Abstract

INTRODUCTION

Exophiala is a genus of saprophytic fungi that have been isolated from environments rich in hydrocarbons (1–3) or from hot, humid, and oligotrophic environments, such as dishwashers (4), steam bath facilities (2), and bathrooms (5). This genus belongs to the family Herpotrichiellaceae and currently consists of more than 30 species. Traditionally, these fungi are considered dematiaceous molds (6). Due to their phenotypic characteristics at the beginning of colony formation, these fungi are often also referred to as “black yeasts” (7), a misnomer which sometimes may mislead the choice of antifungal agents. When the cultures mature, brown hyphae are formed which bear conidiogenous cells referred to as annellides, a typical characteristic of this genus of fungi (8).

Although Exophiala species are environmental fungi, they should not be disregarded as contaminants when they are isolated from clinical specimens. These fungi are causative agents of skin and subcutaneous tissue infections and of systemic infections, such as prosthetic valve endocarditis, dialysis-associated peritonitis, and disseminated infections, especially in immunocompromised patients (9–17). In this study, we report the clinical spectrum of Exophiala infections in our hospital, characterized by phenotypic examination and sequencing of the internal transcribed spacer 1 (ITS1)-5.8S-ITS2 rRNA gene cluster (referred to hereinafter as ITS) and RNA polymerase II's largest subunit gene (Rpb1). During the process, we also discovered a potentially novel Exophiala species from the toe nail clipping of a patient with onychomycosis. The strain, named HKU32^T, exhibited phenotypic characteristics that do not fit into the pattern of any known Exophiala species. Amplification and sequencing of four independent DNA regions showed that it is distinct from all other Exophiala species. On the basis of these studies, we propose a new species, Exophiala hongkongensis sp. nov., to describe this fungus.

MATERIALS AND METHODS

RESULTS

Phenotypic characterization and ITS and Rpb1 sequencing of Exophiala strains and identification of a novel Exophiala species.

On SDA, the colonies of all 12 strains were initially brown to black, moist, and yeastlike. All mature colonies became velvety, with the front color gray-to-black-olivaceous and the reverse black. Microscopic examination of the young cultures of all the 12 strains showed subspherical, budding, yeastlike cells. As the cultures became mature, septate hyphae which bore tubular and rocket-shaped annellides that tapered to form narrow elongated tips were observed. Ellipsoidal conidia, in clusters at the apices of the annellides or at the sides of the conidiophores, were produced from the annellides.

PCR of the ITS and partial Rpb1 gene of the 12 strains and the reference strains showed bands at about 600 bp and 500 bp, respectively. Sequencing of the ITS showed that 11 of the 12 strains were known Exophiala species (Fig. 1). Sequencing of the partial Rpb1 gene revealed results concurring with those of ITS sequencing (data not shown). As for HKU32^T (case 10), although it is most closely related to E. nishimurae CBS 101538^T, there was a 46 (8.0%)-base difference between the ITS of HKU32^T and that of E. nishimurae CBS 101538^T and a 37 (8.1%)-base difference between the partial Rpb1 gene of HKU32^T and that of E. nishimurae CBS 101538^T.

Open in a separate window

Fig 1

Phylogenetic trees showing the relationships of the 12 strains of Exophiala species in this study to known Exophiala species. The tree was inferred from ITS sequence data by the maximum likelihood method with the substitution model GTR+G+I (general time-reversible model, with gamma-distributed rate variation [G] and an estimated proportion of invariable sites [I]). The scale bar indicates the estimated number of substitutions per 50 bases. Numbers at nodes indicate levels of bootstrap support calculated from 1,000 trees. All names and accession numbers are given as cited in the GenBank database.

Partial β-tubulin and β-actin gene sequencing of HKU32^T.

PCR of the partial β-tubulin and β-actin genes of HKU32^T and the reference strains showed bands at about 300 bp and 500 bp, respectively. Sequencing and phylogenetic analysis showed that HKU32^T is most closely related to E. nishimurae CBS 101538^T, concurring with the results of ITS and Rpb1 gene sequencing (Fig. 2). Pairwise alignment showed that there was a 28 (9.5%)-base difference between the partial β-tubulin gene of HKU32^T and that of E. nishimurae CBS 101538^T and a 36 (7.6%)-base difference between the partial β-actin gene of HKU32^T and that of E. nishimurae CBS 101538^T.

Open in a separate window

Fig 2

Concatenated phylogenetic tree showing the relationship of E. hongkongensis HKU32^T to closely related species. The tree was inferred from actin and β-tubulin sequence data by the maximum likelihood method with the substitution model K2+G+I (K2 referring to the Kimura two-parameter model). The scale bar indicates the estimated number of substitutions per 50 bases. Numbers at nodes indicate levels of bootstrap support calculated from 1,000 trees. All names and accession numbers are given as cited in the GenBank database, with the first accession numbers in the parentheses referring to the β-actin gene sequences and the second accession numbers in the parentheses referring to the β-tubulin gene sequences.

TAXONOMY

Description of Exophiala hongkongensis Woo, Ngan, Tsang, Ling, Chan, Leung, Yuen, Lau, sp. nov. MycoBank accession no. MB563298. Teleomorph: unknown. Known distribution: Hong Kong. Etymology: hong.kong.en′sis. N.L. fem. adj., named after Hong Kong, where the type strain was isolated. Specimen examined: Hong Kong; from the big toe nail of a human presenting with onychomycosis in 2010 (holotype: dried culture in NBRC Herbarium H-13132; ex-type cultures: HKU32^T [= NBRC 109366^T = JCM 18697^T = CBS 131511^T]).

On SDA, HKU32^T initially grew slowly as black, slimy, yeast-like colonies with a diameter of 2 mm after incubation at 25°C for 14 days. Subsequently, the centers of the colonies became velvety, gray-olivaceous, and dome-shaped with black reverse, but the margin remained slimy and yeastlike after 28 days of incubation at 25°C (Fig. 3a and ​andb).b). Microscopic examination of the yeastlike colonies at 14 days of incubation after lactophenol cotton blue staining revealed brown yeast cells that were ellipsoidal, had an average size of 6 by 3.5 μm, and possessed short annellated zones to form budding cells (Fig. 3c). Mycelia were usually swollen and formed torulose hyphae with annellated zones found at their ends (Fig. 3c). At the subsequent velvety stage, septate mycelia were not swollen and conidia were formed distinctively in clusters (Fig. 3d and ​ande).e). The clusters of conidia were either intercalary or found at the tips of free conidiophores (Fig. 3d and ​ande).e). Under high-power magnification, intercalary annellides were observed in the form of short annellated pegs (Fig. 3f). These intercalary annellides were present along the creeping septate hyphae. Free conidiophores consisting of 1 or 2 cells with inconspicuous short annellated tips were also observed (Fig. 3g). They possessed the same olivaceous color as the creeping septate hyphae and were often constricted at their bases (Fig. 3g). At both stages, conidia were formed by percurrent growth through annellated zones. Most of them were oval or subglobose, with an average size of 3 by 2 μm, and some with truncated scars. No synanamorphs of Phialophora or Rhinocladiella types were observed. Scanning electron microscopic examination of HKU32^T at 14 days of incubation revealed budding yeast cells and torulose hyphae with distinct annellated zones, consistent with the results observed on light microscopy (Fig. 3h and ​andi).i). In addition, yeast cells with truncated scars were also observed (Fig. 3h).

Open in a separate window

Fig 3

Culture of HKU32^T on SDA after 28 days of incubation at 25°C showed a black slimy colony with gray olivaceous center (a) and black reverse (b). At the initial yeast stage (14 days of incubation), yeast and torulose hyphae with annellated zones were observed with light microscopy (c) and scanning electron microscopy (h and i). Yeast cells with truncated scars were observed (h), and distinct annellated zones were well demonstrated (h and i). At the subsequent velvety stage (28 days of incubation), conidia in clusters were either intercalary or found at the tips of free conidiophores under both low-power (d) and high-power magnification (e). Intercalary annellides with short pegs (f) and free conidiophores with constricted bases (g) were observed.

DISCUSSION

Using the polyphasic approach with a combination of phenotypic and genotypic techniques, we describe a wide variety of Exophiala species associated with different forms of clinical infections. Although Exophiala can often be identified to the genus level by morphological characteristics, identification of Exophiala to the species level by phenotypic characterization alone is very difficult. This difficulty is exemplified by our description of the first case of CAPD peritonitis caused by Exophiala in 2000 (Table 3, case 2) (13). At that time, we could not identify the Exophiala to the species level. Due to the advancement of molecular techniques and availability of DNA sequences of different gene loci in sequence databases such as GenBank, identification of Exophiala to the species level has been made possible. In this study, by sequencing two independent gene loci (ITS and Rpb1), eight different Exophiala species, including E. oligosperma, E. jeanselmei, E. lecanii-corni, E. bergeri, E. cancerae, E. dermatitidis, E. xenobiotica, and a novel Exophiala species, were isolated from 12 patients (Table 3 and Fig. 1). These Exophiala species were found to be associated with both superficial infections and invasive infections, including the case of CAPD peritonitis caused by E. xenobiotica (Table 3, case 2) and another case of pneumonia caused by E. oligosperma (Table 3, case 8). In two patients with cord blood and peripheral blood stem cell transplant, respectively, E. cancerae and E. dermatitidis were the colonizers of their gastrointestinal tracts (Table 3, cases 1 and 12).

Phenotypic and genotypic analysis revealed that HKU32^T is a novel species in the genus Exophiala, which we propose to name E. hongkongensis. Using four independent DNA regions widely used for phylogenetic analysis, including ITS and three housekeeping genes (Rpb1, β-tubulin, and β-actin genes), it was shown unambiguously that E. hongkongensis is closely related to but distinct from other Exophiala species (Fig. 1 and ​and2).2). Among the Exophiala species, E. hongkongensis is most closely related to E. nishimurae, as shown in all phylogenetic trees with high bootstrap supports. Comparison of phenotypic characteristics between E. hongkongensis and those of other closely related Exophiala species shown by ITS and Rpb1 sequencing also revealed unique phenotypic characteristics of E. hongkongensis (Table 4). E. hongkongensis grew very slowly and had a velvety colony texture after 28 days, short conidiophores of the same olivaceous color as the supporting hyphae, numerous spores, and no chlamydospore-like cells.

This study includes the first reported cases of onychomycosis associated with E. hongkongensis, E. bergeri, and E. oligosperma. Although Exophiala species are relatively common causes of subcutaneous and skin infections in the forms of phaeohyphomycosis, chromoblastomycosis, and mycetoma, they have only occasionally been reported to cause onychomycosis. Among the seven cases reported in the literature, four were caused by E. dermatitidis and three were caused by E. jeanselmei (Table 5), whereas for the four patients with Exophiala onychomycosis in the present study, two were caused by E. oligosperma, one was caused by E. bergeri, and one was caused by E. hongkongensis (Table 3). Exophiala species are global pathogens of onychomycosis, with cases reported from Asia, Europe, America, and Africa. Among the seven cases reported in the literature, underlying diseases leading to immunosuppressive states were present in three patients, including diabetes mellitus in two patients and renal transplantation in one (Table 5), whereas for the four patients with Exophiala onychomycosis in the present study, only the patient with E. hongkongensis onychomycosis had underlying hypertension (Table 3). In all 11 patients, the nails were infected by the fungi for one to a few years before presentation, indicating that the disease was a very indolent process with relatively mild symptoms and, hence, patients tended to observe them for a time and delay in seeking medical advice. The nails of the big toes were affected in eight patients (Tables 3 and ​and5).5). Due to underreporting and difficulty in making the microbiological diagnosis, the number of cases of Exophiala onychomycosis is probably underestimated. This is in line with the results of a recent study, which also suggested that the incidence of superficial infections caused by black yeastlike fungi could be underestimated (37). Another study also noted that nonthermophilic Exophiala species may expand in diabetic patients with poor blood circulation (38). A combination of phenotypic and genotypic techniques using the polyphasic approach in microbiology laboratories will facilitate the understanding of the epidemiology of Exophiala onychomycosis, as well as other infections associated with Exophiala species.

ACKNOWLEDGMENTS

Footnotes

REFERENCES