MBE Advance Access originally published online on July 26, 2006
Molecular Biology and Evolution 2006 23(10):1879-1890; doi:10.1093/molbev/msl070
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Research Article |
Recent Evolution of the Human Pathogen Cryptococcus neoformans by Intervarietal Transfer of a 14-Gene Fragment
Department of Molecular Genetics and Microbiology, Duke University Medical Center
E-mail: dietr003{at}mc.duke.edu.
 |
Abstract |
|---|
 TOP Abstract IntroductionMaterials and MethodsResultsDiscussionSupplementary MaterialAcknowledgementsReferences |
|---|
The availability of the whole-genome sequence from the 2 known varieties of the human pathogenic fungus Cryptococcus neoformans provides an opportunity to study the relative contribution of divergence and introgression during the process of speciation in a genetically tractable organism. At the genomic level, these varieties are nearly completely syntenic, share
85–90% nucleotide identity, and are believed to have diverged 18 MYA. Via a comparative genomic approach, we identified a 14-gene region (40 kb) that is nearly identical between the 2 varieties that resulted from a nonreciprocal transfer event from var. grubii to var. neoformans 2 MYA. The majority of clinical and environmental var. neoformans strains from around the world contain this sequence obtained from var. grubii. This introgression event likely occurred via an incomplete intervarietal sexual cycle, creating a hybrid intermediate where mobile elements common to both lineages mediated the exchange. The subsequent duplication in laboratory strains of a fragment of this same genomic region supports evolutionary theories that instabilities in subtelomeric regions promote adaptive evolution through gene amplification and subsequent adaptation. Along with a more ancient predicted transfer event in C. neoformans and a recently reported example from Saccharomyces cerevisiae, these data indicate that DNA exchange between closely related sympatric varieties or species may be a recurrent theme in the evolution of fungal species. It further suggests that although evolutionary divergence is the primary force driving speciation, rare introgression events also play a potentially important role.
Key Words: Cryptococcus neoformans • segmental duplication • parasexual • subtelomeric • introgression • transposons
|
Introduction |
|---|
|
TOPAbstractIntroductionMaterials and MethodsResultsDiscussionSupplementary MaterialAcknowledgementsReferences |
|---|
The availability of relatively simple eukaryotic genomes, represented by the fungi, provides a unique opportunity to employ comparative strategies to characterize the relative contribution of divergence and introgression during the course of speciation. In this research, we examine 2 varieties of the fungal species Cryptococcus neoformans to evaluate these processes in organisms on the cusp of speciation.
Cryptococcus neoformans is a haploid basidiomycete yeast that is the fourth most common life-threatening opportunistic pathogen of immunocompromised patients (Kwon-Chung and Bennett 1992
; Hull and Heitman 2002). Its primary means of reproduction is asexual growth, though examination of markers in wild isolates has demonstrated that outcrossing within varieties does occur in the environment (Kwon-Chung 1976; Chen et al. 1995; Brandt et al. 1996; Boekhout and van Belkum 1997; Litvintseva et al. 2003; Xu and Mitchell 2003; Litvintseva et al. 2005, 2006).
Two varieties are known and have strong phylogenetic support as distinct clades: C. neoformans var. grubii (serotype A) and C. neoformans var. neoformans (serotype D) (Franzot et al. 1998; Xu et al. 2000; Sugita et al. 2001). Both varieties are frequently found in association with pigeon droppings worldwide and have also been isolated from trees, soils, and other sources (Casadevall and Perfect 1998; Hull and Heitman 2002; Idnurm et al. 2005). More than 95% of C. neoformans clinical infections in the United States are caused by var. grubii, although in some regions, var. neoformans infections comprise a significant fraction of the observed cases (Casadevall and Perfect 1998). Up to 12% of infections in New York City may be caused by var. neoformans (Steenbergen and Casadevall 2000). In Europe, clinical var. neoformans strains are more common than in the Americas and may be responsible for up to 30% of infections (Dromer et al. 1996).
Cryptococcus neoformans varieties, neoformans and grubii, are estimated to have diverged 18 MYA (Xu et al. 2000). Despite the phylogenetic analysis indicating distinct clades, var. neoformans and var. grubii have been described as varieties and not as separate species based on their ability to fuse and form viable diploids in the laboratory and in nature. These hybrids arose within the past several million years via an incomplete sexual cycle, producing strains that remain trapped in the diploid state due to genomic differences that prevent the completion of meiosis (Lengeler et al. 2001; Xu et al. 2002; Xu and Mitchell 2003). Molecular analysis of these strains reveals that they are usually aneuploid or diploid, indicating that these strains can undergo parasexual reduction to approach the haploid state (Brandt et al. 1993; Tanaka et al. 1999; Xu et al. 2000). This model has been supported through the creation of artificial hybrids in the laboratory via intervarietal crosses, producing progenies that germinate poorly and are aneuploid or diploid (Lengeler et al. 2001). Interestingly, recent studies have indicated that alleged var. neoformans infections might actually be caused by AD hybrid strains (Boekhout et al. 2001).
The formation of hybrids between closely related sympatric species is not limited to opportunistic fungal pathogens but has been reported in several fungal lineages (Olson and Stenlid 2001; Moon et al. 2004). Hybridization occurs via mating of closely related Candida species and has been proposed as an explanation for the role of genes normally associated with mating being present in asexual species like Candida albicans (Pujol et al. 2004). Interspecies mating can also occur within and between the sensu stricto and sensu lato groups of Saccharomyces species (Marinoni et al. 1999; de Barros Lopes et al. 2002); it has recently been found that some strains of Saccharomyces paradoxus contain a 23-kb region of near identity with Saccharomyces cerevisiae (EJ Louis, personal communication). This region provides evidence of genetic exchange between 2 sensu stricto species, despite their genomes sharing only 85% sequence identity. Another example of potential exchange is in the clinically isolated S. cerevisiae strain YJM789. This strain shares 99% sequence identity with S. cerevisiae strains S288C and RM11-1A but contains a 15-kb region including YAR062W, YAR064W, and YAR066W that is more diverged, with only 96% identity, suggesting that the clinical isolate has acquired this region from a more distantly related uncharacterized variety or species (Tawfik et al. 1989; Goffeau et al. 1996; Gu et al. 2005). Additionally, there is evidence that the S. cerevisiae genome has been the recipient of a number of genes horizontally transferred from bacteria (Gojkovic et al. 2004; Hall et al. 2005).
Here we describe a genome-wide comparison of 2 diverged strains of Cryptococcus neoformans, revealing a 40-kb region containing 14 genes in var. neoformans most likely acquired by direct DNA transfer from var. grubii in a recent nonreciprocal event. This region is distinctive from the rest of the var. neoformans genome because of its near identity to the var. grubii sequence and because of its nonsyntenic genomic location. Following acquisition of this region, var. neoformans appears to have lost its original copy of the region in favor of the sequence obtained from var. grubii. This acquired genomic fragment is nearing fixation in the population. We suggest that nonreciprocal transfer of a multigenic expanse of DNA between strains that are generally genetically isolated may be representative of a general mechanism for DNA exchange among closely related species, potentially providing a selective advantage for the recipient lineage.
|
Materials and Methods |
|---|
|
TOPAbstractIntroductionMaterials and MethodsResultsDiscussionSupplementary MaterialAcknowledgementsReferences |
|---|
Strains
The C. neoformans strains used in this study are listed in table 1. All strains were obtained from the Heitman, Mitchell, or Schell laboratories at Duke University Medical Center. The reference strains used in this study were var. grubii strains KNA-14 (MATa) and H99 (MAT
) (Toffaletti et al. 1993; Nielsen et al. 2003) and the congenic var. neoformans strains JEC20 (MATa) and JEC21 (MAT) (Kwon-Chung et al. 1992; Heitman et al. 1999). Variety and mating-type for selected isolates were verified via polymerase chain reaction (PCR) analysis using primers that amplified the genes STE20 (a/), PAK1, CNA1, and GPA1 as described previously (Lengeler et al. 2001). Based on this analysis, strain DUMC127.92 was reclassified as a var. grubii strain. The molecular VN groups for strains 20020.041, S25C, and UG1712 were determined using PCR fingerprinting as described previously (Meyer et al. 1999).
|
Sequencing
Selected portions of 6 genes were amplified by PCR and sequenced from each of the C. neoformans strains. Information about each gene and the primers used is given in table S1 (Supplementary Material online). The criteria for selecting these genes were that they appear only in single copy in the genome (no paralogs present) and, for the genes outside of the unique genomic feature we identified in this study, that they be scattered across the genome. For the 3 genes in this class, 1 was selected from the left arm of chromosome 5, 1 from the central region of chromosome 1, and 1 from the left arm of chromosome 3. PCR products were purified using the Montage PCRµ96 kit (Millipore, Bedford, MA) and sequenced using standard BigDye chemistry (Applied Biosystems, Foster City, CA).
Southern Blot Analysis
Southern blot analysis used C. neoformans genomic DNA prepared as described by Pitkin et al. (1996). DNA was digested with enzymes and electrophoresed on a 1% agarose gel overnight. The DNA was transferred to a nylon membrane and probed. Probes were produced by PCR from JEC21 or H99 genomic templates using primers given in table S1 (Supplementary Material online), purified using the QIAquick gel extraction kit (QIAGEN, Valencia, CA), and radioactively labeled using the Rediprime II kit (Amersham Biosciences, Little Chalfont Buckinghamshire, England). The probes were detected by exposing the blot to BioMax XAR film (Kodak, Rochester, NY) at –80 °C.
Data Analysis
The MUMmer program was used to perform genome-wide alignments to identify regions of high identity between the JEC21 (var. neoformans) and H99 (var. grubii) genomes (Kurtz et al. 2004). All regions longer than 1,000 bp with nucleotide identity >94% were examined in detail to determine if they represented potential regions of introgression. Several candidate regions were identified but, with the exception of two, these could be explained as long coding regions, repetitive elements, or regions of high gene density (table S2, Supplementary Material online). Additional analyses performed in this study used the Blast, FASTA3, and ClustalX programs (Altschul et al. 1990; Thompson et al. 1997; Pearson 2000; Rice et al. 2000).
Genomic Sequence Data
The genomic sequence data for var. neoformans strains JEC21 and B3501A were obtained from GenBank under accession numbers AE017341–AE017353 and AE017356 (JEC21) and AAEY00000000 (B3501A) (Loftus et al. 2005). Cryptococcus neoformans var. grubii strain H99 sequence data (October 2004 build) and annotation were obtained from the Duke University Center for Applied Genomics and Technology (http://fungal.genome.duke.edu/). Cryptococcus gattii strain R265 sequence was obtained from the Broad Institute Cryptococcus neoformans Serotype B Sequencing Project (http://www.broad.mit.edu). The 8/18/04 assembly represents 6x sequence coverage of the genome that is assembled into 701 contigs in 28 supercontigs (scaffolds). Cryptococcus gattii strain WM276 (Arachne v. 2.0.1 Build3.01 2004-03-01) sequence data were produced at Canadas' Michael Smith Genome Sciences Centre (http://www.bcgsc.ca) and consists of 1,287 contigs. The S. cerevisiae strain RM11-1A sequence was obtained from the Broad Institute (http://www.broad.mit.edu). The RM11-1A Sequencing Project 9/10/2004 release 1 provides 10x sequence coverage of the genome that is assembled into 115 contigs in 17 supercontigs (scaffolds).
Nucleotide Sequence Accession Numbers
All sequence data generated in this study have been deposited in GenBank. Accession numbers are given in table S3 (Supplementary Material online).
Phylogenetic Analysis
The ClustalX program was used to perform multiple alignments of nucleotide sequences (Thompson et al. 1997). Alignments were examined manually, and all columns with questionable homology or gaps were removed. PAUP* was used to build Neighbor-Joining trees using an uncorrected (p) model to determine the distance matrices (Swofford 2002). Bootstrap support values were generated using 2,000 replicates. Trees were rooted with genes from C. gattii strain WM276 as the outgroup. Equivalent topologies were recovered using other tree building methods (data not shown).
|
Results |
|---|
|
TOPAbstractIntroductionMaterials and MethodsResultsDiscussionSupplementary MaterialAcknowledgementsReferences |
|---|
The Genomes of C. neoformans var. neoformans and grubii Are Highly Colinear with
85–90% Nucleotide Sequence IdentityThe genomes of 2 divergent C. neoformans strains have recently been completed, C. neoformans var. neoformans strains, JEC21, and C. neoformans var. grubii strain, H99. Comparison of the
20-Mb genomic sequences of JEC21 and H99 using MUMmer reveals that the 2 genomes are generally colinear and share 85–90% sequence identity at the nucleotide level (fig. 1). Each chromosome displays a region of discontinuity of 40–100 kb that is believed to represent the putative centromere, a highly rearranged, repetitive region of transposon fragments (fig. 1). Although the genomes are largely colinear, there are several apparent inversions and translocations between the current genomic assemblies that may provide a barrier to genetic reassortment between the varieties. Due to these alterations in size, homologous chromosomes have, in some cases, been assigned different chromosome numbers between the H99 and JEC21 assemblies (fig. S1, Supplementary Material online).
|
The Identity Island: A Large Region of High Identity between var. neoformans and grubii
Extensive comparison of the genomes of strains H99 (var. grubii) and JEC21 (var. neoformans) uncovered 2 regions of unexpectedly high sequence identity. The first is
8 kb in length and encompasses the consecutive genes CNM02570, CN02580, CNM02590, and CNM02600. It is in a subtelomeric position in both strains and, in contrast to most of the genome, is nonsyntenic (JEC21, chromosome 13 vs. H99, chromosome 10). It is 95% identical at the nucleotide level, including intronic and intergenic regions, which is 5% above the genome average (table S2, Supplementary Material online).
A second and even more distinctive region involves a subtelomeric site 40-kb long where the 2 genomes are nearly identical by sharing a nucleotide identity of 98.5% (10% higher than the genome average). The current annotation of strain JEC21 predicts 14 protein-coding genes inside this high identity region (fig. 2), all but two of which have support of transcriptional analysis and/or are conserved in other fungal species. Most of the genes within the region are of unknown function with the exception of CNE05250, which appears to be a homolog of enolase, the final enzyme in glycolysis and one of the most abundant proteins in the cell (Entian et al. 1987). Four of the remaining genes have orthologs in S. cerevisiae, none of which are essential in this organism (Winzeler, Shoemaker, et al. 1999; Christie et al. 2004) (table S4, Supplementary Material online). Due to the unusually high level of nucleotide identity between the 2 divergent C. neoformans varieties within this region, we refer to it as the "Identity Island" (fig. 2). All other aligned regions between JEC21 and H99 greater than 1 kb in length with nucleotide identity >94% appeared to have this high nucleotide identity because they contained large coding regions, ribosomal DNA, repetitive elements, or regions of high gene density (table S2, Supplementary Material online).
|
The nearly 10-fold polymorphism reduction between var. neoformans and grubii within the Identity Island is surprising, given that the region consists of
49% noncoding DNA. The average percent identity within introns is 98.6% and within intergenic regions is 97.4%. These genomic regions are not normally subject to strong purifying selection, indicating that strong selective pressure is not responsible for the close similarity between var. neoformans and grubii DNA sequence in this region. In addition to being unusual for its high degree of nucleotide identity, the 40-kb Identity Island is also unusual for its nonsyntenic location in the 2 genomes (fig. 2). Southern analysis was performed to confirm that the Identity Island is in the position that is indicated by the genomic assembly in JEC21. Probes for genes CNE05180 (just outside the Identity Island) and CNE05210 (just inside the Identity Island) demonstrated that these 2 genes are adjacent (data not shown). The altered location, combined with the higher percent identity, suggests that the region has been transferred from one variety to the other since their divergence from a common ancestor.
The Identity Island Was Created by Nonreciprocal Transfer of DNA from var. grubii to var. neoformans
The high sequence identity and nonsyntenic location of the Identity Island suggest that it has been transferred from one C. neoformans lineage to the other via a nonreciprocal event. The direction of transfer was elucidated via comparison with the genome of C. gattii, the closest known relative of C. neoformans. The 2 species are believed to have diverged 40 MYA. Genome sequencing efforts are currently focused on 2 divergent isolates of this organism, providing an outgroup for our analysis.
Comparison with the C. gattii genomes revealed that across the Identity Island boundary this outgroup is colinear with H99 but is discontinuous with JEC21 (fig. 3). Cryptococcus gattii and C. neoformans var. grubii therefore share the genomic arrangement of the common ancestor of these Cryptococcus species, whereas C. neoformans var. neoformans strain JEC21 does not, indicating that the Identity Island was transferred from C. neoformans var. grubii to C. neoformans var. neoformans.
|
Comparison of C. gattii and C. neoformans var. grubii sequences from the Identity Island region indicates that gene order is conserved and that there is no significant difference in the nucleotide identity inside this genomic feature versus outside (80–85%) (fig. 3). This indicates that there is nothing inherently unusual about the Identity Island that restricts its evolutionary rate of divergence, supporting the hypothesis that the Identity Island is the product of a recent nonreciprocal event.
The Identity Island Is Widespread in Natural Populations of C. neoformans var. neoformans
To verify that the Identity Island is an authentic natural feature of the var. neoformans genome and not an abnormality in the sequenced strain or an artifact of genome sequencing, 8 var. grubii and 12 var. neoformans strains collected from around the world were analyzed (table 1). Three genes from inside the Identity Island and three unlinked genes from outside were sequenced and compared with each strain (table 1).
When unlinked genes outside the Identity Island region are compared, it is clear that all the studied var. grubii strains are closely related (1% pairwise polymorphism), all the var. neoformans strains studied are closely related (1% pairwise polymorphism), and that there is a consistent nucleotide divergence between varieties grubii and neoformans (10% pairwise polymorphism) (table 2). These data are in agreement with the observation that the 2 sequenced isolates share 85–90% nucleotide identity across their genomes and confirms that the genome-wide nucleotide identity observed between the sequenced var. neoformans and grubii strains of H99 and JEC21 are generally representative of worldwide populations.
|
In contrast, there is a different pattern when comparing sequence data with genes within the Identity Island. A comparison between var. grubii strains revealed that the level of divergence is equivalent to unlinked genes outside the Identity Island (
1% pairwise polymorphism). However, the var. neoformans strains show a more complex population structure when comparing genes inside the Identity Island. Although most strains are very similar to strain JEC21 (1% pairwise polymorphism), 2 strains are not. Sequence for strains NIH433 (an environmental isolate from Denmark) and NIH430 (a clinical isolate from Denmark) could only be obtained for gene CNE05240. The data demonstrated that the 2 strains differ from the other var. neoformans strains (and by extension the var. grubii strains) by about the same amount as var. neoformans and grubii strains differ outside the Identity Island (table 2). Southern analysis using a CNE05250 gene fragment from JEC21 as a probe demonstrated that NIH433 contained this gene yet showed a different hybridization pattern indicating that it has a different allele of this gene (fig. 5). This suggests that NIH433 and NIH430 bear the original var. neoformans version of the Identity Island.
|
The distinction between genes inside and outside the Identity Island for strains NIH433 and NIH430 is demonstrated by a phylogenetic analysis (fig. 4). For genes outside the Identity Island, var. neoformans and grubii form separate monophyletic clades with strains NIH433 and NIH430 clustering among the var. neoformans strains. However, for the CNE05240 gene inside the Identity Island, var. neoformans and grubii strains group together to form a single clade that excludes NIH433 and NIH430. This pattern is consistent with a recent transfer of the Identity Island from var. grubii to var. neoformans, with NIH433 and NIH430 retaining the original var. neoformans copy. Because NIH433 is one of the parental strains of the congenic pair JEC20 and JEC21, the pair inherited their copy of the Identity Island from the other parent, NIH12 (Heitman et al. 1999
; Fraser et al. 2005). The presence of 2 forms of the Identity Island in var. neoformans provides additional evidence that the Identity Island was transferred into var. neoformans as the recipient population would be expected to contain 2 versions of the region.
|
Based on the H99 genome, the predicted location of the original Identity Island–like sequence in var. neoformans should have been on the right end of chromosome 4. Analysis of the JEC21 genome in this region revealed a
2-kb region at the telomere that appears to be a relic of the original sequence. This region shares 87% sequence identity with var. grubii, reflecting what is observed between the genomes as a whole. This indicates that the original var. neoformans copy of the Identity Island shared the same genomic location as its var. grubii counterpart and that the new location is a result of the transfer event.
Introgression May Have Been Mediated by Repetitive Elements
Analysis of the JEC21 genome revealed that predicted transposons represent 5% of the genome (Loftus et al. 2005). One mobile element associated with the Identity Island is the class 1 non-LTR retrotransposon Cnl1, an element that is found throughout the var. neoformans and grubii genomes in subtelomeric regions (Goodwin and Poulter 2001). In contrast to this normal subtelomeric distribution of Cnl1, 4 partial copies are found grouped together at the interior boundary of the Identity Island in var. neoformans. In addition, a single partial copy is located at the Identity Island boundary in var. grubii (fig. 2). This copy of Cnl1 shares 94% nucleotide identity with 464 bp of the Cnl1 element in JEC21 at the Identity Island boundary. In contrast, the nucleotide identity between the varieties immediately following the Cnl1 sequences jumps to 98.5%. Hence, the Cnl1 elements delineate the Identity Island boundary and may have mediated the translocation of the Identity Island in var. neoformans (fig. S2, Supplementary Material online).
Interestingly, there is also a repetitive element associated with the Identity Island. This putative Cryptococcus specific repetitive element (CSRE1) was introduced into the JEC21 genome during the transfer from var. grubii. CSRE1 has no recognizable homology with any sequences in the GenBank database. In H99, CSRE1 is present at least 22 times, both as full-length copies (4 kb in length) and as defective copies (>500 bp). The best match to this element in JEC21 is 1.4-kb long and is located in the Identity Island. There are 3 JEC21 expressed sequence tags transcribed from the element (CF707481, CF707491, and CF698652); however, all possible large open reading frames from the element are interrupted by stop codons, so it appears to be a nonfunctional variant of the full element (Loftus et al. 2005). There are also several shorter, more highly diverged, copies of this element in the JEC21 genome, suggesting that CSRE1 was present in the common ancestor of C. neoformans var. neoformans and C. neoformans var. grubii, a hypothesis supported by the presence of what appear to be 2 highly diverged fragments, between 300 and 500 bp in length in C. gattii.
Taken together, this evidence suggests that the CSRE1 was present in the common ancestor of C. neoformans but appears to have degenerated in C. gattii and C. neoformans var. neoformans to the point that only a few short remnants remain. When the Identity Island transfer took place, it resulted in the reintroduction of a 1.4-kb copy of CSRE1 into the C. neoformans var. neoformans genome. However, in this case, the transfer has not lead to the proliferation of the element in C. neoformans var. neoformans as the transferred copy is only a fragment of the complete element. Its presence suggests that introgression of DNA between closely related sympatric fungal species could be an important mechanism for spreading repetitive elements.
Genes in the Identity Island Are Possibly Responsible for Selective Retention of This Region in var. neoformans
In the course of sequencing Identity Island genes from multiple strains of var. neoformans and grubii, several strains were found to lack selected genes in the region. Southern analysis demonstrated that some strains appear to have lost some of the genes from the telomere proximal end. To investigate this phenomenon further, we examined a diverse set of var. grubii strains drawn from the 3 molecular groups that are known to exist for this variety (VNI, VNII, and VNB) (Litvintseva et al. 2006). The Identity Island appears in single copy in var. grubii strains, so genes lost in this region cannot be essential for survival. Genes that are retained in all strains are more likely to be essential or provide a selective advantage. This analysis revealed that 5 genes from this subtelomeric location are conserved in all the strains studied (table S5, Supplementary Material online). These conserved genes lack S. cerevisiae homologs: CNE05210, CNE05230, CNE05290, and CNE5300 (all of unknown function) except for the ENO1 ortholog (CNE05250).
Another interesting clue to genes in the Identity Island that may have caused var. neoformans to retain the var. grubii copy comes from a discrepancy between the number of copies of this feature in the 2 sequenced var. neoformans genomes. The JEC21 assembly contains one complete copy of the Identity Island on the right end of chromosome 5 and a segmental duplication of it (14 kb) on the opposite end of the same chromosome (fig. 2). This conflicts with the genome assembly of the closely related strain B3501A that includes only a single complete copy of the Identity Island on the right end of the chromosome. Southern analysis was performed on all strains involved in generation of the JEC20/JEC21 congenic pair (Heitman et al. 1999). The results show that 3 variants of this region exist in var. neoformans, the ancestral copy (NIH433), the acquired copy from var. grubii (most strains), and the segmental duplication (JEC20 and JEC21) (fig. 5). The duplication that has occurred in JEC20 and JEC21 has resulted in the creation of 2 copies of the genes CNE05210, CNE05220, CNE05230, CNE05240, CNE05250 (ENO1), and CNE05260; CNE05210, CNE05230, and CNE05250 (ENO1).
|
Discussion |
|---|
|
TOPAbstractIntroductionMaterials and MethodsResultsDiscussionSupplementary MaterialAcknowledgementsReferences |
|---|
Cryptococcus is typically an environmental fungus that occasionally infects humans to become a pathogen. Disease is thought to initiate in the lung via inhalation of the basidiospore, the product of the a/
sexual cycle and the related cycle of monokaryotic fruiting (Hull and Heitman 2002; Idnurm et al. 2005). Although the sexual cycle is of significant medical interest because it produces the likely infectious basidiospores, studies of natural Crypotcoccal populations indicate that clonal growth is the predominant mode of reproduction. The coexistence of sexual and clonal reproduction may create an avenue for genetic exchange between distantly related strains.
Whereas clonal growth permits propagation and dissemination of aneuploid and rearranged genomes, sexual growth selects against such aberrations because they disrupt proper meiotic segregation. The result of these competing forces could be the creation of hybrid strains where a sexual cycle is initiated, but genomic differences prevent completion of normal meiotic segregation. Hybrids between var. neoformans and grubii are found in nature and can be created in the laboratory, and these serotype AD hybrids are known to be genomically unstable and often aneuploid (Lengeler et al. 2001). These hybrids provide the most likely intermediate through which the genetic exchange that created the Identity Island could have occurred. After the Identity Island was created, the AD hybrid shed its remaining complement of serotype A chromosomes. AD hybrids are thought to sometimes revert to serotype A or D by sequentially losing chromosomes, presumably because this eliminates incompatibilities caused by differences in the serotype's nucleotide sequences (Lengeler et al. 2001). The process of losing only serotype A chromosomes would reduce the AD hybrid to a serotype "D" strain.
The degree of nucleotide identity within the Identity Island supports the AD hybrid transfer hypothesis. The sequence similarity within the Identity Island places its transfer to roughly the same time frame in which AD hybrids are thought to have first evolved (Xu et al. 2002). The fact that the Identity Island transfer and the development of the AD hybrids appear to co-occur lends support to the theory that the transfer occurred through an AD hybrid intermediate.
The uniform level of divergence observed when comparing the var. neoformans and grubii genomes suggests that these 2 varieties have been genetically isolated since their initial divergence. Yet the presence of the Identity Island that appears to have occurred 1/10 as long ago as the varietal divergence based on sequence conservation proves that this is not strictly true. Given that the current estimate for the divergence of var. neoformans and grubii is 18.5 Myr, this would imply that the Identity Island transfer event occurred 2 MYA (Xu et al. 2000). This telomeric acquisition of the sequence is similar to telomeric variation seen in S. cerevisiae (Winzeler, Lee, et al. 1999). A second region, including the consecutive genes CNM02570, CN02580, CNM02590, and CNM02600, may also represent a more ancient exchange event (table S2, Supplementary Material online). This region, including intronic and intergenic sequences, is 95% identical between H99 and JEC21 compared with the genome-wide average of 85–90% identity.
The recombination event that created the Identity Island appears to have been orchestrated by the class 1 non-LTR retrotransposable element Cnl1 (fig. S2). In one possible scenario, the Cnl1 elements provided the homology necessary to allow a reciprocal transfer of DNA between nonhomologous chromosomes in a var. neoformans/grubii hybrid. This model is analogous to the situation in humans where specific Alu repeats have been found to mediate recombination resulting in chromosomal translocations and disease (Hill et al. 2000). Alternatively, interaction between tandemly duplicated Cnl1 elements may have caused a chromosomal break where the resulting fragment then fused to the end of another chromosome through homology of Cnl1 elements at that site. This model is supported by evidence in S. cerevisiae where retrotransposable Ty elements have been identified as fragile sites of elevated chromosomal breakage resulting in chromosomal translocations and loss (Lemoine et al. 2005).
It appears that the Cnl1 elements also played a role in duplicating a portion of the Identity Island in JEC21. A discrepancy exists in the number of copies of the Identity Island between the assembled sequences of var. neoformans (JEC21 vs. B3501A), and the difference in copy number has been verified by Southern blot analysis (fig. 5). Because JEC20 and JEC21 were derived by crosses from NIH12 and NIH433, it appears that the duplication (fig. 2) occurred during this process (Fraser et al. 2005). The appearance of a second copy of the Identity Island is likely the result of homologous recombination of the Cnl1 elements at the 2 ends of the chromosome (fig. S3, Supplementary Material online). The Cnl1 elements just interior of the Identity Island aligned with the Cnl1 elements on the opposite telomere. This caused 14 kb of the Identity Island to be duplicated and appended onto the left end of the chromosome. The duplicated region is inverted relative to the Identity Island sequence on the right end, consistent with this model.
The duplication highlights the general instability of the region occupied by the Identity Island. This region posses features commonly associated with genome instability including a subtelomeric position and repetitive elements. Their coupled effect is to produce translocations and duplications. It has been suggested by other investigators that subtelomeric regions are general genomic locations for adaptive evolution through gene amplification that provides for adaptation (Liti and Louis 2005). Our observations of translocations and duplications within the Identity Island provide support for this hypothesis.
Interestingly, the duplicated region in JEC20 and JEC21 contains the enolase gene (fig. 2). Enolase is one of the most abundant proteins in S. cerevisiae and is the final enzyme in glycolysis. It is encoded by a duplicated gene pair ENO1 and ENO2 that were generated during the S. cerevisiae whole-genome duplication. In JEC20 and JEC21, the duplicated region introduces a duplicated copy of the enolase gene and may have provided a selective advantage under the rich media conditions used during congenic pair generation. Besides enolase, the remainder of the Identity Island genes are of unknown function, and it is unclear what impact their duplication would have. Evidence suggests that some of the same genes in the Identity Island that are duplicated in JEC21 are also highly conserved across var. grubii populations (table S5, Supplementary Material online). If the alleles of these genes perform a function important for survival, as evidenced by their retention in all var. grubii strains, then perhaps they also confer a further advantage by being duplicated though selection on a linked gene or genetic drift could also explain this retention.
Our hypothesis as to the mechanism of creation of the Identity Island can be summarized as follows. First, a hybrid was formed between var. neoformans and var. grubii that was unable to complete meiosis due to genomic incompatibilities. Second, a DNA exchange occurred and the diploid proceeded through a parasexual cycle, losing chromosomes to return to the haploid state. Third, after creation of the Identity Island, the native copy of the Identity Island region was shed by var. neoformans leaving only a 2-kb remnant at the end of the chromosome. Fourth, this strain then spread to near fixation either through the action of selection or through random drift. In a subsequent event, during construction of the congenic pair JEC20 and JEC21, a portion of the Identity Island (14 kb) was duplicated. The duplication appears to have been mediated by Cnl1 retrotransposon elements.
The high degree of sequence similarity among var. grubii genes sequenced in this work supports the recent common ancestry of strains, as previously reported (Boekhout et al. 2001). Similarly, with the exception of data from the Identity Island region, var. neoformans strains are likewise closely related and share common ancestry much more recent than the divergence from var. grubii. Within the Identity Island region, most var. neoformans strains contain the var. grubii sequence that appears to be approaching fixation in the population. Because 2 strains from Denmark were found that do not contain the var. grubii Identity Island copy, it is clear that the transfer occurred since the expansion of var. neoformans. The presence of var. grubii sequence in var. neoformans confirms that genetic exchange between the 2 varieties can occur, albeit rarely.
There are several possible explanations why var. neoformans retained the var. grubii copy of the Identity Island in favor of its own. One is that the var. grubii version confers a selective advantage over the native var. neoformans copy. Southern blot analysis has demonstrated that the var. neoformans native copy of the Identity Island from strain NIH430 has only 5 centromere proximal members of the gene set contained in the var. grubii copy (data not shown). The NIH430 Identity Island contains genes CNE05210, CNE05220, CNE05230, CNE05240, and CNE05250 (ENO1) and is missing CNE05260 through CNE05250 (fig. 2). Acquisition of these distal genes by var. neoformans in the Identity Island transfer from var. grubii may have provided a selective advantage.
For the 5 genes that are shared between the var. grubii Identity Island and the native var. neoformans copy, the standard approach to assessing the action of selection is to use the McDonald–Kreitman Test (McDonald and Kreitman 1991). This method attempts to distinguish the action of selection by comparing the within-species polymorphism with between-species divergence at synonymous and nonsynonymous sites within coding regions. To determine if var. neoformans retained the Identity Island from var. grubii because it offers a selective advantage over the native copy of the Identity Island, it is necessary to have sequence data from var. neoformans stains containing the native copy of the genes. At the present time, only 2 strains (NIH430 and NIH433) have been identified that retain this native copy. We have succeeded in obtaining sequence data from a portion of one of the 5 shared genes, CNE05240 (unknown function), from these strains. These were the data used to perform our phylogenetic analysis (fig. 4). We compared these sequence data to the data from 9 var. grubii strains for this gene. In 486 coding bases, we observed 6 fixed replacement and 10 fixed synonymous between-species differences as well as 4 synonymous within-species polymorphisms (all from var. grubii). This gives a ratio of replacement changes to synonymous changes that are fixed of 0.6 (6/10) and a ratio of replacement to synonymous polymorphisms of 0 (0/4). Fisher's test produces a probability of 0.2, insufficient to make a statistically significant demonstration of selection within this gene.
As an alternative explanation for retention of the Identity Island, it is possible that var. neoformans acquired a deleterious mutation in this region that was then replaced by var. grubii sequence. It is also possible that the Identity Island become widespread through the process of random drift or through linkage to a gene under positive selection. We favor the explanation that the region confers a selective advantage because it seems to have become widespread in the population in a relatively short time. Without an understanding of the function of most of the genes within and flanking the Identity Island, it is difficult to determine with certainty. One possible approach to answering this question would be to perform fitness assays between var. neoformans strains that have the native copy of the Identity Island (NIH433, NIH430) and strains that have the copy acquired from var. grubii to evaluate any relative selective advantages of one copy over the other.
Independent of whether or not the Identity Island confers a selective advantage to var. neoformans, its presence will act to increase the species barrier between the 2 varieties. By introducing a genomic rearrangement between var. neoformans and grubii, it will act to confound intervarietal meiosis and hence reduce the likelihood of a successful hybridization. Thus, although the exchange of sequence by this mechanism may be rare, it may still have a profound effect on the speciation process.
A DNA exchange event has recently been identified involving a transfer of 23 kb from S. cerevisiae to S. paradoxus (EJ Louis, personal communication). This makes it clear that within the fungi even distinct species can exchange genetic material. In both the S. paradoxus and var. neoformans case, the amount of genetic exchange is less than 0.3% of the genome and appears to have occurred as 1 or 2 discrete events. This, combined with the otherwise uniform sequence divergence, suggests that var. neoformans and grubii have been genetically isolated from one another since their initial divergence with the 2 notable exceptions presented in this paper. This makes their relationship to one another analogous to the relationship between S. cerevisiae and S. paradoxus and suggests that by the biological definition of species, var. neoformans and grubii are in fact distinct species.
Given the paucity of closely related sympatric fungal species available for comparison at the genome level, it is unclear how widespread genetic exchange between species or subspecies may be. The phenomenon has been observed twice (var. grubii to var. neoformans; S. cerevisiae to S. paradoxus) with only a few genomes available for analysis; hence, it seems likely that this type of exchange is fairly common. Many of the pathogenic fungi exist in sympatry with divergent varieties (Histoplasma capsulatum, C. neoformans) or closely related sister species (Coccidioides immitis and Coccidioides posadasii, Aspergillus fumigatus and Aspergillus lentulus, and C. gattii and C. neoformans) making it likely that this type of genetic exchange will be observed elsewhere. Although a complete sexual cycle between these varieties or sibling species is expected to be confounded by nucleotide divergence and subsequent inhibition of meiotic recombination, hybridization such as that seen with AD strains of C. neoformans could provide the opportunity for a parasexual cycle to allow introgression of genetic material from one variety to another (Carlile and Watkinson 1994). This argues for continued sequencing of closely related sympatric species to allow evaluation of the significance of introgression in evolutionary process.
Although the amount of DNA exchanged between var. neoformans and grubii appears to be low (0.2%), it may still have a profound impact on the evolutionary development of these varieties. The presences of the Identity Island and of a second, possibly more ancient DNA exchange between var. neoformans and grubii implies that genetic exchange may be a recurrent and important theme in the evolution of this organism. The exchange between S. cerevisiae and S. paradoxus further suggests that this process may be important during speciation of many fungi. Hybridization followed by introgression among organisms on the cusp of speciation provides an avenue for acquiring genetic variation that may provide an immediate selective advantage.
|
Supplementary Material |
|---|
|
TOPAbstractIntroductionMaterials and MethodsResultsDiscussionSupplementary MaterialAcknowledgementsReferences |
|---|
Figures S1–S3 and tables S1–S5 are available at Molecular Biology and Evolution online (http://www.mbe.oxfordjournals.org/).
|
Acknowledgements |
|---|
|
TOPAbstractIntroductionMaterials and MethodsResultsDiscussionSupplementary MaterialAcknowledgementsReferences |
|---|
We are grateful to Kirsten Nielsen, Anastasia Litvintseva, and Wiley Schell for generously providing us with many of the var. neoformans and grubii strains used in this study. Joe Heitman provided access to laboratory equipment and contributed valuable comments and ideas. Jason Stajich also contributed valuable input throughout the course of numerous discussions and Stephanie Diezmann provided assistance with the phylogenetic analysis. We thank Alexander Idnurm and Mark DeLong for critical reading of the manuscript. We also acknowledge the Broad Institute of Harvard and MIT and Canada's Michael Smith Genome Sciences Centre with funding from Genome Canada for genome information prior to publication.
|
Footnotes |
|---|
Laura Katz, Associate Editor
|
References |
|---|
|
TOPAbstractIntroductionMaterials and MethodsResultsDiscussionSupplementary MaterialAcknowledgementsReferences |
|---|
Altschul SF, Gish W, Miller W, Myers EW, Lipman DJ. 1990. Basic local alignment search tool. J Mol Biol 215:403–10.[CrossRef][ISI][Medline]
Boekhout T, Theelen B, Diaz M, Fell JW, Hop WC, Abeln EC, Dromer F, Meyer W. 2001. Hybrid genotypes in the pathogenic yeast Cryptococcus neoformans. Microbiology 147:891–907.
Boekhout T, van Belkum A. 1997. Variability of karyotypes and RAPD types in genetically related strains of Cryptococcus neoformans. Curr Genet 32:203–8.[CrossRef][ISI][Medline]
Brandt ME, Bragg SL, Pinner RW. 1993. Multilocus enzyme typing of Cryptococcus neoformans. J Clin Microbiol 31:2819–23.
Brandt ME, Hutwagner LC, Klug LA, et al. (11 co-authors). 1996. Molecular subtype distribution of Cryptococcus neoformans in four areas of the United States. Cryptococcal Disease Active Surveillance Group. J Clin Microbiol 34:912–7.[Abstract]
Carlile MJ, Watkinson SC. 1994. The Fungi. Boston: Academic Press.
Casadevall A, Perfect JR. 1998. Cryptococcus neoformans. Washington, DC: ASM Press.
Chen F, Currie BP, Chen LC, Spitzer SG, Spitzer ED, Casadevall A. 1995. Genetic relatedness of Cryptococcus neoformans clinical isolates grouped with the repetitive DNA probe CNRE-1. J Clin Microbiol 33:2818–22.[Abstract]
Christie KR, Weng S, Balakrishnan R, et al. (23 co-authors). 2004. Saccharomyces Genome Database (SGD) provides tools to identify and analyze sequences from Saccharomyces cerevisiae and related sequences from other organisms. Nucleic Acids Res 32:D311–4.
de Barros Lopes M, Bellon JR, Shirley NJ, Ganter PF. 2002. Evidence for multiple interspecific hybridization in Saccharomyces sensu stricto species. FEMS Yeast Res 1:323–31.[Medline]
Dromer F, Mathoulin S, Dupont B, Laporte A. 1996. Epidemiology of cryptococcosis in France: a 9-year survey (1985-1993). French Cryptococcosis Study Group. Clin Infect Dis 23:82–90.[ISI][Medline]
Entian KD, Meurer B, Kohler H, Mann KH, Mecke D. 1987. Studies on the regulation of enolases and compartmentation of cytosolic enzymes in Saccharomyces cerevisiae. Biochim Biophys Acta 923:214–21.[Medline]
Franzot SP, Fries BC, Cleare W, Casadevall A. 1998. Genetic relationship between Cryptococcus neoformans var. neoformans strains of serotypes A and D. J Clin Microbiol 36:2200–4.
Fraser JA, Huang JC, Pukkila-Worley R, Alspaugh JA, Mitchell TG, Heitman J. 2005. Chromosomal translocation and segmental duplication in Cryptococcus neoformans. Eukaryot Cell 4:401–6.
Goffeau A, Barrell BG, Bussey H, et al. (16 co-authors). 1996. Life with 6000 genes. Science 274:546, 563–7.
Gojkovic Z, Knecht W, Zameitat E, Warneboldt J, Coutelis JB, Pynyaha Y, Neuveglise C, Moller K, Loffler M, Piskur J. 2004. Horizontal gene transfer promoted evolution of the ability to propagate under anaerobic conditions in yeasts. Mol Genet Genomics 271:387–93.[CrossRef][ISI][Medline]
Goodwin TJ, Poulter RT. 2001. The diversity of retrotransposons in the yeast Cryptococcus neoformans. Yeast 18:865–80.[CrossRef][ISI][Medline]
Gu Z, David L, Petrov D, Jones T, Davis RW, Steinmetz LM. 2005. Elevated evolutionary rates in the laboratory strain of Saccharomyces cerevisiae. Proc Natl Acad Sci USA 102:1092–7.
Hall C, Brachat S, Dietrich FS. 2005. Contribution of horizontal gene transfer to the evolution of Saccharomyces cerevisiae. Eukaryot Cell 4:1102–15.
Heitman J, Allen B, Alspaugh JA, Kwon-Chung KJ. 1999. On the origins of congenic MATalpha and MATa strains of the pathogenic yeast Cryptococcus neoformans. Fungal Genet Biol 28:1–5.[CrossRef][ISI][Medline]
Hill AS, Foot NJ, Chaplin TL, Young BD. 2000. The most frequent constitutional translocation in humans, the t(11;22)(q23;q11) is due to a highly specific alu-mediated recombination. Hum Mol Genet 9:1525–32.
Hull CM, Heitman J. 2002. Genetics of Cryptococcus neoformans. Annu Rev Genet 36:557–615.[CrossRef][ISI][Medline]
Idnurm A, Bahn YS, Nielsen K, Lin X, Fraser JA, Heitman J. 2005. Deciphering the model pathogenic fungus Cryptococcus neoformans. Nat Rev Microbiol 3:753–64.[CrossRef][ISI][Medline]
Kurtz S, Phillippy A, Delcher AL, Smoot M, Shumway M, Antonescu C, Salzberg SL. 2004. Versatile and open software for comparing large genomes. Genome Biol 5:R12.[CrossRef][Medline]
Kwon-Chung KJ. 1976. Morphogenesis of Filobasidiella neoformans, the sexual state of Cryptococcus neoformans. Mycologia 68:821–33.[ISI][Medline]
Kwon-Chung KJ, Bennett JE. 1992. Medical mycology. Philadelphia: Lea & Febiger.
Kwon-Chung KJ, Edman JC, Wickes BL. 1992. Genetic association of mating types and virulence in Cryptococcus neoformans. Infect Immun 60:602–5.
Lemoine FJ, Degtyareva NP, Lobachev K, Petes TD. 2005. Chromosomal translocations in yeast induced by low levels of DNA polymerase a model for chromosome fragile sites. Cell 120:587–98.[CrossRef][ISI][Medline]
Lengeler KB, Cox GM, Heitman J. 2001. Serotype AD strains of Cryptococcus neoformans are diploid or aneuploid and are heterozygous at the mating-type locus. Infect Immun 69:115–22.
Liti G, Louis EJ. 2005. Yeast evolution and comparative genomics. Annu Rev Microbiol 59:135–53.[CrossRef][ISI][Medline]
Litvintseva AP, Marra RE, Nielsen K, Heitman J, Vilgalys R, Mitchell TG. 2003. Evidence of sexual recombination among Cryptococcus neoform