点击显示 收起
Aberdeen Fungal Group, School of Medical Sciences, Institute of Medical Sciences, University of Aberdeen, Aberdeen AB25 2ZD, United Kingdom
HPA Mycology Reference Laboratory, Myrtle Road, Kingsdown, Bristol BS2 8EL, United Kingdom
The Peter Medawar Building for Pathogen Research and Department of Zoology, University of Oxford, OX1 3SY, United Kingdom
ABSTRACT
A system is described for typing isolates of the pathogenic fungus Candida tropicalis, based on sequence polymorphisms in fragments of six genes: ICL1, MDR1, SAPT2, SAPT4, XYR1, and ZWF1a. The system differentiated 87 diploid sequence types (DSTs) among a total of 106 isolates tested or 80 DSTs among 88 isolates from unique sources. Replicate isolates from the same source clustered together with high statistical similarity, with the exception of one isolate. However, a clade of very closely related isolates included replicate isolates from three different patients, as well as single isolates from eight other patients. This clade, provisionally designated clade 1, was one of three clusters of isolates with high statistical similarity. Five of six isolates in one cluster that may acquire clade status were resistant to flucytosine. This study adds C. tropicalis to Candida albicans and Candida glabrata as Candida species for which a multilocus sequence typing (MLST) system has been set up. The C. tropicalis MLST database can be accessed at http://pubmlst.org/ctropicalis/.
INTRODUCTION
Strain typing by sequencing of several housekeeping genes in a microbial species (multilocus sequence typing; MLST) has rapidly developed as a reliable technology for epidemiological studies of infectious disease (5, 23). The method involves determination of DNA sequence polymorphisms between isolates with a set of fragments of five to seven genes, which are ideally under neutral selective pressure and chromosomally dispersed. The data obtained are highly reproducible, amenable to statistical analyses to quantify similarities and putative genetic relationships between isolates, and able to be stored in a single central database for global internet access.
Among fungal diseases, deep-seated Candida infections are the most commonly encountered opportunistic problems that affect seriously immunocompromised or debilitated hosts. Superficial Candida infections are responsible for considerable morbidity among neonates, the elderly, and patients with AIDS (oral infections) and among women of childbearing age (vulvovaginal infections). Candida albicans accounts for the majority of these infections, but other Candida species, particularly C. glabrata, C. parapsilosis, and C. tropicalis, are by no means uncommon causes, with some authorities documenting a rise in prevalence of the latter at the expense of C. albicans. There is therefore a clear need for strain typing of these species for epidemiological purposes. MLST technology has been developed for Candida albicans (3, 4, 22) and C. glabrata (7). MLST could not be used for C. parapsilosis because of the paucity of allelic polymorphisms in this species (20). The MLST approach provides for definition of population structures within a species and can reveal differences in geographical origins, anatomical sources, and other properties between clades of closely related isolates (2, 3, 7, 19, 21).
We have now developed MLST for the species C. tropicalis, a species that has accounted for 5 to 30% of Candida bloodstream infections during the past 15 years (17). C. tropicalis probably has an entirely diploid genome (1, 8, 13) and, as with C. albicans, MLST can take advantage of the additional polymorphisms that are due to allelic heterozygosities.
MATERIALS AND METHODS
Isolates. All but 2 of the 106 C. tropicalis isolates (Table 1) were originally cultured from clinical material; the exceptions were two isolates from the stomach contents of a chameleon. All isolates were reidentified by standard morphological and physiological criteria. Historical isolates came from our collection of pathogenic fungi and fresh isolates beyond those we received for routine test purposes came from the Mycology Reference Laboratory in Bristol, United Kingdom, and the Women's and Children's Hospital, Adelaide, Australia. The sources for 30 isolates were 10 patients and 1 animal from whom two or more cultures were obtained from different anatomical sites or at different times. A set of 88 single-source isolates was therefore available, including one chosen randomly from each of the sets of multiple isolates. The isolates represented probable genetic diversity based on their date, anatomical site, and geographical source of isolation. The yeasts were maintained on Sabouraud agar (Oxoid, Basingstoke, United Kingdom).
Choice of loci for MLST. Initially, 13 gene fragments were chosen for a pilot MLST study of 20 C. tropicalis strains. We reduced the set to the minimum of six gene fragments (Table 2) which, in combination, yielded the highest interstrain discrimination (largest number of different strain types). One fragment was obtained with primers described for the C. albicans ZWF1a gene encoding glucose-6-phosphate dehydrogenase (22), which amplified products from several other Candida species; the remainder were C. tropicalis genes whose sequences were obtained from the GenBank database. Primers were designed to amplify gene fragments of 500 to 750 bp and are also described in Table 2, with the corresponding PCR product sizes given.
DNA extraction. Genomic DNA was extracted from yeasts grown in YPD broth comprising 2% glucose, 2% mycological peptone (Oxoid), and 1% yeast extract (Difco, Detroit, MI). Briefly, cells were harvested in the stationary phase and lysed by vortexing the pellet for 5 min with 0.3 g of glass beads (0.45 to 0.52 mm in diameter; Sigma, St. Louis, MO) in 200 μl of buffer (100 mM Tris HCl, pH 8.0, containing 2% Triton X-100, 1% sodium dodecyl sulfate, 1 mM EDTA) and 200 μl of 1:1 (vol/vol) phenol-chloroform solution. After vortexing, 200 μl of TE (1 mM EDTA, 10 mM Tris-HCl, pH 8.0) was added to the lysate; the mixture was microcentrifuged at full speed for 5 min, and the aqueous phase was transferred to a new tube. DNA was precipitated by addition of 1 ml of ethanol to the supernatant. Samples were centrifuged for 2 min, and the pellet was resuspended in 400 μl of TE containing 100 μg of RNase (10 μl of a 10-mg/ml solution; Sigma). The mixture was incubated for 1 h at 37°C, and then DNA was precipitated with 1 ml of isopropanol and 10 μl of 3 M sodium acetate, dried, and redissolved in 50 μl of TE, pH 8.0.
Amplification and nucleotide sequence determination. PCRs were used to amplify the gene fragments listed in Table 2. Reaction volumes of 50 μl contained 100 ng of genomic DNA, 2.5 U Pfu DNA polymerase (Promega, Madison, WI), 5 μl of 10x buffer (supplied with the enzyme), 200 μM deoxynucleoside triphosphates (dNTPs) (Promega), and 10 μM forward and reverse primers. A Flexigene thermocycler (Techne, Cambridge, United Kingdom) was set up with a first cycle of denaturation for 7 min at 94°C followed by 30 cycles of denaturation at 94°C for 1 min, annealing at 52°C for 1 min, and elongation at 74°C for 1 min 5 s, with a final extension step of 10 min at 74°C. The amplified products were precipitated in microdilution plates. Briefly, 60 μl of a 20% polyethylene glycol (Sigma)-2.5 M NaCl solution was added to each well containing 40 μl of PCR product. The microdilution plate was then sealed, vortexed, incubated at room temperature for 30 min, and centrifuged for 1 h at 2,250 x g (4°C). The supernatant was discarded, and the plate was inverted onto a piece of 3-mm chromatography paper and centrifuged again at 500 x g for 1 min to remove any residual polyethylene glycol from the wells. Pellets were washed with 150 μl of 70% ethanol, precipitated as described above, and resuspended in 60 μl of sterile water. Both strands of purified gene fragments were sequenced on an ABI 3700 DNA analyzer (Foster City, Iowa) with a 2.5 μM concentration of the same primers that were used in the PCR step. The sequence data were coupled with DNASTAR software. Heterozygosities were defined by the presence of two coincident peaks in the forward and reverse sequence chromatograms. The one-letter code for nucleotides from the International Union of Pure and Applied Chemistry (IUPAC) nomenclature was used to define results.
Statistical analysis of MLST data. Phylogenetic analyses by unweighted pair-group method with arithmetic average (UPGMA) and neighbor-joining algorithms were conducted with MEGA version 2.1 (14) applied to modified sequence data. The analyses were based only on the results for variable loci to maximize their power to discriminate between isolates; the data set of only the variable bases was suitable for pairwise difference analysis, which has been used previously with C. albicans MLST (3, 22). To obtain sequences that could be handled by the MEGA software, which is not programmed to analyze heterozygous code data, the following procedure was used. The results for the variable loci from the six gene fragments sequenced were concatenated into a single sequence. For any pair of isolates, each with a diploid genome, the base at each variable locus could be homozygous and identical between the isolates, heterozygous and identical, homozygous and different, or heterozygous in one isolate and homozygous in another. For example, the sequencing result (in the IUPAC single-letter code) for a given locus across a set of strains might appear as A, T, or W (= A+T). Data from the variable loci from the six C. tropicalis alleles were therefore conjoined into a single sequence, and then each base in the sequence was rewritten twice for a homozygous (A, C, G, or T) datum or as the two component bases for a heterozygous (K, M, R, S, W, Y) datum. These revised sequences could then be used to generate an unrooted UPGMA dendrogram based on pair differences in MEGA 2.1. In this form, they were the functional equivalent of scoring a pair of results as 1 for homozygous or heterozygous identical data, 0 for homozygous different data, and 0.5 when one allele had a heterozygous result and the other a homozygous result and then creating a difference matrix. The significance of the cluster nodes was determined by bootstrapping with 1,000 randomizations. The same data were also used to generate unrooted neighbor-joining trees based on p-distance, also with 1,000 random bootstrapping operations to determine significance of nodes.
The eBURST package (http://eburst.mlst.net/) (9) was used to determine putative relationships between isolates. This software scans pairs of alleles and records isolates as related when five of the six alleles are identical between a pair. The eBURST algorithm places all related isolates into clonal complexes and, where possible, predicts the founding, or ancestral diploid sequence type (DST) of each complex. The output is a display of the most parsimonious patterns of descent of each DST from the ancestral type.
Discriminatory power was calculated according to Hunter (11).
RESULTS
C. tropicalis strain differentiation by MLST. The six fragments sequenced allowed for differentiation of 87 DSTs among 106 isolates (Table 1), indicating a discriminatory power of 0.994. For the set of 88 isolates from unique sources used to determine population structure (below), 80 DSTs were found. Of the six gene fragments used for MLST, XYR1 showed the highest typing efficiency, distinguishing 3 genotypes per polymorphism for only 11 polymorphic sites (Table 3), and SAPT2 was the least efficient fragment in terms of genotypes per polymorphism (Table 3). For all of the fragments, the ratios of nonsynonymous to synonymous amino acid changes resulting from the sequence polymorphisms were less than 1, indicating the genes were under neutralizing selective pressure (Table 3).
Only four of the C. tropicalis isolates, WC04-200748 and WC04-200749 (both from the same patient), L634, and AM2005/0005 showed variations at large numbers of the polymorphic sites identified. These variations were mostly seen as heterozygosities at multiple loci rather than as homozygous differences from other isolates, and they were particularly abundant in the SAPT2 and SAPT4 fragments. The majority of the isolates were discriminated on the basis of a smaller number of polymorphisms in each fragment than is implied by the data in Table 3.
Reproducibility of MLST data. A total of 10 different C. tropicalis isolates were submitted for full or partial MLST in duplicate on different occasions and blinded to the person conducting the sequencing. The duplicate tests involved a total of 1,286 fragment comparisons, of which 10 (0.8%) gave different results between duplicate tests. In eight cases, the difference resulted from a single-locus homozygous/heterozygous difference, and in two cases there were homozygous/heterozygous discrepancies at two sites in a single fragment. The sequence reproducibility was therefore better than 99%.
Nucleotide polymorphisms and amino acid changes. The 142 nucleotide polymorphisms among the six sequenced fragments resulted in 24 nonsynonymous changes in amino acids encoded by sequence-variable triplets. Of these, 18 changes were nontrivial (e.g., acidic to basic side chains, aliphatic to aromatic side chains). In ZWF1a, the only nonsynonymous amino acid change was between a tyrosine residue and a stop codon. However, all isolates with the ZWF1a stop codon polymorphism were heterozygous at the site. In the SAPT2 sequence, a switch between tyrosine and a stop codon also resulted from the polymorphism at position 125 in the sequence. Only 2 of the 99 isolates sequenced encoded the stop codon, but in both cases, the alleles were homozygous for the stop codon.
Similarity of isolates from the same source. A neighbor-joining dendrogram for all 30 duplicate and multiple isolates from single sources is shown in Fig. 1. Twelve isolates from three different patients coclustered with very high similarity (topmost cluster in Fig. 1). Ten isolates from sources referred to as patients 1 and 2 in Table 1 were all superficial surveillance isolates from individuals undergoing chemotherapy for hematological malignancy in the same hospital in the 1980s (15); the other two were blood and urine isolates from a patient in a different hospital in 2004. Isolate L634 was a further isolate from patient 1, but it did not cluster anywhere close to the remainder of patient 1 isolates. The considerable difference of L634 from the other patient 1 isolates may have resulted from an error in maintenance of the isolate in our collection. Because of its obvious difference, L634 was treated as an isolate from a separate source for subsequent analysis. All other sets of duplicate and triplicate isolates from individual patients formed separate clusters according to their source, with high isolate similarity confirmed by the high bootstrap values at the dendrogram nodes. The ability of MLST to distinguish possible errors in labeling or storage of isolates in the way we recognized a problem with L634 is a further advantage of the system: we have noted similar "maverick" strain types by MLST in groups of isolates from single sources with our C. albicans system.
Population structure of 88 isolates from separate sources. eBURST analysis of the genotypes and DSTs for 88 C. tropicalis isolates from separate sources revealed one cluster of 8 DSTs and another of 6 DSTs as the only related sets of any size to emerge from this analysis. There were clusters with three and four DSTs and four pairs of related DSTs, but most of these comprised multiple isolates from the same patients; the majority of isolates appeared as unrelated singletons.
A UPGMA dendrogram based on pairwise differences between MLST sequences (Fig. 2) indicated a population structure for C. tropicalis isolates that included three clades of isolates that were reasonably robust (bootstrap values of at least 68%) and closely related (within a p-distance of 0.02 or less). These clusters were provisionally designated C. tropicalis clades 1 through 3. All but two of the isolates in the group designated clade 1 were in the largest eBURST clonal cluster. The other two UPGMA clusters also correlated with eBURST data. Among all 106 isolates typed in this study, a total of 20 isolates from 11 patients belonged to provisional clade 1. Six isolates, loosely clustered on the UPGMA dendrogram, were all resistant to flucytosine (Fig. 2), but the low bootstrap value for the node of this cluster precludes its designation as a putative clade.
Properties of isolates in different clades. Numbers of isolates in the three definable clades were too small to permit statistical analysis of isolate properties by clade. Fifty-six (59.7%) of the 106 isolates shown in Fig. 2 came from the United Kingdom. Clade 1 was dominated by isolates from the United Kingdom (8 of 11 isolates), and there were no clade 1 isolates from North America, whereas clades 2 and 3 showed a small majority (4/6 and 5/9, respectively) of isolates from other countries. The distribution of isolates from blood or other sterile source, among all isolates where the source was known, was uneven: 1/9 in clade 1, 2/4 in clade 2, and 3/5 in clade 3. The set of loosely clustered flucytosine-resistant isolates (Fig. 2) was the only group in the set tested that showed a possible cluster-related property. There was no conspicuous relationship between individual alleles and flucytosine resistance.
DISCUSSION
Our data show it is possible to differentiate isolates of C. tropicalis with a high degree of reproducibility by multilocus sequence typing with the set of six gene fragments described in this study. The six genes chosen meet the requirement for MLST fragments to encode a low ratio of nonsynonymous to synonymous amino acid changes, suggesting they are not under selective pressure (12). Only two of the fragments yielded more than one genotype per polymorphic locus (Table 3). However, these data are biased by the unusual occurrence of a small subset of isolates showing polymorphisms at many more sites in some of the gene fragments than the rest of the isolates. If the five isolates AM2003/0076, AM2005/0005, L634, WC04-200748, and WC04-200749 are excluded from the analysis in Table 3, the genotype/polymorphic site ratio rises to 1.7 for ICL1, 2.2 for MDR1, 2.3 for SAPT2, 1.4 for SAPT4, and 2.1 for ZWF1a, with the ratio for XYR1 only slightly increased to 3.7. These revised ratios approach those seen with C. albicans MLST (4). The unusually polymorphic isolates may represent a subtype within the species C. tropicalis but are unlikely to constitute a possible separate species, since they form PCR products of the same size as all the isolates with all six primers. A totally different situation was encountered with subtypes of C. parapsilosis when we attempted to devise an MLST system for strains within the species and found two stable subtypes that formed products with only a minority of primers and were confirmed as separate species (20). The overall discriminatory power of the C. tropicalis MLST system is very high: at >99% probably greater than that of other fungal strain typing systems (10, 18).
As with C. albicans (19), replicate C. tropicalis isolates show high levels of similarity on an individual patient basis and high levels of difference between patients. The exception concerns isolates in the highly similar cluster here provisionally designated clade 1 (Fig. 2), where isolates from three patients could not be separated by analysis of MLST data. Within C. albicans MLST clade 1, a similar subset of indistinguishable isolates was found which also came from several patients. For both species, there are therefore a particularly common DST and a set of very closely related DSTs that constitute a clonal cluster by eBURST analysis and may represent the strain type best adapted to coexistence with humans. Subsequent clonal reproduction of the species leads to generation of mutants with close affinities to this "optimized" ancestral type.
The major exception to the normal tendency of isolates from the same patient to emerge as highly similar by MLST was L634, a vaginal surveillance isolate from patient 1 in a study published 20 years ago (15). The isolates from this patient have been stored in a collection that has been transported from England to Belgium and thence to Scotland, thus increasing the possibility that cross-contamination or other error means this isolate is not the original one. However, L634 was the first example we encountered of the small subset of isolates with high numbers of (usually heterozygous) polymorphisms in the six gene sequences used for MLST, and the other four examples of such isolates were received subsequent to the first sequencing of L634. The isolate typed as L634 must therefore have already existed in our collection, whether or not its true origin was the same patient as for the other isolates shown as originating from "patient 1" in Fig. 1. Isolates with such unusually high multiple sequence heterozygosities have not been encountered in our extensive experience of C. albicans MLST. We shall prospectively investigate the possibility that such isolates can be generated from less heterozygous strains by application of stresses such as antifungal exposures.
The sizes of the three definable clades in the present study were too small to afford insights into possible associations with geographical areas or sites of infection; indeed, the diversity of DSTs as seen by eBURST analysis is greater than was encountered in MLST surveys of smaller isolate sets of C. albicans (3, 21) and C. glabrata (7). Nevertheless, the observation of a loose cluster of flucytosine-resistant isolates is intriguing, and we are now investigating the possibility that a common mutation underlies flucytosine resistance in this set in the same way as was found for C. albicans clade 1 isolates (6, 19).
The relationships between C. tropicalis strain population structure and clinically relevant properties of clades will inevitably become clearer as the database of typed isolates (http://pubmlst.org/ctropicalis/) expands. Meanwhile we have added C. tropicalis to the growing list of Candida species for which MLST, with its advantages of reproducibility and portability, is now available.
ACKNOWLEDGMENTS
This study was supported by grants from the Wellcome Trust (069615, 074898).
We are grateful to the many colleagues who have supplied us with C. tropicalis isolates, in particular C. C. Kibbler and D. H. Ellis for several recent clinical isolates.
REFERENCES
Blandin, G., O. Ozier-Kalogeropoulos, P. Wincker, F. Artiguenave, and B. Dujon. 2000. Genomic exploration of the hemiascomycetous yeasts. 16. Candida tropicalis. FEBS Lett. 487:91-94.
Bougnoux, M.-E., A. D. M., S. Morand, M. Theraud, B. G. Spratt, and C. d'Enfert. 2004. Multilocus sequence typing of Candida albicans: strategies, data exchange and applications. Infect. Genet. Evol. 4:243-252.
Bougnoux, M.-E., S. Morand, and C. d'Enfert. 2002. Usefulness of multilocus sequence typing for characterization of clinical isolates of Candida albicans. J. Clin. Microbiol. 40:1290-1297.
Bougnoux, M. E., A. Tavanti, C. Bouchier, N. A. R. Gow, A. Magnier, A. D. Davidson, M. C. J. Maiden, C. d'Enfert, and F. C. Odds. 2003. Collaborative consensus for optimized multilocus sequence typing of Candida albicans. J. Clin. Microbiol. 41:5265-5266.
Chan, M. S., M. C. J. Maiden, and B. G. Spratt. 2001. Database-driven multi locus sequence typing (MLST) of bacterial pathogens. Bioinformatics 17:1077-1083.
Dodgson, A. R., K. J. Dodgson, C. Pujol, M. A. Pfaller, and D. R. Soll. 2004. Clade-specific flucytosine resistance is due to a single nucleotide change in the FUR1 gene of Candida albicans. Antimicrob. Agents Chemother. 48:2223-2227.
Dodgson, A. R., C. Pujol, D. W. Denning, D. R. Soll, and A. J. Fox. 2003. Multilocus sequence typing of Candida glabrata reveals geographically enriched clades. J. Clin. Microbiol. 41:5709-5717.
Doi, M., M. Homma, A. Chindamporn, and K. Tanaka. 1992. Estimation of chromosome number and size by pulsed-field gel electrophoresis in medically important Candida species. J. Gen. Microbiol. 138:2243-2251.
Feil, E. J., B. C. Li, D. M. Aanensen, W. P. Hanage, and B. G. Spratt. 2004. eBURST: inferring patterns of evolutionary descent among clusters of related bacterial genotypes from multilocus sequence typing data. J. Bacteriol. 186:1518-1530.
Hunter, P. R. 1991. A critical review of typing methods for Candida albicans and their applications. Crit. Rev. Microbiol. 17:417-434.
Hunter, P. R. 1990. Reproducibility and indices of discriminatory power of microbial typing methods. J. Clin. Microbiol. 28:1903-1905.
Jones, N., J. F. Bohnsack, S. Takahashi, K. A. Oliver, M.-S. Chan, F. Kunst, P. Glaser, C. Rusniok, D. W. M. Crook, R. M. Harding, N. Bisharat, and B. G. Spratt. 2003. Multilocus sequence typing system for group B Streptococcus. J. Clin. Microbiol. 41:2530-2546.
Kamiryo, T., N. Mito, T. Niki, and T. Suzuki. 1991. Assignment of most genes encoding major peroxisomal polypeptides to chromosomal band V of the asporogenic yeast Candida tropicalis. Yeast 7:503-511.
Kumar, S., K. Tamura, I. B. Jakobsen, and M. Nei. 2001. MEGA2: molecular evolutionary genetics analysis software. Bioinformatics 17:1244-1245.
Odds, F. C., C. C. Kibbler, E. Walker, A. Bhamra, H. G. Prentice, and P. Noone. 1989. Carriage of Candida species and C. albicans biotypes in patients undergoing chemotherapy or bone marrow transplantation for haematological disease. J. Clin. Pathol. 42:1259-1266.
Rex, J. H., C. R. Cooper, W. G. Merz, J. N. Galgiani, and E. J. Anaissie. 1995. Detection of amphotericin B-resistant Candida isolates in a broth-based system. Antimicrob. Agents Chemother. 39:906-909.
Sandven, P. 2000. Epidemiology of candidemia. Rev. Iberoam. Micol. 17:73-81.
Soll, D. R. 2000. The ins and outs of DNA fingerprinting the infectious fungi. Clin. Microbiol. Rev. 13:332-370.
Tavanti, A., A. D. Davidson, M. J. Fordyce, N. A. R. Gow, M. C. J. Maiden, and F. C. Odds. 2005. Population structure and properties of Candida albicans, as determined by multilocus sequence typing. J. Clin. Microbiol. 43:5601-5613.
Tavanti, A., A. D. Davidson, N. A. R. Gow, M. C. J. Maiden, and F. C. Odds. 2005. Candida orthopsilosis and Candida metapsilosis spp. nov., to replace Candida parapsilosis groups II and III. J. Clin. Microbiol. 43:284-292.
Tavanti, A., N. A. R. Gow, M. C. J. Maiden, F. C. Odds, and D. J. Shaw. 2004. Genetic evidence for recombination in Candida albicans based on haplotype analysis. Fungal Genet. Biol. 41:553-562.
Tavanti, A., N. A. R. Gow, S. Senesi, M. C. J. Maiden, and F. C. Odds. 2003. Optimization and validation of multilocus sequence typing for Candida albicans. J. Clin. Microbiol. 41:3765-3776.
Urwin, R., and M. C. J. Maiden. 2003. Multi-locus sequence typing: a tool for global epidemiology. Trends Microbiol. 11:479-487.