OUP user menu

Multilocus sequence typing confirms synonymy but highlights differences between Candida albicans and Candida stellatoidea

Mette D. Jacobsen, Teun Boekhout, Frank C. Odds
DOI: http://dx.doi.org/10.1111/j.1567-1364.2008.00392.x 764-770 First published online: 1 August 2008

Abstract

We used multi-locus sequence typing (MLST) to investigate 35 yeast isolates representing the two genome-sequenced strains plus the type strain of Candida albicans, four isolates originally identified as Candida stellatoidea type I and 28 representing type strains of other species now regarded as synonymous with C. albicans. DNA from all 32 C. albicans synonyms readily formed PCR products with the C. albicans MLST primer sets. Their sequences placed all of them within the existing C. albicans clade structure, represented by 1516 isolates. One isolate, originally received as Mycotorula sinensis, was resistant to flucytosine, but no other unusual susceptibilities were found to polyene, azole or echinocandin antifungal agents. The four isolates of C. stellatoidea type I coclustered with two other sucrose-negative isolates, originally identified as examples of Candida africana, in a group of strains highly distinct from the majority of C. albicans. Our results not only confirm the synonymity of all the isolates with C. albicans but also confirm an obvious genotypic difference in the case of C. stellatoidea type I.

Keywords
  • Candida albicans
  • Candida stellatoidea
  • MLST

Introduction

Candida albicans is the Candida species most commonly associated with human infections, and the species most intensively studied among the more than 160 accepted species within the genus Candida (Meyer et al., 1998). In the late 19th and early 20th centuries, the fungus now known as C. albicans was rediscovered many times. In her Ph.D. thesis, Berkhout (1923) studied the many conflicting accounts of species that had been proposed in the genera Monilia, Oidium, Oospora and Torula, examined the phenotypic properties of the published isolates and proposed a classification that erected the genus Candida with an isolate of Candida tropicalis (originally designated Candida vulgaris) as the type strain. The species C. albicans was named to accommodate several yeasts with identical characteristics, which were thus synonyms of C. albicans. The genus name Candida was accepted and, by 1930, recommended by others in the field (Ciferri & Redaelli, 1929; Ashford, 1930). Candida was adopted by the Eighth Botanical Congress in 1954 as a nomen conservandum.

In the most recent edition of the major taxonomic resource, The Yeasts, A Taxonomic Study, 163 species or varietal names are listed as synonyms of C. albicans (Meyer et al., 1998). Of these, 95 were first published after 1930, and 10 since 1960. Most of these species names have received no further attention after they were reclassified as C. albicans synonyms on the basis of morphologies and physiologies indistinguishable from the species defined by Berkhout. A notable exception is Candida stellatoidea. This was first published in 1938 as a new species, Monilia stellatoidea, which did not utilize sucrose as a carbon source, formed few chlamydospores on corn meal agar and was nonpathogenic for rabbits (Jones & Martin, 1938). It was formally renamed as C. stellatoidea the following year (Langeron & Guerra, 1939). In 1979, DNA reassociation studies showed clearly that isolates of C. stellatoidea were conspecific with C. albicans (Meyer, 1979), a finding supported by confirmatory DNA reassociation data (Kamiyama et al., 1989) and by antigenic comparisons (Montrocher, 1980). However, many publications continued to list C. stellatoidea at least as a variety, if not still a species separate from C. albicans.

Kwon-Chung. (1988) showed that isolates of C. stellatoidea could be divided into two subtypes on the basis of electrophoretic karyotype profiles. Type I isolates constituted authentic examples of C. stellatoidea sensu strictu: negative for sucrose assimilation and with low mouse virulence, whereas type II isolates were essentially sucrose-negative variants of C. albicans. The identity of C. stellatoidea type II with C. albicans was confirmed by the demonstration that type II isolates could be induced to revert to growth on sucrose (Kwon-Chung et al., 1990). However, it was subsequently found that reversion to sucrose assimilation could also be induced in type I C. stellatoidea, a shift of phenotype that resulted from chromosomal rearrangements (Wickes et al., 1991). This observation finally laid to rest any lingering doubts concerning the conspecificity of C. stellatoidea and C. albicans.

Notwithstanding the acknowledged conspecificity of C. stellatoidea type I with C. albicans, several studies showed that the former were notably distinct from other C. albicans isolates at the DNA level, as shown by cytochrome b sequences (Biswas et al., 2001), multi-locus enzyme electrophoresis (Pujol et al., 1997), restriction fragment length polymorphisms (Magee et al., 1987), electrophoretic karyotyping (Kwon-Chung et al., 1988, 1989) and electrophoretic pattern of tRNAs (Santos et al., 1994).

Multi-locus sequence typing (MLST), in which strain differences in single nucleotide polymorphisms are determined for fragments of six or seven housekeeping genes, has now been extended from its original applications in bacteriology (Maiden, 2006) to provide highly discriminatory strain typing schemes for several pathogenic fungi, including C. albicans (Bougnoux et al., 2002, 2003; Tavanti et al., 2003). The MLST approach has so far been applied mainly to fresh clinical isolates of C. albicans and has been used to provide information on the epidemiology (Bougnoux et al., 2004, 2006; Odds et al., 2006) and phylogenetics (Tavanti et al., 2004, 2005; Aliyu et al., 2006; Chen et al., 2006) of the species. In this study, we have used MLST to type authentic isolates of strains now regarded as C. albicans but originally described as distinct species, including four isolates originally identified as C. stellatoidea.

Materials and methods

Yeast isolates

The material for this study comprised 35 yeast isolates (Table 1). They included C. albicans SC5314 and WO-1, the two strains that have been used for whole genome sequencing, C. albicans CBS562, the type strain of the species, and 32 other isolates representing yeasts that were originally reported as species other than C. albicans, but which are currently regarded as synonyms of C. albicans. Three strains identified as C. stellatoidea type I were the kind gift of Dr K.J. Kwon-Chung. The remaining isolates, including the type strains for C. stellatoidea and Candida claussenii, came from the collection of the Centraalbureau voor Schimmelcultures in Utrecht, the Netherlands. A further five isolates from our C. albicans collection were tested for sucrose assimilation. These were IHEM17984, HK03M120736, AM2005/0411, JIMS500006 and AM2005/0411. The first four of these were also included in a previous publication on MLST (Odds et al., 2007).

View this table:
1

Details of yeast strains typed and results of MLST, ABC typing and MTL typing

Reference no.Details/alternative namesOriginal source/ country of originStrain typing data
DSTABCMTLClade
B4257Candida stellatoidea; ≡CBS 8190, ATCC 36232USA1031BhetS
B4404Candida stellatoideaWS Riggsby collection1031BhetS
B4406Candida stellatoideaWS Riggsby collection1031BhetS
CBS562Type strain of Candida albicans (Robin) Berkhout; ≡ATCC18804Interdigital lesion1030Bhet5
CBS1899Type strain of Candida truncata VanbreuseghemSkin/Republic of Congo124Bhet4
CBS1905Neotype of Candida stellatoidea (Jones & Martin) Diddens & Lodder; ≡ATCC11006Vagina/USA1031BhetS
CBS1912Type strain of Candida langeronii Dietrichson ex van Uden & BuckleySputum/Norway119Ahet2
CBS1949Type strain of Candida claussenii Lodder & Kreger-van Rij; type strain of Syringospora claussenii Van der Walt; ≡ATCC188141048Bhet5
CBS2312Received as Monilia butantanensis; ≡ATCC28776Lung1032Bhet3
CBS2689Type strain of Mycelorrhizoides gruetzii OtaInterdigital/Germany1033Chet7
CBS2690Type strain of Cryptococcus copellii Froilano de MelloTongue1049Bhet3
CBS2691Type strain of Monilia tumefaciens-albus Fullerton; formerly named Saccharomyces tumefaciens-albusPharyngitis1034Aα/α1
CBS2692Type strain of Myceloblastanon favrei OtaDermatitis/Germany1035Bα/α3
CBS2695Syntype of Monilia psilosis AshfordSprue/Puerto Rico1036Ahet2
CBS2696Type strain of Mycotoruloides ovalis Langeron & TaliceOropharynx/Germany170Ahet1
CBS2697Type strain of Mycotoruloides triadis Langeron & TaliceSputum/France170Ahet1
CBS2698Type strain of Blastodendrion erectum Langeron & Talice; Endomyces albicans SabouraudOral thrush/France1037Chet5
CBS2700Type strain of Monilia aldoi Pereira; also named Mycotoruloides aldoi; Candida aldoiTongue/Brazil1038Ahet15
CBS2702Type strain of Candida desidiosa Ciferri & RedaelliPigeon droppings/Italy1039Ahet1
CBS2703Type strain of Candida mycotoruloidea Redaelli & CiferriThroat/Italy1040Aa/a1
CBS2704Type strain of Cryptococcus pinoysimilis Castellani; earlier called Mycocandida pinoysimilisSkin lesion69Ahet1
CBS2705Type strain of Mycotorula verticillata Redaelli & CiferriDermatitis/Italy840Aa/a7
CBS2706Type strain of Monilia periunguealis NiñoNail/Argentina1041Ahet2
CBS2707Type strain of Monilia alvarezsotoi Mazza & NiñoSkin1071Ahet15
CBS2710Type strain of Blastodendrion oosporoides ZachNail/Austria1042Chet5
CBS2712Type strain of Mycotorula sinensis ReissSputum/China1072Ahet13
CBS5137Type strain of Syringospora stellatoidea Van der Walt; ≡ATCC32077Sputum/Netherlands69Ahet1
CBS5144Type strain of Candida intestinalis Batista & SilveiraFaeces/Brazil1044Bhet3
CBS5145Type strain of Candida biliaria Batista & SilveiraBile/Brazil344Bhet3
CBS5703Types strain of Procandida grubyii Novák & VitézSputum/Hungary1043Ahet7
CBS5736Neotype of Syringospora albicans (Robin) Dodge (designated Van der Walt, 1970)Vagina/South Africa918Ahet9
CBS6552Type strain of Candida nouvelii SaëzPharynx of Cephalophus dorsalis/France1045Aa/a15
CBS8781Type strain of Candida africana TietzBalanitis/Germany182Ahet13
SC5314Candida albicans strain used for whole genome sequencingGeneralized Candida infection/USA52Ahet1
WO-1Candida albicans white–opaque switcher, used for whole genome sequencingBlood isolate/USA383Aα/α6

Strain typing and susceptibility testing

The yeasts were submitted to typing by MLST and for homozygosity at the mating-type locus (MTL) as previously described (Bougnoux et al., 2003; Tavanti et al., 2003). The result of MLST was expressed as a diploid sequence type (DST) and DSTs were assigned to clades according to criteria defined previously (Odds et al., 2007). MTL typing led to designation of strains as a/α, a/a or α/α according to their heterozygosity or homozygosity at the MTL. ABC typing by PCR (McCullough et al., 1999) was used to detect the presence or absence of an intron in the ITS1 region of DNA encoding rRNA. Strains without the intron in both alleles are designated type A; those with the intron in both alleles are type B and strains with a heterozygous distribution of the intron are type C. This PCR would also have detected any strains that were examples of Candida dubliniensis rather than C. albicans (McCullough et al., 1999) but none were found. The susceptibility of the yeast panel was determined to amphotericin B, fluconazole, itraconazole, posaconazole, voriconazole, flucytosine and caspofungin by the EUCAST microdilution method (Cuenca-Estrella et al., 2007) with spectrophotometric endpoints read at 24 and 48 h.

Assimilation tests

Sucrose assimilation by yeast isolates was tested by inoculation of 5 mL volumes of yeast nitrogen base with added 1% sucrose or glucose, sterilized by membrane filtration. The inoculated tubes were incubated at 30 °C with constant rotation at 20 r.p.m. at an angle of 4° from the horizontal, to provide continuous turbulence. Tubes were inspected after 24 and 48 h. Sucrose assimilation was recorded as negative when growth in the tube containing glucose was apparent as strong turbidity but minimal or no turbidity was evident in the tube containing sucrose.

Statistical analysis

A dendrogram for 1516 C. albicans isolates from different sources or from the same source but representing different DSTs was constructed from the database of MLST results (http://test1.mlst.net/) in August 2007. The 1516 isolates included the 35 in the present study. The dendrogram was based on the unweighted pair-group method using arithmetic averages (UPGMA) and determined by p-distance as implemented by mega3 software (Kumar et al., 2004).

Results

Typing of C. albicans synonym isolates

Detailed strain typing results for the panel of 35 yeasts are given in Table 1. The 35 yeasts represented 30 different DSTs and 10 different clades. Several of the DSTs were novel and details have been added to the online MLST database (http://test1.mlst.net/). By ABC typing, 20 isolates in the panel were type A, 12 were type B and three were type C. Six of the yeasts were homozygous at the MTL: three were a/a and three were α/α. All the isolates were susceptible to all the antifungal agents tested, with the exception of CBS2712, which had intermediate flucytosine susceptibility (MIC=16 μg mL−1).

The relationship of the yeasts to other C. albicans strains was demonstrated by the UPGMA dendrogram (Fig. 1). Seven of the C. albicans synonyms and strain SC5314 were members of clade 1: CBS2704 and CBS5137 were DST 69, the most common DST for C. albicans isolates globally (Odds et al., 2007). Three of the synonyms were members of clade 2, five of clade 3 and one belonged to clade 4 (Table 1 and Fig. 1). None of the strains fell into clade 11, the fifth most populous clade and one which is dominated by European isolates of C. albicans (Odds et al., 2007). The C. albicans type strain CBS562 and three other isolates coclustered in clade 5. Three isolates clustered in minor clade 15. One isolate, CBS2712, coclustered with the Candida africana type strain MYA-2669 in clade 13.

1

UPGMA dendrogram for 1516 Candida albicans isolates. The paths highlighted are for the 35 yeasts listed in Table 1: the shape of the rest of the diagram is sketched in gray shadow. The position of the four major C. albicans clades is indicated.

The four isolates of C. stellatoidea, including the species type strain, formed a cluster of indistinguishable strains with DST 1031. In the UPGMA dendrogram, they were well separated from most other C. albicans strains (Fig. 1). Their closest neighbors in the full 1516-isolate UPGMA dendrogram (Fig. 2) were isolates P2216 and P2246, originally identified as C. africana. Only these six isolates coclustered at a similarity with p-distance <0.04, the previous arbitrary cut-off for clade distinctions (Odds et al., 2007). All six isolates in this small cluster were negative in tests for sucrose assimilation. As shown in Fig. 2, a further 15 isolates coclustered with the six sucrose-negative isolates beyond the limit of p-distance <0.04. For all but five of these isolates, the MLST data came from other laboratories or from our own laboratory on the basis of a DNA sample, not the live isolate; hence we had no material for conducting a sucrose assimilation test. For the five isolates in our collection, sucrose assimilation was positive. CBS5137, nominally the type strain of Syringospora stellatoidea, considered by Van der Walt (1970) to be the teleomorph of C. stellatoidea, was also negative for sucrose assimilation, but was DST 69, the most common C. albicans type. We have tested sucrose assimilation for two other randomly chosen DST 69 isolates: both were sucrose-positive.

2

Detail for the portion of the dendrogram at the bottom of Fig. 1. This shows the positions of the six sucrose-negative isolates (designated Candida stellatoidea or Candida africana) and the other Candida albicans isolates within this region of the dendrogram. An asterisk indicates the isolate was not available to us for study: MLST data were provided from other laboratories or were determined by our laboratory from DNA samples only.

Discussion

Our study confirms the existing phenotypic and genetic data that reclassified the diverse putative novel yeast species we studied as synonyms of C. albicans (Robin) Berkhout. All 32 synonymous isolates were typeable with C. albicans MLST PCR primers and coclustered with existing C. albicans isolates in the UPGMA dendrogram based on MLST results. We have tested four isolates of C. dubliniensis with the same set of PCR primers (unpublished data). All four formed PCR products readily with the ADP1 primers and, with prolongation of the PCR cycles, with the MPIb primers. Reactions were weak or nonexistent with the other five sets of primers. Although the MLST primer sets were designed for typing strains within a single species and not for species identification, if isolates of such a closely related species as C. dubliniensis (Fitzpatrick et al., 2006) react only with a subset of those primers, it seems unlikely that isolates of any species other than C. albicans would readily form PCR products with the full primer set. Our results therefore lend no support to any possibility of reclassifications for the alternative species names represented in our isolate panel.

Our study reveals that the species type strain for C. albicans, CBS562, is neither a member of the most populous C. albicans clade (clade 1) nor the most commonly encountered ABC type (type A). With the DNA sequence information now available, a more rational choice for a C. albicans type strain would be an example of DST 69, type A, to represent the most common C. albicans strain type known (Odds et al., 2007). Of course, once designated in a Latin description, strain types retain their authority as representatives of the phenotypic properties of a species: however, the level of detail to which DNA from a single isolate can now be scrutinized means that isolates other than phenotypically designated strain types are more important tools for experimental investigation. The two C. albicans strains that have so far been the subject of whole genome sequencing were chosen over and above the species strain type.

Isolates at one time designated as C. africana have been described in several publications concerned with phenotypically atypical strains of C. albicans (Tietz et al., 1995, 2001; Forche et al., 1999). Most such isolates have already been shown by MLST to cocluster at a high level of similarity away from the majority of C. albicans strains (Odds et al., 2007). However, two isolates we received as examples of C. africana coclustered in this study with the four isolates originally classed as C. stellatoidea type I (Fig. 2) and, like the C. stellatoidea, failed to grow in broths with sucrose as the sole carbon source – the main phenotypic property that differentiated C. stellatoidea from C. albicans (Kwon-Chung et al., 1989). The portion of the dendrogram shown in Fig. 2 seems to include two strain subclusters on the basis of the sucrose assimilation properties of the isolates (but with the limitation that the majority of these isolates were not available to us for testing). The fate of most C. albicans isolates that were first purported to be new species was to become synonyms of C. albicans with some rapidity. However, C. stellatoidea retained its separate status from C. albicans for 40 years until 1979 when Meyer first proposed conspecificity on the basis of DNA reassociation studies (Meyer, 1979) and was still referred to at least as a variety for many years afterwards. Our data confirm beyond any question that C. stellatoidea is a synonym of C. albicans– the sucrose-negative isolates are unequivocally coclustered with sucrose-positive C. albicans isolates. However, our data extend the previous studies showing a marked genetic difference between C. stellatoidea type I isolates and other C. albicans (Magee et al., 1987; Kwon-Chung et al., 1988, 1989; Santos et al., 1994; Pujol et al., 1997; Biswas et al., 2001). It is, perhaps, most impressive overall that many isolates, earlier regarded by investigators equipped only with experience and phenotypic data as sufficiently different from C. albicans to constitute a separate species, still emerge as a separate strain cluster with high dissimilarity from most of the other C. albicans types. However, the typing of the sucrose-negative isolate originally named as S. stellatoidea as C. albicans DST 69 suggests that this isolate is an example of C. stellatoidea type II, rather than the better demarcated subcluster represented by the six sucrose-negative isolates CBS1905, B4404, B4406, B4257, P2216 and P2246 (Fig. 2).

Acknowledgements

We thank Bart Theelen for skilled technical assistance. M.D.J. is supported by a grant from the European Union (EURESFUN).

Footnotes

  • Editor: Richard Calderone

References

View Abstract