1 Department of Biology, McMaster University, Hamilton, Ontario, Canada L8S 4K1
2 Department of Molecular Genetics and Microbiology, Duke University Medical Center, Durham, NC 27710, USA
Correspondence
Jianping Xu
jpxu{at}mcmaster.ca
![]() |
ABSTRACT |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
![]() |
INTRODUCTION |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
C. neoformans is a basidiomyceteous yeast, typically with a haploid nucleus. It has a heterothallic life cycle. The mating type locus has two alternative alleles, a and . Under suitable conditions, yeast cells with opposite mating types can fuse to form dikaryotic hyphae. The terminal cell of a dikaryotic hypha forms a basidium where nuclear fusion and meiosis occur, and where four chains of haploid basidiospores are produced (Kwon-Chung, 1976
). Most current genetic analyses of C. neoformans involve strains of serotype D, as mating between strains of opposite mating types within this serotype occurs readily and karyogamy, meiosis and basidiospore formation proceed normally. However, when strains of serotypes A and D mate in the laboratory, meiotic progeny often contain alleles from both parental strains and are diploid or aneuploid (e.g. Lengeler et al., 2000
). Recent molecular analyses have shown that most environmental and clinical strains of serotype AD are diploid or aneuploid and contain alleles typical of both serotypes A and D (Brandt et al., 1993
, 1995
; Boekhout et al., 2001
; Lengeler et al., 2000
; Xu et al., 2002
). In addition, genealogical analyses indicated that strains of serotype AD are recent hybrids between strains of serotypes A and D and involved multiple hybridization events (Xu et al., 2002
). As in a previous study (Xu et al., 2002
), the terms hybridization and hybrid denote the process or a product of the process, respectively, by which an offspring is generated from a mating between two genetically divergent parental strains (e.g. strains of different breeds, varieties, subspecies, species or genera). Because strains of serotypes A and D diverged millions of years ago (Xu et al., 2000b
), mating between strains of serotypes A and D is referred to as hybridization, comparable to the terminology applied to plants and animals. In contrast, recombination is generally defined as the formation of new combinations of genes in progeny that did not exist in the parents, resulting from the processes of mating, crossing-over, and independent assortment among genes located on different chromosomes during meiosis.
Despite extensive population surveillance and strain typing studies of C. neoformans, several questions regarding its population biology remain unresolved. One question is whether recombination occurs in populations within a serotype. Current evidence indicates that there are clones and clonal lineages within each of the serotypes, consistent with extensive asexual reproduction, clonal dispersal and lack of recombination. For example, epidemiological surveys in four areas of the US revealed only a few multilocus enzyme electrophoretic genotypes, and most genotypes were represented by multiple strains that were often distributed over wide geographical areas (Brandt et al., 1996).
The extensive clonality observed in natural populations of C. neoformans might reflect a bias in sampling. More than 90 % of all clinical and environmental isolates of C. neoformans of serotype D possess the mating type (MAT
), and nearly 100 % of isolates of serotype A are MAT
. For example, among 355 clinical and environmental strains of serotype A, D or AD, which were obtained from four areas in the USA, only one strain contained only the mating type a (MATa) allele at the STE20 gene, which is located within the mating type region, and this strain had serotype AD (Yan et al., 2002
). Clinical isolates of serotype A or D strains are predominantly MAT
, probably because pathogenicity is linked to this locus (Kwon-Chung et al., 1992
). The predominance of MAT
among environmental isolates may similarly reflect a relationship of this locus to fitness, although there is no experimental evidence to support this hypothesis. The second hypothesis is related to haploid fruiting ability of MAT
strains. Under starvation conditions in the laboratory, MAT
strains produce hyphae, basidia and haploid fruiting (Wickes et al., 1996
). Because of their small size, basidiospores are more effectively dispersed than the encapsulated vegetative yeast cells. Thus, prolific haploid fruiting of MAT
strains may explain the dominance of this mating type among environmental isolates that might also engage in true sexual reproduction in the presence of both mating types.
In contrast to the biased mating type ratios in clinical and environmental isolates of serotypes A and D, mating type ratios in strains of serotype AD are generally balanced. This is because most strains of serotype AD possess both MATa and MAT loci (Lengeler et al., 2001
; Yan et al., 2002
). In addition, in one study, a significant proportion (14 of 19) of serotype AD isolates contained the MATa allele from serotype A parents (Aa) (Yan et al., 2002
). Strains of Aa were previously thought to be extinct and were identified only recently (Lengeler et al., 2000
; Viviani et al., 2001
). Because of the balanced ratios of mating type alleles among AD strains and the existence of strains with mating type allele Aa, we decided to test strains of serotype AD for the possibility of recombination for serotypes A and D.
There are several methods to detect recombination in natural populations of microorganisms (Xu & Mitchell, 2003). One method is to compare genealogies of different genes for the same set of strains. In strictly asexual organisms with no sexual recombination, genealogies of different genes are expected to be congruent. Conversely, incongruent gene genealogies would be consistent with recombination. In a previous study (Xu et al., 2002
), we analysed a portion of the LAC gene sequence from each of 14 serotype AD strains. These 14 strains were all from the USA, and 12 were collected from the San Francisco area. In this study, we sequenced a portion (642 nucleotides) of another gene, orotidine monophosphate pyrophosphorylase (URA5) for the same 14 strains and compared genealogies derived from these two genes. URA5 was selected for study because it is present in only one copy in haploid strains of C. neoformans and known to be highly polymorphic (Edman & Kwon-Chung, 1990
; Casadevall et al., 1992
; Xu et al., 2000b
). To strengthen the analyses, we included additional strains of serotypes A and D. The results revealed gene genealogy incongruence, consistent with sexual recombination within populations of both serotypes A and D.
![]() |
METHODS |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
|
A typical PCR reaction contained 10 µl (1 ng) of diluted genomic DNA template, 0·5 unit Amplitaq DNA polymerase, 0·2 µM of each primer and 0·2 mM of each deoxyribonucleotide triphosphate in a total volume of 50 µl. The following PCR conditions were used: 3 min at 94 °C, followed by 40 cycles of 30 s at 94 °C, 30 s at 50 °C and 30 s at 72 °C, and lastly, 7 min of extension at 72 °C. PCR products were cleaned by using Wizard spin columns (Promega) and sequenced using an Applied Biosystems Prism 373 (ABI373) automated sequencer with dRhodamine-labelled terminators (PE Applied Biosystems) following the manufacturer's instructions. However, we were unable to obtain clear URA5 sequences from any of the 14 strains by the direct PCR sequencing method. Similar to those for the LAC gene, this result suggested sequence heterogeneity at the URA5 locus within each strain. We therefore cloned the PCR product from each strain using a pGEM-T cloning kit (Promega) and transformed the cloned PCR products into competent Escherichia coli cells following the manufacturer's instructions. For each of the 14 strains, ten random E. coli colonies were picked, amplified with the URA5 primers, and digested with the restriction enzyme DdeI to screen for different alleles. Here, the term allele refers to a distinct nucleotide sequence from a locus within a strain. The term haplotype indicates a unique sequence in the whole collection of strains. For each strain, two clones representing each of two DdeI restriction digest patterns were sequenced (i.e. four clones total for each strain) to control for DNA sequence variation generated due to PCR, cloning and sequencing. Among the 14 strains, only one (MAS94-0241) showed a single base pair difference between two presumably identical alleles (i.e. two clones with the same DdeI restriction digest pattern). To check the authenticity of these two clones, two additional clones were sequenced. The additional sequences were the same as one of the two original clones and identical to the allele from MAS94-0244, a strain from a different body site of the same patient. These sequences were aligned and optimized and imported to the software PAUP* (Swofford, 2002
).
Data analyses.
Phylogenetic analysis was performed with PAUP* (Swofford, 2002). Maximum parsimony (MP) trees were identified using heuristic searches based on 500 random sequence additions (Swofford, 2002
). Statistical support for phylogenetic groupings was assessed by bootstrap analysis using 1000 replicate datasets (sampled from phylogenetically informative characters) with the random addition of sequences during each heuristic search. This analysis identified phylogenetically distinct and statistically well-supported sequence clusters.
The partition homogeneity (PH) test was used to determine statistical significance of congruence between gene genealogies and to infer recombination within populations of serotypes A and D. Because this test requires that the rates of molecular evolution did not differ significantly among lineages, we first determined if the two genes evolved according to a molecular clock model using sequences from the 14 strains of serotype AD. Maximum-likelihood estimates of the MP trees (Felsenstein, 1981) with and without a molecular clock showed little difference (P>0·2 for both genes), demonstrating that the evolution of LAC and URA5 was not significantly different from molecular clock models. The PH test was then used to compare the gene genealogies of the serotype A and serotype D haplotype clusters (see below).
To determine the relationships among haplotypes of strains of serotype AD and those of strains of serotypes A or D, we included published URA5 sequences from strains of serotypes A and D (Xu et al., 2000b; GenBank accession numbers AF140185AF140217). For strains of serotypes A, B, C and D (and some strains of serotype AD) sequenced previously, direct sequencing using PCR products without cloning revealed no sequence ambiguity, and each strain had only one allele at each of four genes analysed, including URA5 (Xu et al., 2000b
; see also Edman & Kwon-Chung, 1990
; Casadevall et al., 1992
). Since introns in both URA5 and LAC genes from strains of serotypes B and C were difficult to align with those from strains of serotypes A and D (Xu et al., 2000b
), strains of serotypes B and C were not included in the analyses here.
![]() |
RESULTS |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
|
|
As shown in Fig. 1(a), the URA5 gene genealogy supported the conclusion of multiple recent hybrid origins for the 14 strains with serotype AD, as previously determined by the analyses of the LAC gene (Xu et al., 2000b
, 2002
). Each serotype AD strain contained two distinct alleles for both genes, with one allele identical or highly similar to those from strains of serotype D (allele 1) and the other to those from strains of serotype A (allele 2). These serotype-specific haplotype groups were used to analyse genealogical congruence between genes and to infer recombination within populations of serotypes A and D.
Despite the overall similarities between the genealogies from these two genes, there were several differences within each of the two haplotype groups. First, for the same 14 serotype AD strains, there were more URA5 haplotypes than LAC haplotypes within the serotype D group (Xu et al., 2002; Table 2
; Fig. 1
). Specifically, ten haplotypes were found among the 14 sequences for the URA5 gene but only five haplotypes were found for the LAC gene. However, the numbers of haplotypes clustered with serotype A strains were very similar between the two genes (five for LAC and four for URA5).
Second, based on phylogenetic patterns and bootstrap values (Fig. 1a), analysis of URA5 identified at least four hybridization events: (i) one generated strain MAS94-0351; (ii) one generated MAS93-0315 and MAS93-0610, two strains from the same patient; (iii) one generated strains MAS92-0022, MAS92-0793 and MAS92-0855, each from a different patient; and (iv) one generated the remaining eight bybrids. Similar analyses of the LAC gene identified at least three hybridization events (Fig. 1b
, Xu et al., 2002
): (i) one for strains MAS92-0793 and MAS92-0022; (ii) one for strains MAS92-0855 and MAS94-0351; and (iii) one for the remaining ten strains. The combined analyses would suggest at least five hybridization events: (i) one for strain MAS94-0351; (ii) one for strain MAS92-0855; (iii) one for strains MAS93-0315 and MAS93-0610; (iv) one for strains MAS92-0022 and MAS92-0793; and (v) one for the remaining eight strains. It should be noted that there are minor sequence differences among AD strains generated by each proposed hybridization event (Table 2
).
Evidence for recombination within both serotype A and serotype D haplotype groups
Phylogenetic analyses identified gene genealogy incongruence between the two genes within both A and D haplotype groups (Fig. 1). To determine the statistical significance of the observed incongruence, a partition homogeneity (PH) test was performed for the two haplotype groups: one group included all the 14 allele 1's of both genes from each serotype AD strain (i.e. those that clustered with the serotype D sequences) and the other included the 14 allele 2's of both genes from each serotype AD strain (i.e. those that clustered with the serotype A sequences). The PH test of each group showed that the two genealogies were significantly incongruent for both A and D serotype clusters (ILD<0·05 for both groups). This result is consistent with recombination within each of the two serotypes. Visual inspection determined that the genealogical incongruence was caused by the following strains: MAS94-0351, MAS93-0315, MAS93-0610 and MAS92-0855. When these strains were excluded from the PH tests, the two genealogies were congruent (ILD>0·2).
![]() |
DISCUSSION |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Previous population studies indicated a predominantly clonal population structure of C. neoformans. Evidence of a clonal population structure included: (i) over-representation of certain genotypes; (ii) lack of segregation among alleles at different loci; and (iii) overall congruence of gene genealogies (e.g. Brandt et al., 1996; Xu et al., 2000b
). However, gene genealogical analyses in our earlier study also revealed hybridization and recombination among serotypes (Xu et al., 2000b
). Because those earlier samples were from geographically diverse areas around the globe and included only one or two strains of each serotype from each area, the question of whether recombination occurred within a serotype could not be addressed (Xu et al., 2000b
). Though the possibility of recombination within serotypes has been suggested [Franzot et al., 1999
; Taylor et al. (1999)
based on data from Brandt et al. (1996)
; Xu et al. (2000b)
], the evidence was largely circumstantial and inconclusive.
This study differs from previous reports on natural populations of serotypes A and D. First, the 14 strains analysed here are unique. Previous population studies of serotypes A and D focused on strains that were serotype A or D, but not serotype AD. Intuitively, that was logical. However, natural samples of serotype A or serotype D are often greatly skewed, favouring MAT. Such biased mating type ratios could manifestly contribute to the appearance of a clonal population structure. In contrast, the serotype AD strains used here have a balanced mating type ratio, with most strains having both MATa and MAT
(Table 1
). The chances of detecting recombination should be greater in samples with comparable numbers of each mating type. Second, 12 of these 14 serotype AD strains were from one geographical region (the San Francisco area; the other two strains were one each from Georgia and Texas). Although we have little knowledge of where or when the patients initially acquired their infections, the possibility of detecting recombination is probably higher among strains from the same geographical area, where there could be the opportunity for contact and sexual recombination. Third, the two genes analysed here are highly polymorphic. Highly polymorphic gene sequences permit the construction of robust phylogenies, leading to unambiguous inferences regarding hybridization and recombination.
Our studies showed that combined analyses of the two genes identified more hybridization events than those based on each individual gene. Analyses of additional genes might reveal more hybridization and recombination events. Recent studies using gene genealogical comparison to infer recombination and speciation have typically used four or more genes (for a review see Taylor et al., 1999). Indeed, our earlier study used four genes located at different parts of the genome [LAC, URA5, the internal transcribed spacer region of rRNA (ITS), and the mitochondrial large ribosomal gene (mtLrRNA)]. However, neither the ITS nor the mtLrRNA gene showed phylogenetically informative variation among geographically diverse strains within serotypes A and D (Xu et al., 2000b
). We have also screened portions of other genes using PCR-RFLP, including the nuclear genes GPA1 and ADE2 and the mitochondrial gene NADH dehydrogenase subunit 2. Unfortunately, no phylogenetically informative variation among strains within either serotype A or D was found (unpublished data; Xu, 2002
). Nevertheless, using the two highly polymorphic loci, our analyses clearly identified multiple hybridizations between strains of serotypes A and D and recombination within populations of both serotypes A and D.
While our analyses of URA5 and LAC genes using the 14 serotype AD strains provided unambiguous evidence for recombination within populations of both serotypes A and D, we would like to stress that our data are insufficient for estimating the frequencies of recombination in environmental populations of these two serotypes. To obtain such estimates, larger sample sizes are needed. In addition, increasing the number of genes could also enhance the probability of detecting recombination events. Due to differences in the geographical distribution of these serotypes, it is likely that populations of C. neoformans from different locations will show different degrees of clonality and recombination.
Implications for environmental populations of C. neoformans
Our inferred population structures of serotypes A and D derived from clinical strains of serotype AD are different from direct analyses of strains of serotypes A and D from clinical samples. Based on our analyses, we believe serotype AD strains are more reflective of environmental serotype A and D populations than of clinical serotype A and D populations. For example, one of the consistent observations of clinical samples is that strains of MAT predominate. Indeed, no MATa strain was found among strains of serotypes A (324 strains), B (3 strains) and D (12 strains) isolated from four different geographical areas in the USA (San Francisco, Georgia, Texas and Alabama). All 339 strains had the MAT
allele (Yan et al., 2002
). However, analyses of mating types from strains of serotype AD clearly suggested that both mating types existed in this sample. Among the 14 strains of serotype AD analysed here, nine contained the MATa allele from serotype A (Aa), three contained the MAT
allele from serotype A (A
), and the remaining two contained neither mating type allele from serotype A (Table 1
, Yan et al., 2002
). Similarly, ten strains had the D
allele, three strains had the Da allele and one contained neither mating type allele from serotype D. Though the mating type allele ratios were not 1 : 1 among the 14 strains for either serotype A or D, they were much more balanced than in typically analysed clinical populations of serotypes A or D. Furthermore, because these serotype AD strains arose relatively recently as a result of hybridization between strains of serotypes A and D (Xu et al., 2002
), strains of serotype A with MATa must exist in the environment in North America at present or have existed in the recent past. Strains of serotype A, MATa were until very recently thought to be extinct (Lengeler et al., 2000
). Our results here suggest that these strains not only exist in North America, but that they are recombining with strains of serotype A and MAT
. Coupled with evidence for recent hybridization between strains of serotypes A and D, the recombining populations of both serotypes A and D indicate continuously evolving natural populations of this important human pathogenic fungus.
![]() |
ACKNOWLEDGEMENTS |
---|
![]() |
REFERENCES |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Brandt, M. E., Bragg, S. L. & Pinner, R. W. (1993). Multilocus enzyme typing of Cryptococcus neoformans. J Clin Microbiol 31, 28192823.[Abstract]
Brandt, M. E., Hutwagner, L. C., Kuykendall, R. J. & Pinner, R. W. (1995). Comparison of multilocus enzyme electrophoresis and random amplified polymorphic DNA analysis for molecular subtyping of Cryptococcus neoformans. J Clin Microbiol 33, 18901895.[Abstract]
Brandt, M. E., Hutwagner, L. C., Klug, L. A. & 9 other authors (1996). Molecular subtype distribution of Cryptococcus neoformans in four areas of the United States. J Clin Microbiol 34, 912917.[Abstract]
Casadevall, A. & Perfect, J. R. (1998). Cryptococcus neoformans. Washington, DC: American Society for Microbiology.
Casadevall, A., Freundlich, L. F., Marsh, L. & Scharff, M. D. (1992). Extensive allelic variation in Cryptococcus neoformans. J Clin Microbiol 30, 10801084.[Abstract]
Edman, J. C. & Kwon-Chung, K. J. (1990). Isolation of the URA5 gene from Cryptococcus neoformans var. neoformans and its use as a selective marker for transformation. Mol Cell Biol 10, 45384544.[Medline]
Felsenstein, J. (1981). Evolutionary trees from DNA sequences: a maximum likelihood approach. J Mol Evol 17, 368376.[Medline]
Franzot, S. P., Salkin, I. F. & Casadevall, A. (1999). Cryptococcus neoformans var. grubii: separate varietal status for Cryptococcus neoformans serotype A isolates. J Clin Microbiol 37, 838840.
Kwon-Chung, K. J. (1976). A new species of Filobasidiella, the sexual state of Cryptococcus neoformans B and C serotypes. Mycologia 68, 942946.
Kwon-Chung, K. J., Bennett, J. E. & Rhodes, J. C. (1982). Taxonomic studies on Filobasidiella species and their anamorphs. Antonie van Leeuwenhoek 48, 2538.[Medline]
Kwon-Chung, K. J., Edman, J. C. & Wickes, B. L. (1992). Genetic association of mating types and virulence in Cryptococcus neoformans. Infect Immun 60, 18691874.[Abstract]
Lengeler, K. B., Wang, P., Cox, G. M., Perfect, J. R. & Heitman, J. (2000). Identification of the MATa mating-type locus of Cryptococcus neoformans reveals a serotype A MATa strain thought to have been extinct. Proc Natl Acad Sci U S A 97, 1445514460.
Lengeler, K. B., Cox, G. M. & Heitman, J. (2001). Serotype AD strains of Cryptococcus neoformans are diploid or aneuploid and are heterozygous at the mating-type locus. Infect Immun 69, 115122.
Mitchell, T. G. & Perfect, J. R. (1995). Cryptococcosis in the era of AIDS 100 years after the discovery of Cryptococcus neoformans. Clin Microbiol Rev 8, 515548.[Abstract]
Swofford, D. L. (2002). PAUP*: Phylogenetic Analysis Using Parsimony. Sunderland, MA: Sinauer Associates.
Taylor, J. W., Geiser, D. M., Koufopanou, V. & Burt, A. (1999). The evolutionary biology and population genetics behind fungal strain typing. Clin Microbiol Rev 12, 26146.
Viviani, M. A., Esposto, M. C., Cogloati, M., Montagna, M. T. & Wickes, B. L. (2001). Isolation of a Cryptococcus neoformans serotype A MATa strain from the Italian environment. Med Mycol 39, 383386.[Medline]
Wickes, B. L., Mayorga, M. E., Edman, U. & Edman, J. C. (1996). Dimorphism and haploid fruiting in Cryptococcus neoformans: association with the -mating type. Proc Natl Acad Sci U S A 93, 73277331.
Xu, J. (2002). Mitochondrial DNA polymorphisms in the human pathogenic fungus Cryptococcus neoformans. Curr Genet 41, 4347.[CrossRef][Medline]
Xu, J. & Mitchell, T. G. (2003). Population genetic analyses of medically important fungi. In Pathogenic Fungi in Humans and Animals, 2nd edn, pp. 703722. Edited by D. H. Howard. New York: Marcel Dekker.
Xu, J., Ramos, A. R., Vilgalys, R. & Mitchell, T. G. (2000a). Clonal and spontaneous origins of fluconazole resistance in Candida albicans. J Clin Microbiol 38, 12141220.
Xu, J., Vilgalys, R. & Mitchell, T. G. (2000b). Multiple gene genealogies reveal recent dispersion and hybridization in the human pathogenic fungus Cryptococcus neoformans. Mol Ecol 9, 14711481.[CrossRef][Medline]
Xu, J., Luo, G., Vilgalys, R., Brandt, M. E. & Mitchell, T. G. (2002). Multiple origins of hybrid strains of Cryptococcus neoformans with serotype AD. Microbiology 148, 203212.
Yan, Z., Li, X. & Xu, J. (2002). Geographic distribution of mating type alleles of Cryptococcus neoformans in four areas of the United States. J Clin Microbiol 40, 965972.
Received 11 December 2002;
revised 30 April 2003;
accepted 6 May 2003.