Identification of Trichosporon asahii by PCR Based on Sequences of the Internal Transcribed Spacer Regions

Abstract

Trichosporon asahii is a major causative agent of deep-seated trichosporonosis, which has a high mortality rate. To detect T. asahii, we have developed specific oligonucleotide primers based on the internal transcribed spacer regions of this organism’s genome. Amplification products were selectively obtained from only T. asahii DNA; the DNAs of other Trichosporon species, as well as those of medically relevant yeasts such as Candida albicans, Cryptococcus neoformans, and Malassezia furfur, were not amplified. This detection system will be useful as a microbiological tool for the diagnosis of trichosporonosis.

Trichosporon Behrend is a medically important genus that includes the causative agents of deep-seated, mucosa-associated, and superficial infections, including white piedra. Recently, the taxonomy of the genus Trichosporon was significantly revised on the basis of partial sequences of large-subunit (LSU) rRNA and DNA relatedness (3, 18, 22). The causative agent of trichosporonosis was previously believed to be Trichosporon cutaneum, but on the basis of the new taxonomy, it has been demonstrated that six Trichosporon species, T. asahii, T. asteroides, T. cutaneum, T. inkin, T. mucoides, and T. ovoides, are all associated with this infection (2, 5, 19, 20). It was also shown that the major causative agents of trichosporonosis differ according to the site of infection. T. asahii and T. mucoides are involved in deep-seated infections. T. asteroides and T. cutaneum are associated with superficial infections. T. ovoides is involved in capital white piedra, and T. inkin is associated with white piedra of the genital area. Since the first report of a brain abscess due to Trichosporon infection, there have been an increasing number of reports concerning this infectious agent (13, 26, 28). The majority of those patients with fatal disseminated fungemia were afflicted with leukemia or lymphoma and were in a profound neutropenic state when the Trichosporon infection developed. Deep-seated trichosporonosis is especially life threatening, with a high mortality rate; the prognosis for patients is very poor. In taxonomic investigations, most of the isolates obtained from patients with deep-seated trichosporonosis were identified as T. asahii (2, 5, 19, 20). Early diagnosis and treatment are therefore of paramount importance to trichosporonosis patients. Since we had already developed genus-specific primers for Trichosporon species, including nonpathogenic species (23), in the present study we developed a rapid PCR-based approach to the detection of a major causative agent of trichosporonosis, T. asahii.

The 78 strains used in this study, including 14 strains of T. asahii, are listed in Table 1. They include all species (17 species and 5 varieties) of the genus Trichosporon, as well as related medically relevant yeasts. Trichosporon clinical isolates have been identified on the basis of a nuclear DNA-DNA hybridization method (19, 20). For basidiomycetous yeasts, DNA was extracted by the method of Makimura et al. (9). For ascomycetous yeasts, a DNA extraction kit (Nucleon MiY; Amersham International plc, Buckinghamshire, United Kingdom) was used according to the manufacturer’s instructions. The internal transcribed spacer (ITS) regions were amplified with primers pITS1 (5′-TCCGTAGGTGAACCTGCGG-3′) and pITS4 (5′-TCCTCCGCTTATTGATATG-3′), which were derived from conserved regions of the small-subunit (SSU) and LSU rRNA genes, respectively. Direct sequencing of the PCR-amplified ITS1, ITS2, and 5.8S rRNA gene was performed with an ABI PRISM cycle sequencing kit (Applied Biosystems, Foster City, Calif.). To determine the sequences, two external primers, pITS1 and pITS4, were used. T. asahii var. asahii CBS 2479, T. asahii var. coremiformis CBS 2482, T. asahii var. faecalis CBS 4828, and T. asteroides CBS 2481 were sequenced. They are all type strains. To select primers that would specifically amplify only T. asahii, the sequences of the ITSs of potentially pathogenic yeasts, obtained from DNA sequence libraries, were aligned. The primers, chosen to align with regions which were not conserved in other medically relevant yeasts, were TAAF (forward; 5′-GGATCATTAGTGATTGCCTTTATA-3′) and pITS4 (reverse; 5′-TCCTCCGCTTATTGATATG-3′). The oligonucleotide primers were obtained from Greiner Japan (Tokyo, Japan). Amplification reactions were performed in PCR buffer containing the following: 50 mM KCl; 10 mM Tris-HCl (pH 8.3); 1.0 mM MgCl₂; 25 mM each dATP, dCTP, dGTP, and dTTP; 2 mM each oligonucleotide primers; and 0.5 U of Taq DNA polymerase (Takara, Shiga, Japan). The reaction mixtures were amplified in a Perkin-Elmer 9700 thermal cycler, using the following program: 94°C for 3 min, followed by 30 cycles consisting of 94°C for 30 s, 58°C for 30 s, and 72°C for 40 min, with a final extension period at 72°C for 10 min. After thermal cycling, 3 μl of the amplified product was run on a 1.5% (wt/vol) agarose gel, stained with ethidium bromide, and visualized under a UV light.

TABLE 1.

Specificity of the primers for T. asahii and related species

Species	Strain	Sourcea	PCR productb
T. asahii var. asahii	M 9306c	CBS 2479	+
	M 9402	Blood	+
	M 9403	Blood	+
	M 9415	Lung	+
	M 9416	Lung	+
	M 9417	Lung	+
	M 9411	Sputum	+
	M 9410	Feces	+
	M 9405	Urine	+
	M 9406	Urine	+
T. asahii var. coremiformis	M 9309c	CBS 2482	+
	M 9404	Urine	+
T. asahii var. faecalis	M 9312c	CBS 4828	+
	M 9409	Urine	+
T. asteroides	M 9308c	CBS 2481	−
	M 9329	CBS 7623	−
	M 9330	CBS 7624	−
T. aquatile	M 9317c	CBS 5973	−
	M 9321	CBS 5988	−
T. brassicae	M 9322c	CBS 6382	−
T. cutaneum	M 9304c	CBS 2466	−
	M 9307	CBS 2480	−
	M 9423	Skin	−
	M 9425	Skin	−
T. domesticum	M 9401c	House	−
	M 9412	Nail	−
T. dulcitum	M 9337c	CBS 8257	−
T. inkin	M 9316c	CBS 5585	−
	M 9333	CBS 7926	−
T. gracile	M 9334c	CBS 8189	−
T. jirovecii	M 9326c	CBS 6864	−
T. loubieri var. loubieri	M 9327c	CBS 7065	−
T. loubieri var. laibachii	M 9319c	CBS 5790	−
T. montevideense	M 9323c	CBS 6721	−
T. mucoides	M 9331c	CBS 7625	−
	M 9332	CBS 7626	−
	M 9422	Skin	−
	M 9424	Skin	−
T. moniliiforme	M 9305c	CBS 2467	−
T. ovoides	M 9328c	CBS 5585	−
	M 9315	CBS 5580	−
	M 9407	Urine	−
T. pullulans	M 9339c	CBS 2532	−
	M 9440	CBS 2533	−
	M 9441	CBS 2538	−
T. sporotrichoides	M 9336c	CBS 8245	−
Candida albicans	M 1001c	CBS 562	−
	M 1447	NIHA207	−
	M 1445	NIHB792	−
Candida glabrata	M 4002c	IFO 0622	−
Candida guilliermondii	M 1002c	CBS 566	−
	M 2054	Skin	−
	M 1023	ATCC 9058	−
Candida kefyr	M 1018c	JCM 9559	−
Candida lusitaniae	M 1010c	CBS 4413	−
Candida parapsilosis	M 1015c	ATCC 22019	−
Candida tropicalis	M 1017c	ATCC 7349	−
Cryptococcus neoformans var. neoformans	M 9010c	CBS 132	−
	M 9026	NIHB3501	−
	M 9116	Cerebrospinal fluid	−
	M 9117	Skin	−
	M 9125	VUT 77063	−
	M 9216	Cerebrospinal fluid	−
Cryptococcus neoformans var. gattii	M 9028	NIH191	−
	M 9228	CDCB3174	−
	M 9229	CDCB3175	−
	M 9230	CDCB3183	−
	M 9246	CDCB3184	−
	M 9247	CDCB3185	−
Cryptococcus albidus	M 9001c	CBS 142	−
Debaryomyces hansenii var. hansenii	M 5012c	JCM 1990	−
	M 5111	CBS 1793	−
	M 5033	JCM 1521	−
Malassezia furfur	M 9701c	CBS 1878	−
	M 9702	CBS 6093	−
	M 9703	CBS 7019	−
Malassezia pachydermatis	M 9704c	CBS 1079	−
Rhodotorula mucilaginosa	M 9501c	CBS 17	−

^aAbbreviations: ATCC, American Type Culture Collection, Rockville, Md.; CBS, Centraal Bureau voor Schimmelcultures, Baarn, The Netherlands; CDC, Centers for Disease Control and Prevention, Atlanta, Ga.; JCM, Japan Collection of Microorganisms, Saitama, Japan; M. Meiji College of Pharmacy, Tokyo, Japan; NIH, National Institutes of Health, Bethesda, Md.; VUT, School of Veterinary Medicine, Faculty of Agriculture, University of Tokyo, Tokyo, Japan. Blood, lung tissue, sputum, feces, urine, skin, nail, and cerebrospinal fluid specimens were clinical isolates. 

^b+, product obtained; −, no product obtained. 

^cType strain. 

Table 1 shows the specificity of the oligonucleotide primers that we designed against the medically relevant yeasts. The primers amplified the DNAs of only T. asahii, including all three varieties, and produced approximately 500-bp fragments (Fig. 1). We used 14 strains of T. asahii, including clinical isolates obtained from blood, feces, sputum, lung tissue, and urine. All of the strains produced the specific DNA fragment. DNAs of the other Trichosporon species, as well as those of medically relevant yeasts such as Candida albicans, Cryptococcus neoformans, and Malassezia furfur, were not amplified by the detection system. Our data show that of the DNAs of the pathogenic yeasts, these primers selectively amplify only that of T. asahii.

FIG. 1.

Agarose gel electrophoresis of PCR products of T. asahii and related species. Lane 1, molecular size marker (X174 DNA, HaeIII digest; Takara, Shiga, Japan); lane 2, T. asahii var. asahii M 9306; lane 3, T. asahii var. coremiformis M 9309; lane 4, T. asahii var. faecalis M 9312; lane 5, clinical isolate T. asahii var. asahii M 9402; lane 6, clinical isolate T. asahii var. asahii M 9415; lane 7, clinical isolate T. asahii var. asahii M 9411; lane 8, clinical isolate T. asahii var. asahii M 9405; lane 9, T. asteroides M 9308; lane 10, T. ovoides M 9328; lane 11, T. mucoides M 9331; lane 12, T. cutaneum M 9304; lane 13, Candida albicans M 1001; lane 14, Candida parapsilosis M 1015; lane 15, Cryptococcus neoformans var. neoformans M 9010; lane 16, Cryptococcus neoformans var. gattii M 9028; lane 17, M. furfur M 9701.

To design specific oligonucleotide primers that would amplify only T. asahii sequences, we sequenced the ITS regions, including the end of the SSU rRNA, of three varieties of T. asahii and the closest phylogenetically related species, T. asteroides. Our primers amplified all three varieties of T. asahii. Their ITS1 region sequences were almost identical, and differential sequences were not found. Molecular phylogenetic analysis based on the sequences of SSU and/or LSU rRNA genes demonstrated that the three varieties of T. asahii are located in the same cluster (1, 3, 16, 17). Moreover, they showed intermediate levels of nuclear DNA relatedness (40 to 56%) to each other in a DNA-DNA hybridization experiment (18). Taxonomic investigations showed that most of the clinical isolates obtained from deep-seated trichosporonosis patients were T. asahii (2, 5, 19, 20). In addition, of the three varieties of T. asahii, the primary causative agent is the variety asahii. T. asahii var. coremiformis and faecalis are also occasionally isolated from clinical specimens (19, 20). Detection of all three varieties of T. asahii at once would be clinically significant. To detect these pathogenic fungi, primers for PCR have been designed on the basis of the sequences of SSU or LSU rRNA genes (4, 25). However, the relatively high level of sequence similarity observed between some species appears to limit the value of SSU rRNA for differentiating phylogenetically closely related species.

The ITS region is located between the SSU and LSU rRNA genes. The ITS region is subdivided into the ITS1 region, which separates the SSU and 5.8S rRNA genes, and the ITS2 region, which is found between the 5.8S and LSU rRNA genes. It is generally thought that the ITS regions have higher rates of divergence than SSU, 5.8S, or LSU rRNA genes. Therefore, on the basis of the sequences of the ITS regions, highly specific primers for PCR can be designed. By analyzing the sequences of the ITS regions, all Trichosporon isolates may be easily identified to the species level.

Several techniques for identification of Trichosporon species have been previously reported. Guého et al. (2) mentioned that six pathogenic species were clearly differentiated by several key characteristics: a combination of assimilation of carbon compounds, cycloheximide resistance, and ability to grow at 37°C. Immunohistochemical identification techniques were also reported by some researchers (8, 12). However, the cell wall antigens of Trichosporon species cross-react with the capsular polysaccharide of Cryptococcus neoformans (10, 11, 15). Shinoda and coworkers prepared specific-factor sera for Trichosporon species, and their studies indicated that Trichosporon species have at least four different serotypes: I, II, III, and I-III (7, 14). In addition, the serotypes correlated well with the molecular phylogenetic tree based on LSU rRNA sequences (16). Pathogenic Trichosporon species have serotype I (T. cutaneum and T. mucoides) or serotype II (T. asahii, T. asteroides, T. inkin, and T. ovoides), and serotype III and I-III species, such as T. gracile, are not responsible for infection. Serotyping provides useful information for tentative identification (7). While these physiological and biochemical identification techniques are easy and convenient, the pathogen is not rapidly detected. Deep-seated trichosporonosis is life threatening, with a high mortality rate, and the prognosis for the patient is very poor. Some authors have reported that the mortality rate (despite antifungal therapy, including amphotericin B) is 64 to 88% (6, 24). Our primers amplified T. asahii DNA from clinical specimen, although a significant number of specimens have not yet been studied (unpublished data). A rapid identification method such as PCR is a useful microbiological tool for the diagnosis of trichosporonosis.

In conclusion, we successfully developed species-specific primers for T. asahii based on the sequences of the ITS regions. We expect this detection system to be applicable to the clinical diagnosis of trichosporonosis.