Molecular identification of Pichia guilliermondii, Debaryomyces hansenii and Candida palmioleophila.

Traditional phenotypic methods and commercial kits based on carbohydrate assimilation patterns are unable to consistently distinguish among isolates of Pichia guilliermondii, Debaryomyces hansenii and Candida palmioleophila. As result, these species are often misidentified. In this work, we established a reliable method for the identification/differentiation of these species. Our assay was validated by DNA sequencing of the polymorphic region used in a real-time PCR assay driven by species-specific probes targeted to the fungal ITS 1 region. This assay provides a new tool for pathogen identification and for epidemiological, drug resistance and virulence studies of these organisms.

Introduction

The precise identification of some fungal species is often very difficult when using only biochemical or phenotypic methods (Dooley ; Fenn ). However, the advent of DNA-based methods largely overcame the limitations of traditional methods and studies using molecular approaches revealed a greater diversity in fungi (Odds ; Chen ).

Some closely related species are often misidentified because of the great similarity in their biochemical and morphological characteristics. Desnos-Ollivier reported the misidentification of Pichia guilliermondii (teleomorph Candida guilliermondii), Debaryomyces hansenii (teleomorph Candida famata) and Candida palmioleophila. Their results showed that only 23 of 36 isolates identified as P. guilliermondii and three of 26 identified as D. hansenii were confirmed by sequencing the ITS1-5.8S-ITS2 and D1/D2 regions of the ribosomal cistron. Other species such as Candida albicans (Odds ; Jabra-Rizk ; Tietz ) and Candida parapsilosis (Lasker ) show the same problem.

In our laboratory, we have had problems differentiating (1) P. guilliermondii, C. palmioleophila and D. hansenii, (2) Candida krusei and Candida inconspicua, and (3) Candida pelliculosa and Candida subpelliculosa when using the commercial kit API® 20 C AUX (Biomérieux, France) (Table S1). The Vitek Yeast Biochemical Card and the ID 32C, two widely used methods for yeast identification, were tested by different groups (Dooley ; Lo ; Burton ) and their findings confirmed the problem of incorrect or inconsistent identification. Lo suggested that the laboratory routine should include at least two methods for yeast identification.

Misidentification of the fungal species can compromise epidemiological or antibiotic susceptibility studies and over- or underestimate the species abundance. Precise identification is therefore necessary and molecular approaches can provide the tools for a fast method. In this study, we developed a real-time PCR method that can differentiate/identify P. guilliermondii, D. hansenii and C. palmioleophila, and sequencing the ITS 1 region of these species confirmed our results.

Materials and Methods

Total DNA was extracted as described by Philippsen . DNA was quantified with a Qubit® fluorometer and a Quant-iT PicoGreen® dsDNA BR assay kit (Invitrogen, Life Technologies, Eugene, OR, USA) according to the manufacturers recommendations.

A Go® Taq Flexi DNA Polymerase kit (Promega, Madison, WI, USA) was used to amplify the ITS 1 region with ITS1/ITS2 primers (White and Lee, 1990) and a TopoTA Cloning® kit (Invitrogen) was used for the constructions. The sequences were obtained using a Big Dye® Terminator v3.1 Cycle Sequencing kit followed by automatic sequencing in an ABI 3100 (Applied Biosystems, Life Technologies, Foster City, CA, USA).

The sequences obtained from the three American Tissue Culture Collection (ATCC) species, D. hansenii ATCC 36239, C. famata ATCC 62894 and P. guilliermondii ATCC 6260, were compared by the BLASTN program (Zhang ) against the GenBank non-redundant database (nr), EMBL, DDBJ and PDB nucleotide collections. The sequences obtained were aligned with ClustalW software (Larkin ) and then used to create a consensus sequence for each species and to choose target regions (data not shown). The primers and TaqMan® Minor Groove Binder (MGB) probes were designed using the software Primer Express v. 2.0 and default parameters (Applied Biosystems).

A real-time PCR was done using the TaqMan® Universal PCR Master Mix (Applied Biosystems) in the following singleplex reaction mixture: 10 ng of sample DNA, 1X TaqMan® Universal PCR Master Mix, 200 nM of each primer, 300 nM of TaqMan® MGB probe (Table 1) and water to a total volume of 25 μL. All reactions were done in duplicate. The cycling conditions were set in an ABI 7300 real-time PCR cycler fitted with SDS software v. 1.2.3 (Applied Biosystems) as follows: 10 min at 95 °C followed by 35 cycles of 95 °C for 30 s and 60 °C for 1 min.

Table 1

Sequences of primers and probes used in this work. In the probes, the reporter dye is indicated in bold and the “No fluorescent quencher (NFQ)” region is underlined. The probe sequence is inserted between these two regions.

Primers/probes	Sequence	Tm (ºC)
ITS1 Fw	TGAACCTGCGGAAGGATCAT	59
ITS2 Rev	TCCGTTGTTGAAAGTTTTGAAGATT	59
TaqMan® MGB probes
D. hansenii	VICTTGTTATTACAAGAACTTTTGCMGBNFQ	70
P. guilliermondii	6FAMTTGATACAGAACTCTTGCTTTGMGBNFQ	70
C. palmioleophila	NEDCTTTATTAGAAACTATTGCTTTGGCMGBNFQ	70

The efficiency of multiplex reactions was tested by mixing all of the probes together in the same concentrations as shown above.

Results and Discussion

The sequences obtained from the ATCC strains (data not shown) were matched in a search against the (nr) nucleotide bank using the BLAST program (the alignment files are provided in Supplementary Material Figure S2) in order to choose the potential target region for genotyping by real-time PCR (Figure 1).

Figure 1

Sequence alignment of the polymorphic ITS 1 region. The ITS 1 regions of the three species were aligned using ClustalW. The numbers above the sequences represent base positions, the gray boxes are the targets of forward and reverse primers and the clear boxes indicate the probe targets.

Comparison of the two methods of identification (API 20C AUX and ribotyping by sequencing) revealed the difficulty in differentiating P. guilliermondii and D. hansenii, in the correct identification of C. parapsilosis versus D. hansenii (Burton and Table S1) and the impossibility of differentiating C. krusei and C. insconspicua.

Initially, three polymorphic domains (ITS 1, ITS 2 and D1/D2) of the ribosomal cistron from these species were aligned using ClustalW (Larkin ). Of these three regions, only the ITS 1 region provided suitable discrimination (Figure 1). This region is ideal because it is flanked by two conserved domains: the end of 18S rRNA and the beginning of 5.8S rRNA. The amplicon is short (about 300 bp) and the sequence is variable among different species but conserved among strains of the same species.

Figure 2 shows the amplification plot of the real-time PCR genotyping done using species-specific probes. The region indicated as NTC (no template control) confirmed the specificity of each probe since only in the presence of the specific target was there amplification. To confirm this finding, we analyzed the real-time PCR products in a 2% agarose gel stained with ethidium bromide and detected the expected amplicons (data not shown).

Figure 2

Plots showing real-time PCR amplification of singleplex assays monitored with fluorescent TaqMan® MGB probes. Delta Rn versus Cycle amplification plots were obtained using probes for (A) P. guilliermondii, (B) C. palmioleophila and (C) D. hansenii templates. The NTC (no template control) curves show no detection in the absence of templates and in the presence of non-specific targets, i.e., C. palmioleophila and D. hansenii in (A), P. guilliermondii and D. hansenii in (B) and P. guilliermondii and C. palmioleophila in (C).

The multiplex reaction worked as well as the singleplex test (Figure S1) and can be used for fungus identification, thereby reducing the costs of the assay.

The close relationship between C. palmioleophila and D. hansenii has previously been shown by a phylogenetic analysis using data from the D1/D2 and ITS regions (Desnos-Ollivier ). In our study, the strain C. famata ATCC 62894, used as a positive control, was identified as C. palmioleophila by sequence analysis of the D1/D2 region of rRNA (Table S1) and by the real-time PCR assay described here (Figure 2).

In conclusion, the three species examined here are difficult to identify using standard laboratory tests. Ribosomal RNA sequencing is the gold standard for identification but is generally expensive and time consuming. The real-time PCR assay described here is a very effective, rapid, low-cost alternative. The method can unambiguously identify isolates and confirm the identification of strains analyzed by traditional methods, with the advantage of measuring species abundance if necessary.

Supplementary Material

The following online material is available for this article:

Comparison of results obtained DNA sequencing.

Real-time PCR amplification of multiplex assays.

BLAST alignments for C. palmioleophila; D. hansenii and P. guilliermondii.

This material is available as part of the online article from http://www.scielo.br/gmb.