroup (Cyphellophora fusarioides MUCL 44033 and C. eucalypti CBS 124764). RAxML analysis of the combined dataset yielded the best scoring tree, with a final log likelihood value of −38,143.750648. The matrix was comprised of 1880 distinct alignment patterns with 53.45% undetermined characters or gaps. Estimated base frequencies were recorded as follows: A = 0.2297, C = 0.2831, G = 0.2311, T = 0.2561; substitution rates AC = 1.2386, AG = 4.4108, AT = 0.9986, CG = 0.8412, CT = 7.1418, and GT = 1.0000. The gamma distribution shape parameter alpha was equal to 0.3965 and the Tree-Length was equal to 12.0800. Using BI analysis, the final average standard deviation of the split frequencies at the end of the total MCMC generations was estimated to be 0.00513. In terms of topology, the phylograms of the ML and BI analyses were similar (data not shown). The phylogram generated from the ML analysis is shown in Figure 4. Our phylogenetic tree was constructed concordantly and is supported by previous studies [4,18]. The phylogram separated all fungal strains in this study into four monophyletic clades with high BS and PP support values. These clearly formed distinct lineages from previous known Exophiala species with high BS and PP support values. The results of our study revealed that two fungal strains, namely SDBR-CMU417 and SDBR-CMU418 (introduced as E. siamensis), were clearly separated from the previously known species of Exophiala. Moreover, five fungal strains, SDBR-CMU404, SDBR-CMU405, SDBR-CMU406, SDBR-CMU407, and SDBR-CMU408 (introduced as E. lamphunensis), formed a sister taxon to the two strains SDBR-CMU415 and SDBR-CMU416 (described here as E. saxicola), with 80% and 1.00 BS and PP support values, respectively. Notably, E. lamphunensis and E. saxicola formed a sister clade to E. xenobiotica, with high BS (98%) and PP (1.0) support values. Moreover, our six strains, SDBR-CMU409, SDBR-CMU410, SDBR-CMU411, SDBR-CMU412, SDBR-CMU413, and SDBR-CMU414 (introduced as E. lapidea), formed a sister taxon to E. moniliae (BS = 99% and PP = 1.0).     GenBank: ON555798 (ITS), ON555813 (nrSSU), ON544242 (tef), ON544227 (tub), and ON544257 (act).
Growth temperature: growth occurred within a range of 10-37 • C, optimum at 28 • C, while no growth at 4 and 40 • C.
The phylogenetic analyses of the combined ITS, nrSSU, tub, tef, and act sequences confirmed that E. lamphunensis formed a monophyletic clade that clearly distinguished it from E. nagquensis, E. oligosperma, E. saxicola, and E. xenobiotica. Exophiala lamphunensis formed a sister clade to E. saxicola. However, sequence similarity and pairwise nucleotide comparison of tef data also showed that E. lamphunensis differs from E. saxicola in 97% and 3.1% (5/162 bp), respectively. Differences in the morphological characteristics and the optimum growing temperature were found between E. lamphunensis and E. saxicola. Exophiala lamphunensis produces soluble pigment on PDA and chlamydospore production is absent while this was not the case for E. saxicola. The slightly wider size of the germinating cells in E. lamphunensis (3.1-7.3 × 2.4.-5.8 µm) distinguished it from E. saxicola (3.6-6.0 × 1.9-3.7 µm). Additionally, E. lamphunensis had a lower optimum temperature (28 • C) than E. saxicola (30 • C). Therefore, E. lamphunensis and E. saxicola were considered as different species based on their morphological, optimal growth temperature, and tef sequence data.
The multi-gene phylogenetic analyses (ITS, nrSSU, tub, tef, and act genes) confirmed that E. lapidea formed a monophyletic clade that clearly separated it from the other previous known Exophiala species and closely related species. A phylogram showed that E. lapidea formed a sister taxon to E. moniliae (Figure 4). However, the shorter size of conidia in E. moniliae (2.3-3.9 × 1.6-2.2 µm) clearly distinguished it from E. lapidea [15].
The phylogenetic analyses of the combined ITS, nrSSU, tub, tef, and act sequences confirmed that E. saxicola formed a monophyletic clade that clearly distinguished it from the other closely related species, namely E. nagquensis, E. oligosperma, and E. xenobiotica. Furthermore, E. saxicola formed a sister clade to E. lamphunensis. However, differences in the morphological characteristics, optimal growth temperature, and tef sequence data of E. saxicola and E. lamphunensis were observed and described above.
Culture characteristics: Colonies on PDA, MEA, and OA were described at 25 °C after 28 days of incubation ( Figure 8A). Colonies on PDA were 14-21 mm in diameter, restricted, irregular, convex in elevation, and velvety with brownish-grey and dark-brown margins. Reverse black. Colonies on MEA and OA restricted, circular, flat, velvety. Colonies on MEA grew to 9-11 mm in diameter with dark-green to greyish-green and white margins. Reverse dark green. Colonies on OA reached a diameter of 15-16 mm with darkgreen and greyish-green margins. Reverse black and olive margin. Budding cells rarely, GenBank: ON555811 (ITS), ON555826 (nrSSU), ON544255 (tef ), ON544240 (tub), and ON544270 (act).
Moreover, a multi-gene phylogenetic analysis confirmed that E. siamensis formed a well-supported monophyletic clade that was distinctly separated from other Exophiala species.
In this study, four new species of Exophiala, consisting of E. lamphunensis, E. lapidea, E. saxicola, and E. siamensis, were introduced. The different morphological characteristics identified between the four new species indicate that only E. lamphunensis produced soluble pigments around the colonies on PDA. Chlamydospore formations were observed in E. saxicola and E. siamensis, but this was not the case for E. lamphunensis and E. lapidea. Additionally, the budding cells of E. siamensis were larger and wider than those of E. lapidea and E. lamphunensis. However, the germinating cells and conidia of our four species were not observed to be different. The optimum growth temperature of E. lapidea and E. saxicola was 30 • C, which was higher than for E. lamphunensis (28 • C) and E. siamensis (28 • C). Additionally, the maximum growth temperature of E. siamensis (30 • C) was lower than for the other three new species (37 • C). Subsequently, our phylogenetic analyses of the combined five genes (ITS, nrSSU, tub, tef, and act) revealed that the four new species formed distinct lineages within the genus Exophiala. Therefore, a combination of the morphological characteristics and the molecular analyses conducted in our study strongly support the recognition of four new Exophiala species.
Exophiala species have been isolated in various habitats throughout the world as shown in Table 1. Several Exophiala species have been identified as potential agents of human and animal diseases. However, in some studies, certain Exophiala species have been employed in agricultural and biotechnological applications. In this study, four new Exophiala species were isolated from rock samples collected from natural forests located in northern Thailand. Our findings are similar to those of previous studies, which reported that some Exophiala species (e.g., E. bonaiae, E. cinerea, E. clavispora, E. ellipsoidea, and E. nagquensis) have been successfully isolated from rock samples. However, there have been no prior reports involving investigations of rock-inhabiting fungi in Thailand. Therefore, our study is the first of its kind to report on the discovery of Exophiala on rocks in Thailand. Prior to our study, a total of three Exophiala species (E. dermatitidis, E. jeanselmei, and E. spinifera) were known from Thailand [43,46,67]. Therefore, the successful identification of the Exophiala species in this study has increased the number of species found in Thailand to 7 species and has led to 67 global species. The outcomes of this present study will provide scientists and researchers with valuable information that can stimulate deeper investigations of rock-inhabiting fungi in Thailand. Ultimately, these findings will help researchers gain a better understanding of the distribution and ecology of Exophiala.