Several mechanisms enforce species boundaries by either preventing the formation of hybrid zygotes, known as pre-zygotic barriers, or preventing the viability and fecundity of hybrids, known as post-zygotic barriers. Despite these barriers, interspecific hybrids form at an appreciable frequency, such as hybrid isolates of the human fungal pathogenic species, Cryptococcus neoformans and Cryptococcus deneoformans, which are regularly isolated from both clinical and environmental sources. C. neoformans x C. deneoformans hybrids are typically highly aneuploid, sterile, and display phenotypes intermediate to those of either parent, although self-fertile isolates and transgressive phenotypes have been observed. One important mechanism known to enforce species boundaries or lead to incipient speciation is the DNA mismatch repair system, which blocks recombination between divergent DNA sequences during meiosis. The aim of this study was to determine if genetically deleting the DNA mismatch repair component Msh2 would relax the species boundary between C. neoformans and C. deneoformans. Progeny derived from C. neoformans x C. deneoformans crosses in which both parental strains lacked Msh2 had higher viability, and unlike previous studies in Saccharomyces, these Cryptococcus hybrid progeny had higher levels of aneuploidy and no observable increase in meiotic recombination at the whole-genome level.
The mixing of species through sexual reproduction can result in hybrid offspring. While hybridization can have beneficial consequences in some cases (e.g. hybrid vigor and the emergence of novel hybrid species), sexual reproduction between diverging lineages or different species is typically deleterious and results in hybrid progeny with reduced fitness or sterility [1–3]. Thus, mechanisms preventing such events, such as pre- and post-zygotic reproductive barriers, tend to be favored by natural selection. Pre-zygotic species barriers block the formation of a hybrid zygote, and in the event that a hybrid zygote forms, several post-zygotic barriers exist that inhibit the viability or fecundity of the hybrid . Post-zygotic barriers include 1) gross chromosomal rearrangements that prevent effective meiotic recombination, 2) Bateson-Dobzhansky-Muller incompatibilities in which interactions between nuclear elements or between mitochondrial and nuclear factors are detrimental or lethal, and finally, 3) the mismatch repair (MMR) pathway, which has an important role in blocking meiotic recombination between diverged DNA sequences.
The MMR pathway was first identified in prokaryotes as a mechanism to repair replication errors or damage-induced mismatches in DNA . The MMR pathway is highly conserved, playing similar roles in unicellular eukaryotes, such as Saccharomyces species, and in multicellular eukaryotes, including humans . Prokaryotic and eukaryotic cells lacking functional MMR components typically have increased mutation rates and therefore display a hypermutator phenotype . The MMR pathway also plays an additional role in maintaining species boundaries by inhibiting homeologous chromosome pairing and subsequent recombination during meiosis. The inability of chromosomes to properly and stably pair, and subsequently undergo recombination, can lead to chromosome nondisjunction during meiosis, resulting in high frequencies of aneuploidy. This aneuploidy can be lethal if a progeny fails to inherit one or more essential chromosomes or if a lethal combination of alleles is inherited. Rayssiguier et al. demonstrated that mutation of the MutL, MutS, or MutH MMR components relaxed the species boundary between Escherichia coli and Salmonella typhimurium , two bacterial species whose genomes are 20% divergent, such that recombination during conjugational and transductional crosses increased up to 1,000-fold . The involvement of MMR in recombination is conserved and has also been shown to enforce the species boundary between Saccharomyces cerevisiae and the closely related species Saccharomyces paradoxus , whose genomes are ~15% divergent . Mutants lacking the eukaryotic MutS homolog, Msh2, or the MutL homolog, Pms1, produce hybrid S. cerevisiae x S. paradoxus progeny with higher rates of viability as well as approximately ten-fold increased frequencies of meiotic recombination [10,11].
Despite the numerous pre- and post-zygotic species barriers and the robust fitness defects associated with hybrids, hybridization occurs at an appreciable frequency. In an exciting recent study, researchers witnessed the origin and monitored the evolution of a novel finch species that arose through a hybridization event in the Galapagos Islands . Hybridization has impacted recent human evolution as well, with several modern human populations having introgressed genomic regions from Neanderthals or Denisovans . Ligers, the hybrid progeny of a male lion and female tiger, and mules, the hybrid progeny of a male donkey and female horse, are examples of hybrids that normally only occur through human intervention. Ligers are bred for their size, as they are larger than either parent (an instance of hybrid vigor), while mules are bred for their endurance, docile demeanor, and intelligence . It is important to note however, that interspecific hybrids are often sterile .
Hybridization is also important in many microbial pathogens and model organisms. For instance, the diploid model organism S. cerevisiae is thought to have arisen following a whole-genome duplication that was a direct consequence of interspecies hybridization [16–18]. An instance of relatively recent hybridization is also thought to have led to the emergence of the novel widespread fungal plant pathogen species Zymoseptoria pseudotritici, originating from fusion between two diverged haploids followed by mitosis and meiosis to generate a recombinant haploid F1 hybrid . In recent years, the hybrid nature of several emerging human opportunistic pathogens has been uncovered [20,21], suggesting hybridization might be a mechanism underlying the emergence of novel pathogens .
Several Cryptococcus species are microbial human fungal pathogens and are responsible for over 200,000 infections in both immunocompromised and immunocompetent individuals annually . Cryptococcal infections are associated with high mortality rates and occur globally. There are currently eight recognized species in the pathogenic Cryptococcus species complex that form two well-supported subgroups, the Cryptococcus neoformans species complex and the Cryptococcus gattii species complex, which consist of two and six species, respectively [24,25]. The present study focuses on the two members of the C. neoformans species complex: C. neoformans and C. deneoformans.
Previously, C. neoformans and C. deneoformans were recognized as a single species with two varieties and two serotypes: C. neoformans var. grubii (serotype A) and C. neoformans var. neoformans (serotype D) . However, there is clear genetic evidence separating these two groups, and molecular phylogenetics along with whole-genome sequencing suggest they diverged ~18 million years ago [26–29]. There are also several phenotypes that differentiate C. neoformans and C. deneoformans as distinct species, such as differences in thermotolerance, capsular agglutination reactions, morphology during murine infection, and human disease manifestations and outcomes [30–34]. Of the pathogenic Cryptococcus species, C. neoformans and C. deneoformans are the two most commonly isolated species from clinical and environmental settings and both species serve as model pathogenic eukaryotic organisms . Both species have bipolar mating-type systems in which a single mating-type (MAT) locus encodes either the MATa or MATα mating-type allele. C. neoformans has only been observed to undergo bisexual reproduction, between cells of opposite mating types, while C. deneoformans is capable of both bisexual and unisexual mating, which occurs either between two cells of the same mating type or via endoreplication . Due to the large prevalence of MATα strains isolated from clinical and environmental settings, C. neoformans and C. deneoformans are thought to largely reproduce through unisexual reproduction or asexually as haploids, with infrequent instances of bisexual reproduction in nature.
Despite differences between C. neoformans and C. deneoformans, these two groups produce hybrids in the laboratory and in nature. C. neoformans x C. deneoformans hybrids, also known as AD hybrids, make up ~7.5% of environmental isolates and up to 30% of clinical isolates in Europe and North America [37–40]. Spores produced by genetic crosses between C. neoformans and C. deneoformans isolates are known to have poor germination frequencies (~5%) relative to intraspecific crosses (~80% germination), and are typically highly aneuploid or heterozygous diploids . This poor viability is likely due to a combination of gross chromosomal rearrangements between the two parental genomes along with ~15% sequence divergence between the parental species [27,29,42], leading to a compromised meiosis that produces genetically imbalanced meiotic progeny. C. neoformans x C. deneoformans hybrids also have unstable karyotypes, can be self-fertile, and display phenotypes intermediate of either parent, although instances of transgressive phenotypes (i.e. phenotypes that fall outside of the range between either parental phenotype) and hybrid vigor have been observed [41,43–45]. Several studies of C. neoformans x C. deneoformans hybrid genomes have utilized restriction fragment length polymorphism analysis or PCR with sequence-specific primers to assess hybrid genomes. Through these methods it has been demonstrated that while C. neoformans x C. deneoformans hybrids are heterozygous at most loci, some chromosomes seem to be recombinant, indicative of potential mitotic or meiotic recombination [44,46–48].
In the present study, we generated C. neoformans and C. deneoformans strains lacking Msh2 to determine if loss of MMR relaxed the boundary between these two species. As expected, C. neoformans and C. deneoformans msh2 Δ mutants displayed hypermutator phenotypes. Hybrid progeny derived from genetic crosses in which both parental strains lacked Msh2 displayed increased germination frequencies compared to wild-type crosses and also exhibited phenotypes and genotypes in accordance with previous findings . Several instances of genetic recombination were observed in hybrid progeny derived from both wild-type C. neoformans x C. deneoformans crosses and msh2Δ mutant crosses, although interestingly, increased frequencies of meiotic recombination were not observed in hybrid progeny derived from crosses involving msh2Δ mutants. Additionally, lower rates of loss of heterozygosity (LOH), higher rates of aneuploidy, and more instances of chromosome breaks and de novo telomere addition were observed in hybrid progeny from msh2Δ mutant crosses. These results suggest that although Msh2 plays a role in the viability of hybrid progeny, other pathways and mechanisms are responsible for blocking homeologous meiotic recombination in Cryptococcus.
To assess the role of the MMR pathway in maintaining species boundaries in Cryptococcus, we first determined if Msh2 plays a similar role in DNA MMR in C. deneoformans as in other fungi. Deletion mutants lacking MSH2 were generated via biolistic transformation and homologous recombination in the C. deneoformans JEC20a genetic background (S1A Fig). After transformation, mutants were selected on medium containing nourseothricin, and PCR was employed to confirm that the deletion allele had replaced the wild-type MSH2 allele at its endogenous locus (S1B Fig).
Following isolation and confirmation of the desired msh2Δ mutants, we determined if the mutants displayed hypermutator phenotypes similar to those observed in msh2 mutants of other fungi [49,50]. To assess mutation rate, fluctuation assays were performed on either YPD medium supplemented with a combination of rapamycin and FK506 at 37°C (where calcineurin, the target of FKBP12-FK506, is essential) or YNB medium supplemented with 5-fluoroorotic acid (5-FOA) at 30°C. Resistance to the combination of rapamycin and FK506 is mediated by mutations in their common target, FKBP12, which is encoded by the gene FRR1, while resistance to 5-FOA arises following loss-of-function mutations in URA5 or URA3, genes encoding enzymes involved in the de novo pyrimidine biosynthesis pathway. In this analysis, the progenitor strain of the genetic deletion mutants, JEC20a, served as the negative control, and a C. neoformans msh2Δ mutant in the KN99α genetic background from the Cryptococcus deletion mutant collection  served as the positive control. Although fluctuation assays do not provide genome-wide mutation rates, the assays allow us to compare the mutation rates at two unique coding loci of the JEC20a msh2Δ mutants to those of the controls. On the rapamycin and FK506 antifungal drug combination, the msh2Δ-1, msh2Δ-2, msh2Δ-3, and msh2Δ-4 mutants exhibited hypermutator phenotypes with significantly higher mutation rates (msh2Δ-1: 1.02 × 10−6 (95% confidence interval (CI) 7.76 × 10−7–1.28 × 10−6); msh2Δ-2: 1.30 × 10−6 (95% CI 1.04 × 10−6–1.59 × 10−6); msh2Δ-3: 1.16 × 10−6 (95% CI 9.01 × 10−7–1.45 × 10−6); and msh2Δ-4: 1.16 × 10−6 (95% CI 9.01 × 10−7–1.45 × 10−6) mutations per cell per generation) than the parental JEC20a strain (8.59 × 10−8 (95% CI 4.84 × 10−8–1.31 × 10−7 ) mutations per cell per generation) (Fig 1A). Mutation rates of three of the four independent msh2Δ mutants (msh2Δ-2, msh2Δ-3, and msh2Δ-4) were also significantly higher than the mutation rate of the parental strain, JEC20a , on 5-FOA (S2A Fig). Interestingly, the msh2Δ-1 mutant failed to produce any 5-FOA resistant isolates during the fluctuation assay, which was explained by a single base deletion in a homopolymeric nucleotide run, causing a frameshift in FUR1 , which encodes a uracil phosphoribosyl transferase involved in the pyrimidine salvage pathway. This mutation led to cross resistance to the antifungal drug 5-fluorouridine (5FU) and the clinically relevant antifungal drug 5-fluorocytosine (5FC) (S2B-S2D Fig) (see Materials and Methods for further details) .
Insertion/deletion mutations (INDELs) in homopolymeric nucleotide runs are a hallmark mutation pattern of msh2 Δ mutants [49,53]. To determine if the four independent JEC20a msh2Δ mutants also displayed similar mutation patterns, FRR1, the gene encoding FKBP12 (the common target of FK506 and rapamycin), was PCR amplified and mutations were identified through Sanger sequencing. This analysis revealed single base pair INDELs in homopolymeric nucleotide runs within FRR1 in all ten independent colonies analyzed for the msh2Δ-1, msh2Δ-2, and msh2Δ-4 strains and nine of ten independent colonies analyzed for the msh2Δ-3 strain (Fig 1B). This is significantly different from the frequency of 1-bp INDEL mutations observed in the wild-type JEC20a parental strain, in which only one of nine colonies had a 1-bp INDEL mutation in a homopolymer run (p<0.001, Fisher’s exact test). As anticipated, the KN99α msh2Δ mutant produced resistant colonies with 1-bp INDEL mutations in homopolymer runs in FRR1 in all (10/10) colonies analyzed (Fig 1B).
Interspecific crosses involving MMR-deficient Saccharomyces strains produce progeny with increased viability [10,11]. To determine if a similar increase in viability would be observed in Cryptococcus, genetic crosses were conducted between C. neoformans and C. deneoformans wild-type strains as well as corresponding msh2Δ mutants. A cross between H99α (a laboratory standard C. neoformans reference strain) and JEC20a served as a control for the msh2Δ hybrid crosses. H99α x JEC20a interspecific crosses produced robust mating structures, including hyphae, basidia, and basidiospores (Fig 2A). Interestingly, bilateral crosses in which both parental strains lacked MSH2 produced similarly abundant hyphae as those produced by the H99α x JEC20a cross, but produced a significantly greater number of bald basidia and therefore fewer basidiospore chains (34%, 33%, and 63% bald basidia on average for the wild-type, unilateral, and bilateral msh2Δ hybrid crosses, respectively, p <0.001, one-way ANOVA, Tukey’s HSD) (Fig 2A and 2B and S1 Table).
Basidiospores from wild-type, unilateral msh2Δ x wild type, and bilateral msh2Δ x msh2Δ C. neoformans x C. deneoformans genetic crosses involving each of the four independent JEC20a msh2Δ mutants and the KN99α msh2 Δ strain were randomly dissected via micromanipulation onto nutrient-rich YPD medium. The spores were allowed to germinate for up to three weeks on YPD medium at room temperature, and total germination frequencies were calculated (Fig 2C and S2 Table). Dissected basidiospores from wild-type H99α x JEC20a crosses germinated at a frequency of 5.8%, which is similar to the previously published frequency of 5% . Basidiospores from unilateral crosses in which only one parent lacked MSH2 germinated at a frequency of 7.5% (Fig 2C and S2 Table). Hybrid progeny from bilateral crosses involving msh2Δ mutants in both parents displayed significantly increased germination frequencies compared to the progeny from unilateral and wild-type crosses, with an average germination frequency of 29% (p<0.001, one-way ANOVA, Tukey’s HSD). Progeny germination frequencies were as high as 41% for some bilateral msh2Δ x msh2 Δ sexual crosses (S2 Table), nearing the expected upper germination frequency limit for C. neoformans x C. deneoformans crosses based on the presence of one large reciprocal chromosomal translocation between C. neoformans H99α and C. deneoformans JEC21α, a strain congenic to JEC20a with the exception of the MAT locus [29,42].
The germination frequencies of progeny from the hybrid C. neoformans x C. deneoformans crosses were much lower than germination frequencies from intraspecific crosses (S3 Fig and S2 Table). For instance, progeny from an intraspecific C. neoformans cross between the wild-type strains H99α and KN99a germinated at an average frequency of 83%. In a unilateral msh2Δ C. neoformans intraspecific cross, progeny germinated at an average frequency of 86%, while progeny from a bilateral msh2Δ C. neoformans intraspecific cross germinated on average only 71% of the time, which was significantly lower than either the unilateral msh2Δ or wild-type C. neoformans intraspecific progeny (p<0.05, one-way ANOVA, Tukey’s HSD). Similarly, progeny from wild-type (JEC20a x JEC21α) and unilateral msh2Δ C. deneoformans intraspecific crosses had high average germination frequencies (78% average). In contrast, progeny from bilateral msh2Δ C. deneoformans intraspecific crosses germinated only 36% of the time, a significantly lower frequency compared to the wild-type and unilateral crosses (p <0.001, one-way ANOVA, Tukey’s HSD) (S3B Fig and S2 Table). The decreases in germination frequencies in progeny from bilateral msh2Δ intraspecific crosses is likely due to the high mutation rates associated with loss of Msh2, as has been observed in other studies, although this effect appears to be more pronounced in C. deneoformans than in C. neoformans [10,54].
During the prolonged incubation period in which dissected progeny were allowed to germinate, YPD agar germination plates were kept sealed with parafilm on the benchtop. Surprisingly, after 14 days of incubation, aerial filaments began to emerge from the periphery of several germinated hybrid progeny. Through microscopic analysis we found that these progeny had produced structures resembling those observed during sexual reproduction under mating-inducing conditions. Continued incubation and additional microscopic analysis revealed that after approximately three weeks of incubation, a number of these progeny produced basidia and sparse basidiospores under conditions previously not known to support sexual reproduction for any Cryptococcus species or strains (rich medium, incubated in the light in sealed plates) (Fig 3). This phenotype was measured across all hybrid progeny, parental strains, and standard laboratory reference strains. Although there was no significant difference between the ability to produce hyphae on YPD between the progeny from the different types of hybrid crosses (one-way ANOVA), progeny from bilateral msh2Δ hybrid crosses tended to be able to produce hyphae more often on average than progeny from wild-type or unilateral msh2 Δ hybrid crosses (45% compared to 25% or 19%, respectively) (Fig 3B and S3 Table). Interestingly, no parental strains, progeny from intraspecific crosses, or common laboratory reference strains were able to produce hyphae on YPD except for XL280α, a hyper-filamentous C. deneoformans strain .
The hybrid progeny from both wild-type C. neoformans x C. deneoformans crosses, as well as those from unilateral and bilateral msh2Δ crosses, exhibited typical Cryptococcus hybrid genotypes: high levels of aneuploidy and inheritance of the MATa and MAT α mating-type alleles from both parents [41,45]. With sequence-specific primers for the gene STE20, which exists as two mating-type specific alleles and encodes a kinase involved in the pheromone response signaling cascade, the mating types of the hybrid progeny were determined. Nearly all progeny inherited and maintained both the STE20α allele from the C. neoformans parent and the STE20a allele from the C. deneoformans parent, which was expected due to the diploid genome characteristic of C. neoformans x C. deneoformans hybrids (S4 Fig) [41,45]. Based on flow cytometry, the ploidy of the majority of the hybrid progeny was diploid or aneuploid (S5 Fig). Three of the 27 hybrid progeny for which whole-genome sequencing (WGS) was obtained (progeny YX1, YX3, and YX6) were, however, estimated to be close to haploid, and all three were from a wild-type C. neoformans x C. deneoformans cross (S5 Fig). These results are in stark contrast to the FACS analysis results for 38 intraspecific progeny from the C. neoformans and C. deneoformans wild-type, unilateral msh2Δ, and bilateral msh2Δ crosses, all of which were estimated to be haploid, with one exception: KN99α msh2Δ x KN99a progeny 5 appeared diploid (S6 Fig).
Although the majority of naturally occurring C. neoformans x C. deneoformans hybrid strains are not self-fertile, C. neoformans x C. deneoformans hybrids produced under laboratory conditions, similar to those presented in this study, are often self-fertile [41,43]. This fertility is characterized by the ability to produce hyphae when incubated alone on mating-inducing media, such as MS agar plates, at room temperature in the dark. All but one of the hybrid progeny assessed produced hyphae on MS medium, and many progeny were also capable of producing basidia and basidiospores (S7 Fig). The only progeny that was not self-fertile (YX3) had lost the C. deneoformans MATa locus (S4 and S7 Figs). Interestingly, progeny YX1, which only inherited the C. deneoformans MATa locus but not the C. neoformans MAT α locus, was self-fertile (S4 and S7 Figs). While the C. deneoformans JEC20a parent is not self-fertile, other C. deneoformans MATa strains are self-fertile , and genetic mechanisms underlying this fertility may be similar to those observed in YX1.
Following isolation and characterization of hybrid progeny from the wild-type and unilateral and bilateral msh2 Δ mutant crosses, we generated WGS data for 27 hybrid progeny. The chromosomal composition of the hybrid progeny, along with the identification of potential sites of meiotic recombination, LOH events, and heterozygosity across the genome, were assessed by employing the analytic pipeline described in S8 Fig (see methods for details). To determine ploidy, read depth was assessed in conjunction with flow cytometry data (Figs 4, and S5 and S9).
Hybrid progeny from crosses between wild-type C. neoformans and C. deneoformans parental strains, as well as those from unilateral and bilateral msh2 Δ hybrid crosses, displayed high rates of aneuploidy (Figs 4, S9 and S10 and S4 Table). We characterized the number of instances in which each chromosome was aneuploid by determining which chromosomes had been gained or lost relative to the euploidy estimated by FACS analysis (i.e. variations from 1n or 2n). Aneuploidies involving all chromosomes were observed, with the exception of two homologous chromosome pairs: C. neoformans Chr5/C. deneoformans Chr4 and C. neoformans Chr 6/C. deneoformans Chr5 (S10A and S10B Fig and S4 Table). All hybrid progeny had three or fewer aneuploid chromosomes, with the exception of a progeny from a wild-type cross, YX6, which was disomic for five chromosomes. We also observed a trend in which larger chromosomes were less likely to be aneuploid, although there was not a significant correlation between chromosome length and likelihood of aneuploidy (S10C Fig and S4 Table).
The majority (16/27) of the hybrid progeny were diploid (2n, 12/27 progeny) or nearly diploid (2n+1 or 2n-1, 4/27 progeny), and all progeny derived from unilateral and bilateral msh2 Δ crosses were close to diploid (S4 Table). Interestingly, hybrid progeny from bilateral msh2 Δ mutant crosses were nearly completely heterozygous across a majority of the genome and displayed fewer LOH events (Figs 4, S9 and S11 and S5 Table). Quantifying this heterozygosity, hybrid progeny derived from bilateral msh2Δ x msh2Δ mutant crosses had significantly more heterozygosity (p<0.05, Kruskal-Wallis test, Dunn’s test), with a genome-wide average of 96% heterozygosity relative to hybrid progeny derived from wild-type H99α x JEC20a crosses, which had an average of 67% heterozygosity across their genomes (S11 Fig and S5 Table).
In contrast to the results from the hybrid progeny, progeny derived from intraspecific wild-type, unilateral msh2Δ, and bilateral msh2Δ C. neoformans and C. deneoformans crosses were largely haploid based on the combination of FACS data and WGS (S6 and S12 Figs). Out of the 38 intraspecific progeny for which WGS was obtained, only three progeny from C. neoformans intraspecific crosses were not haploid: progeny 5 from a unilateral KN99α msh2Δ x KN99a cross, which was estimated to be near diploid (2n-1; missing one copy of Chr13), and progeny 3 and 4 from a bilateral KN99α msh2Δ x KN99a msh2Δ cross (both 1n +1; gained one copy of Chr2 and Chr7, respectively) (S12 Fig). All progeny from all C. deneoformans intraspecific crosses were haploid by both FACS analysis and WGS (S6 and S12 Figs).
Whole-genome sequencing revealed instances of potential meiotic recombination in the hybrid progeny derived from the wild-type, unilateral msh2Δ, and bilateral msh2Δ crosses. Due to the high levels of heterozygosity across the genomes of many C. neoformans x C. deneoformans hybrid progeny, and because all progeny were recovered through random spore dissection, we were not able to detect heteroduplex DNA, an indicator of recombination in MMR-deficient strains. Additionally, it is possible that some recombination events might not be detected by merely assessing the depth of reads aligning to each parental reference strain. For instance, if a hybrid progeny inherited both homeologous chromosomes involved in a meiotic reciprocal recombination event, read depth coverage would be equal across both parental chromosomes. To ensure that additional recombination events like these were not being missed, Illumina paired-end reads were aligned to a combined reference that included both parental genomes to identify additional possible recombination sites. This analysis was based on the assumption that if a recombination event occurred, the forward and reverse reads of a pair would align to different chromosomes. Different filtering thresholds were applied based on mapping quality and the number of read-pairs supporting the event. The results of these analyses, using more or less stringent filtering thresholds to detect recombination events, show similar frequencies of recombination events for each of the different types of hybrid crosses (Fig 5 and S6 Table). With the strictest filtering thresholds (at least 15 read-pairs supporting the event and a mapping quality of 60), a median of 1 recombination event was detected across the whole genomes of progeny derived from wild-type, unilateral KN99α msh2Δ x JEC20a, and bilateral msh2Δ crosses, while a median of 1.5 events were detected in progeny from unilateral H99α x JEC20a msh2Δ-1 crosses. Furthermore, with the least strict thresholds (at least 5 read-pairs supporting and mapping quality greater than or equal to 0), a median of 10 recombination events were detected in progeny from a wild-type cross, a median of 8.5 and 6.5 events were detected in progeny from both unilateral crosses (H99α x JEC20a msh2Δ-1 and KN99α msh2Δ x JEC20a , respectively), and a median of 5.5 events were detected in progeny from bilateral crosses (Fig 5 and S6 Table). These results showed no increase in meiotic recombination in hybrid progeny derived from parents lacking Msh2.
High-confidence recombination sites were mapped onto the H99 reference genome along with the distribution of SNPs differing between H99α and JEC20a , as well as the distribution of repetitive elements (Fig 6A). One might expect that recombination events in hybrid progeny would occur at regions with higher homology between the two parental genomes (i.e. regions with lower SNP densities). However, analysis of the SNP densities within 1 kb on either side of each recombination site showed that high-confidence recombination events did not occur in regions with significantly lower or higher SNP densities (Fig 6B). Moreover, no high-confidence recombination events were detected within the ~40-kb region that shares 98.5% identity between the two parents, known as the identity island . Additionally, no tracts ≥ 300 nucleotides in length with complete homology between the two parental genomes were identified, including within the identity island, suggesting that recombination events were not being missed due to limitations of the sequencing methods used. It also did not appear that recombination events were more likely to occur at repetitive elements (Fig 6A). Interestingly, one of the recombination events was near the mating-type locus, a known hotspot for recombination in Cryptococcus .
Another unexpected finding was a phenomenon associated with hybrid progeny derived from unilateral or bilateral msh2Δ crosses in which chromosome breaks occurred at centromeres, transposable elements, and other regions in the genome and appeared to have been repaired by de novo addition of telomeric repeats (Figs 4C, S9 and S13 and S7 Table). Of the 15 instances of de novo telomere addition in the hybrid progeny, 5 occurred at repetitive elements, such as centromeres and transposons. We also assessed whether or not this phenomenon occurred in any of the progeny derived from intraspecific crosses, with the hypothesis that this type of chromosomal breakage and de novo telomere addition would only be viable in the context of a diploid, such that no essential genes are lost if only one homolog breaks. Accordingly, in the intraspecific progeny, we only identified one instance of de novo telomere addition at the end of chromosome 13 in progeny 3 from a bilateral KN99α msh2Δ x KN99a msh2 Δ cross (S14 Fig and S7 Table). This event occurred ~17 kb from the end of the chromosome and only resulted in the loss of three genes encoding hypothetical proteins (CNAG_07919, CNAG_06256, and CNAG_07920) as well as a gene encoding a transporter belonging to the major facilitator superfamily (MFS) (CNAG_06259) (S14 Fig), indicating these are not essential genes.
In this study, we investigated the role of MMR in maintaining the species boundary between two closely related and prevalent opportunistic human fungal pathogens, C. neoformans and C. deneoformans . The MMR pathway is known to play a highly conserved role in blocking recombination between homeologous DNA sequences during meiosis, serving as a post-zygotic barrier in addition to its role in DNA repair. Findings from previous studies in prokaryotic and eukaryotic models indicated that inactivating the MMR pathway by genetically deleting pathway components, particularly Msh2, allows increased homeologous recombination during meiosis [8,10,11]. Therefore, we hypothesized that lack of Msh2 in Cryptococcus would allow increased pairing and recombination between homeologous chromosomes during meiosis, leading to decreased rates of chromosome nondisjunction, and ultimately the production of more viable progeny with fewer instances of aneuploidy.
Fluctuation analysis and characterization of the mutational spectra of C. deneoformans msh2Δ mutants confirmed that Msh2 plays a similar role in DNA MMR in C. deneoformans as has been previously observed across eukaryotes and in other pathogenic Cryptococcus species [49,50]. Similar to results from previous studies in prokaryotic and eukaryotic microorganisms, germination frequencies of C. neoformans x C. deneoformans hybrid progeny derived from bilateral msh2 Δ crosses increased significantly relative to wild-type hybrid germination frequencies [8,10,11]. This was in contrast to the reduced viability observed in bilateral msh2Δ intraspecific crosses, which decreased by 12% compared to wild-type C. neoformans intraspecific crosses and 42% compared to wild-type C. deneoformans intraspecific crosses; this reduced viability is likely due to the hypermutator phenotype of msh2Δ mutants, as has been observed in previous studies in S. cerevisiae [10,54].
Hybrid progeny derived from genetic crosses between C. neoformans and C. deneoformans isolates are known to display unique phenotypes. For example, hybrids, especially those generated under laboratory conditions, can exhibit self-fertility, producing hyphae and in some cases basidia and spores under mating-inducing conditions [41,45]. The majority of the hybrid progeny derived from the crosses described here displayed phenotypes in accord with previously published findings for C. neoformans x C. deneoformans hybrids [41,43]. However, a novel transgressive phenotype was exhibited by many hybrid progeny: the ability to produce hyphae and in some cases, basidia and sparse basidiospores, on nutrient-rich YPD medium in sealed plates in the light. This was surprising, because environmental cues known to suppress mating in Cryptococcus include nutrient-rich environments, such as YPD medium, as well as light and high levels of humidity and carbon dioxide [33,36,58–60]. The formation of sexual reproduction structures by the hybrid progeny under nutrient-rich, light, sealed conditions represents sexual reproduction of Cryptococcus in a novel environment, which has never been shown to induce filamentation for any C. neoformans or C. deneoformans strain previously. The isolation of self-fertile progeny provides an important example of how hybridization can enable Cryptococcus to access a dimorphic state under conditions that are not conducive for the parental isolates to sexually reproduce. The ability to produce spores, the infectious propagules of Cryptococcus human pathogens, in a previously prohibitive environment also provides the opportunity to produce more infectious propagules, which could potentially result in a higher rate of infection. This is in line with previous studies which found that up to 30% of clinical isolates in Europe were C. neoformans x C. deneoformans hybrids [38–40]. These results, along with recent studies that identified Msh2-deficient C. neoformans clinical isolates , suggest that certain Cryptococcus hybrids could potentially contribute to the emergence of a new pathogen, similar to the recent emergence of a wheat stem rust pathogen lineage that is the result of a hybridization with no subsequent recombination or chromosomal reassortment .
Unlike their haploid parents, C. neoformans x C. deneoformans hybrid progeny typically have relatively unstable aneuploid or diploid karyotypes [43–45]. All hybrid progeny derived from C. neoformans x C. deneoformans crosses in this study were aneuploid based on both FACS analysis and whole-genome sequencing data. Three hybrid progeny isolated from wild-type C. neoformans x C. deneoformans crosses were close to haploid and all other hybrid progeny derived were diploid or nearly diploid. Additionally, all hybrid progeny were euploid or nearly euploid, with three or fewer aneuploid chromosomes, with only one exception. These findings are similar to those from Parry and Cox, who found that S. cerevisiae progeny dissected from a triploid were close to euploid, suggesting only a limited number of aneuploid chromosomes are tolerated . We also observed instances where the read-depth of several aneuploid chromosomes was lower than expected, which likely represent events of chromosome loss among a fraction of the cells that were sequenced at the whole-genome level, reflecting the karyotypic instability of these progeny.
The higher ploidy levels in progeny derived from crosses involving msh2 Δ mutants were unexpected because MMR mutants in yeast produce progeny with lower levels of aneuploidy . It is possible that the diploid or near diploid C. neoformans x C. deneoformans progeny were the meiotic products of a tetraploid, generated by the fusion of two diploids that formed within the mating patch. However, interspecific progeny derived from divergent Saccharomyces tetraploids display much higher germination frequencies (>90%), which is very different from the frequencies associated with these C. neoformans x C. deneoformans hybrid progeny, making it unlikely that they are derived from tetraploids . It is also possible that the high frequency of diploid progeny could be indicative of a failed meiosis. However, Cryptococcus mutants lacking known meiotic genes, such as those encoding the meiosis-specific endonuclease Spo11 or the key meiotic regulator Dmc1, are either unable to produce viable progeny or unable to efficiently produce basidiospores, respectively [64,65]. In contrast to the abnormal sexual reproduction structures produced by these meiosis-deficient mutants, the basidia produced by hybrid crosses in this study generated four lengthy spore chains. On the other hand, it is possible that without Msh2 during the hybrid meiosis, initiation of recombination between homeologous sequences will not be prevented effectively, and meiosis will therefore proceed to a degree while homeologous chromosomes are linked by strand invasions, increasing the number of viable progeny. This could also lead to increased chromosomal nondisjunction in meiosis I or potentially a skipping of the reductional meiotic division if homeologous chromosomes remain linked by strand invasions, which could explain the near diploid genomes of hybrid progeny from the unilateral and bilateral msh2Δ crosses. The strand invasions between homeologous chromosomes might be resolved as non-crossovers or crossovers, although the crossovers may be inviable, which could also explain the increased number of basidia lacking spore chains in the bilateral msh2Δ hybrid matings. The higher ploidy in hybrid progeny from bilateral msh2Δ x msh2Δ crosses was also associated with significantly more heterozygosity across their genomes than their counterparts from wild-type C. neoformans x C. deneoformans crosses. It is possible that this higher level of heterozygosity contributes to the increased viability by masking deleterious alleles or overcoming Bateson-Dobzhansky-Muller genetic incompatibilities, as has been observed in previous studies of C. neoformans x C. deneoformans hybrids [48,66].
An interesting phenomenon associated with progeny derived from unilateral and bilateral hybrid msh2Δ crosses as well as a single intraspecific progeny from a bilateral msh2Δ cross was the de novo addition of telomeric repeat sequences to various locations in the genome including centromeric repeat sequences, non-centromeric transposable elements, and the rDNA repeat locus following chromosomal breaks. Although the addition of telomeric repeats at non-telomeric sites can promote the stabilization of a broken chromosome, it also often leads to the loss of a large portion of a chromosome, which would normally threaten cell viability. However, the presence of both homeologous chromosomes from each parental species in these hybrid progeny alleviates this problem. In the haploid intraspecific progeny, only one instance of de novo telomere addition was observed, and in this case only a small portion of the chromosome (~17kb) was lost, thus avoiding the deleterious effects that might be associated with the larger losses observed in the hybrid progeny if they were to happen in a haploid background. Loss of Msh2 has been shown to promote telomeric recombination in yeast, and our results suggest Msh2 might be mediating a similar anti-recombination mechanism at telomeric and other repetitive loci in Cryptococcus . Further supporting this, de novo telomeric repeat addition has been observed in Cryptococcus following CRISPR-mediated double-stranded breaks at centromeres . Overall, our results suggest that in Cryptococcus crosses involving msh2Δ mutants, chromosomes may be more prone to double-stranded breaks, or that the normally occurring double-stranded breaks are unable to be properly repaired, and loss of Msh2 promotes de novo telomere addition at these sites.
Many instances of LOH and recombination were observed in the hybrid progeny assessed in this study. One caveat to note regarding the recombination events identified through WGS, is that C. neoformans x C. deneoformans hybrids have been known to experience loss of heterozygosity during mitotic growth , and results from other previous studies also suggest that mitotic recombination can occur during mating itself [47,69]. Unexpectedly, instances of recombination were detected in progeny derived from each of the different types of hybrid crosses (wild-type, unilateral, and bilateral msh2Δ), but no increase in meiotic recombination was observed in hybrid progeny derived from parents lacking Msh2. Although Cryptococcus has lower frequencies of meiotic recombination compared to S. cerevisiae – approximately 1.27 crossovers per chromosome per progeny derived from intraspecific bisexual crosses , and this frequency is estimated to decrease by six- to seven-fold in C. neoformans x C. deneoformans hybrid progeny  – we expected to see significantly higher rates of recombination in hybrid progeny from msh2 Δ crosses compared to wild-type crosses based on previous studies in both prokaryotes and eukaryotes [8,10,11]. Based on read-depth analysis as well as detecting recombination events by aligning whole-genome sequencing from the progeny to a combined reference genome with both parental species, no increase in recombination was observed in crosses involving either a single parent lacking Msh2 or both parents lacking Msh2. One potential explanation for this observation could be that the level of sequence divergence between C. deneoformans and C. neoformans is large enough that meiotic recombination will be inefficient, even in the absence of MMR. However, previous studies on meiotic recombination in S. cerevisiae indicate that although recombination is less efficient at high levels of sequence divergence, loss of MMR leads to approximately a 24-fold increase in meiotic recombination between 15% divergent sequences, the same level of divergence as between C. neoformans and C. deneoformans [27,29,71]. Furthermore, studies on mitotic recombination in S. cerevisiae also similarly indicate that even at 26% sequence divergence, loss of MMR leads to a 55-fold increase in mitotic recombination [72,73]. Conversely, studies in other models have observed instances where loss of Msh2 did not lead to increased recombination frequencies between substrates with as little as 1% sequence divergence, up to 25% divergence [74–76].
In summary, this study illustrates several key findings on the roles of MMR in Cryptococcus. In C. deneoformans, msh2Δ mutants behave as hypermutators on various selective media and acquire INDELs in homopolymeric nucleotide runs. Hybrid C. neoformans x C. deneoformans progeny dissected from genetic crosses involving msh2Δ mutants generally displayed increased germination frequencies compared to those from wild-type crosses. Hybrid progeny derived from msh2Δ crosses displayed phenotypes and karyotypes characteristic of C. neoformans x C. deneoformans hybrid strains, such as diploidy/aneuploidy and self-fertility, and some progeny displayed a novel, transgressive phenotype in which they were capable of producing hyphae, basidia, and basidiospores on a glucose- and nutrient-rich medium in the light. The increased viability of hybrid progeny derived from bilateral msh2Δ crosses suggests that loss of Msh2 in Cryptococcus may allow homeologous chromosomes to pair more efficiently during meiosis. However, the observation that loss of Msh2 did not seem to increase the frequency of meiotic recombination between C. neoformans and C. deneoformans homeologous chromosomes was highly unexpected based on many previous studies, particularly those in yeast. These results suggest that Msh2 plays a role in maintaining the species boundary between C. neoformans and C. deneoformans, albeit an unexpected one. Additionally, these results suggest alternative pathways or additional MMR components may play different or more important roles in maintaining species boundaries in Cryptococcus than in other previously studied organisms; proteins like the DNA helicases Mph1 and Sgs1, which have been shown to block homeologous recombination and play significant roles in chromosome nondisjunction in budding yeast, would be ideal candidates for further investigation [77–79]. Thus, future studies identifying the robust post-zygotic mechanisms that ultimately maintain integrity by blocking homeologous recombination between these two closely related species will be of great interest.
The C. neoformans and C. deneoformans strains described in this study are listed in S8 Table. Strains were stored at -80°C in liquid yeast extract peptone dextrose (YPD) supplemented with 15% glycerol. Strains were inoculated on YPD agar plates, initially grown at 30°C for 3 days, and then maintained at 4°C. Due to the hypermutator phenotypes associated with msh2Δ strains and the genomic instability associated with C. neoformans x C. deneoformans hybrids, strains used in the experiments of this study were not maintained for more than two weeks at 4°C on YPD agar plates and fresh cells from frozen glycerol stocks were inoculated to new YPD agar plates at the end of each two-week period.
The open-reading frame of the MSH2 gene (gene ID: CNA07480) in the C. deneoformans JEC20a  genetic background was replaced with the gene encoding the dominant drug-resistance marker for nourseothricin resistance, NAT , by homologous recombination via biolistic transformation as previously described . Following transformation and selection on YPD + 100 μg/mL nourseothricin agar medium, genomic DNA was isolated from candidate mutants with the MasterPure DNA purification kit (Epicentre) and PCR followed by gel electrophoresis confirmed correct integration of the NAT dominant resistance marker. The locations of the primers used to generate the deletion allele and confirm deletion mutants are depicted in S1A Fig and their sequences are given in S9 Table. To generate congenic strains of opposite mating types for the intraspecific C. neoformans and C. deneoformans msh2Δ crosses, KN99a  was crossed with the KN99α msh2Δ mutant and JEC20a msh2Δ-1 was crossed with JEC21α , respectively. Progeny were isolated and PCR was used to confirm that they inherited the msh2Δ deletion construct at the endogenous MSH2 locus as well as the appropriate mating type, and did not inherit a functional MSH2 allele.
Fluctuation analysis was utilized to quantify the mutation rates, or the number of mutations per cell per generation, of the JEC20a msh2Δ mutants. The wild-type strain JEC20a served as a negative control and the msh2 Δ mutant from the 2015 Madhani deletion collection in the KN99α genetic background served as a positive control . For each strain, including controls, ten 5 mL YPD liquid cultures were each inoculated with a single colony from a YPD agar stock plate. Cultures were incubated overnight at 30°C. After incubation, cultures were pelleted at 3,000 x g, washed twice with 5 mL dH2O, and resuspended in 4 mL dH2 O. 100 μL of undiluted, washed cells was plated directly to YPD + 100 ng/mL rapamycin + 1 μg/mL FK506 or yeast nitrogen base (YNB) + 1 mg/mL 5-fluoroorotic acid (5-FOA) solid agar medium. Washed cells were diluted 1:100,000 and 100 μL of the dilution was plated to YPD solid agar medium. Inoculated YPD and YNB+5-FOA plates were incubated at 30°C for 4 or 14 days, respectively. Inoculated YPD+rapamycin+FK506 plates were incubated at 37°C for 14 days. Following incubation, colonies on each of the media were counted and mutation frequencies were calculated with the FluCalc program, which utilizes the Ma-Sandri-Sarkar maximum-likelihood estimation (MSS-MLE) equation for calculations .
Single resistant colonies from fluctuation analyses were streak purified to YPD + rapamycin + FK506 medium and grown for 3 days at 37°C. Genomic DNA was isolated using the MasterPure Yeast DNA Purification Kit (Epicenter Biotechnologies, Madison, WI), and the FRR1 gene was PCR-amplified with Phusion High-Fidelity DNA Polymerase (NEB, Ipswich MA, USA). PCR products were subjected to gel electrophoresis, extracted using a QIAgen gel extraction kit, and mutations were identified through classical Sanger sequencing at Genewiz. Fisher’s exact probability test was used to calculate statistically significant differences between the frequencies of 1-bp INDEL mutations compared to other types of mutations in YPD + rapamycin + FK506-resistant colonies from strains lacking MSH2 compared to the wild-type JEC20a strain using the VassarStats online software (http://vassarstats.net).
Papillation assays were conducted on YNB solid agar medium supplemented with 1 mg/mL 5-FOA, 100 μg/mL 5-fluorocytosine (5FC), or 100 μg/mL 5-fluorouridine (5FU). For this assay, ten independent YPD liquid cultures were inoculated with ten single colonies from YPD agar plates and incubated overnight at standard laboratory conditions. Following overnight culture, cells were pelleted at 3,000 x g and resuspended in 2 mL dH2O. Sterile cotton swabs were then used to inoculate quadrants of the agar plates with each independent overnight culture. The inoculated agar plates were incubated at 30°C for 6 days to allow sufficient growth to visualize resistant colonies.
All genetic crosses were conducted on Murashige and Skoog (MS) agar medium following Basic Protocol 1 for mating assays as previously described . The frequency of bald basidia was calculated by imaging random areas surrounding two mating patches per cross, counting the number of bald basidia and basidia producing basidiospores, and the frequencies from the two mating patches were averaged. For each genetic cross, two independent mating patches were assessed; for each mating patch, over 130 total basidia were assessed across at least 11 images. Images used to quantify bald basidia were taken on a Zeiss Axio Scope.A1 with camera. Random basidiospore dissection was performed as described in Basic Protocol 2 . Following dissection, the micromanipulated basidiospores were germinated for up to 3 weeks on YPD agar plates sealed with parafilm and incubated on the laboratory benchtop. Images of self-fertile hybrid progeny derived from bilateral msh2Δ crosses producing sexual structures on YPD agar plates following dissection were taken with an Accu-Scope EXC-500 microscope with an attached Nikon DXM1200F microscope camera.
Primers designed to specifically amplify only the C. neoformans STE20a, C. neoformans STE20α, C. deneoformans STE20a, or C. deneoformans STE20 α alleles aided in identifying the mating types of the hybrid progeny (primers listed in S9 Table). To assess self-filamentation, hybrid progeny were spotted onto MS agar plates and incubated for 14 days at room temperature (approximately 24°C) in the dark as described in Basic Protocol 1 . Filamentation was assessed via microscopy after 14 days of incubation with an Accu-Scope EXC-500 microscope with an attached Nikon DXM1200F microscope camera.
Cells were patched onto YPD agar medium and incubated at 30°C overnight; strains that exhibited slow growth were incubated at 30°C for three days. Cells were harvested by scraping a 2 mm sized colony with a toothpick and resuspending in 1 mL PBS buffer. Cells were washed once with 1 mL PBS and then fixed in 1 mL 70% ethanol overnight at 4°C. After fixation, cells were washed once with 1 mL NS buffer (10 mM Tris-HCl pH = 7.6, 150 mM sucrose, 1 mM EDTA pH = 8.0, 1 mM MgCl2, 0.1 mM ZnCl2, 0.4 mM phenylmethylsulfonyl fluoride, 1 mM β-mercaptoethanol) and resuspended in 180 μl NS buffer supplemented with 5 μl propidium iodide (0.5 mg/mL) and 20 μl RNase A (10 mg/mL). Cells were incubated covered, overnight at 4°C with shaking. Prior to analysis, 50 μl of cells were diluted in 2 mL of 50 mM Tris-HCl (pH = 8.0) in a 5 mL falcon tube. Flow cytometry was performed on 10,000 cells and analyzed on the FL1 channel on a Becton-Dickinson FACScan.
Single colonies of strains for whole-genome sequencing were inoculated into 50 mL of liquid YPD and grown overnight at 30°C in standard laboratory conditions. Overnight cultures were then pelleted at 3,000 x g in a tabletop centrifuge and subsequently lyophilized. High molecular weight DNA was extracted from lyophilized cells following the CTAB protocol as previously described . Genomic DNA libraries were constructed with a Kapa HyperPlus library kit for 300 bp inserts and sequenced at the Duke Sequencing and Genomic Technologies Shared Research core facility. Libraries were sequenced using paired-end, 2 x 150 bp reads on an Illumina HiSeq 4000 platform. The BioProject accession numbers for each sample are provided in S10 Table.
C. deneoformans strain JEC21α is the congenic mating partner of JEC20a, which was obtained in an earlier study by selecting MATα progeny after ten rounds of backcrossing to JEC20a . The genomes of C. deneoformans JEC21α and JEC20a strains are therefore nearly identical, with the exception of 5,322 SNPs mainly distributed over three genomic regions (including the mating-type locus) . Because a highly contiguous de novo genome assembly of JEC20a is not currently available, the genome assembly of strain JEC21α (GCA_000091045.1) was used for all comparisons. The nuclear genomes of the two parental strains, C. neoformans H99α and C. deneoformans JEC21α, were compared by performing whole-genome alignments with Satsuma (https://github.com/bioinfologics/satsuma2) , using default parameters. The output of Satsuma was input to the visualization tools “BlockDisplaySatsuma” and “ChromosomePaint”, included in the same package to generate a postscript file. For representation purposes, chromosome color codes were modified in Adobe Illustrator, and centromeres and other genomic features (rDNA and MAT loci) were superimposed at scale based on their respective genomic coordinates. Characterization of the chromosome composition of the hybrid progeny followed the procedure summarized in S8A Fig. The combined nuclear reference genome used in this study was built with the genome assemblies of the two parental strains after reordering and reorienting the JEC21α contigs to maximize collinearity with the H99α assembly (S8B Fig), using the Mauve Contig Mover tool . A dot plot analysis comparing the H99α assembly with the JEC21α rearranged assembly (S8B Fig) was performed with D-Genies application , which uses minimap2  for aligning the two genomes. Raw Illumina paired-end reads of the selected progeny and the combined reference genomes were input into the sppIDer pipeline , which is a wrapper that sequentially maps the Illumina short reads to the combined reference, performs quality filtering (MQ > 3), and generates depth of coverage plots (Figs 4 and S9). For each progeny, the number of chromosomes and the ploidy were estimated from the sppIDer plots in conjunction with the flow-cytometry data. Chromosomal aberrations, e.g. due to recombination, or chromosome breakage followed by de novo telomere addition, were inferred from the sppIDer plots, and further validated by visual inspection of the mapped reads in IGV .
To inspect the heterozygosity levels of the hybrid progeny, the same set of Illumina paired-end reads were mapped to the C. neoformans H99α reference genome, using the BWA-MEM short-read aligner (v0.7.17-r1188) with default settings . SNP discovery, variant evaluation, and further refinements were carried out with the Genome Analysis Toolkit (GATK) best-practices pipeline [94,95] (v22.214.171.124), including the use of Picard tools to convert SAM to sorted BAM files, fix read groups (module: ‘AddOrReplaceReadGroups’; SORT_ORDER = coordinate), and mark duplicates. Variant sites were identified with HaplotypeCaller from GATK and only high-confidence variants that passed filtration were retained (the “VariantFiltration” module used the following criteria: DP < 20 || QD < 2.0 || FS > 60.0 || MQ < 40.0 || SOR > 4.0). The genome-wide level of heterozygosity was defined as the ratio of the number of heterozygous SNPs divided by the total number of SNPs (i.e. heterozygous and non-reference homozygous SNPs), and was calculated from the resulting VCF files on a per-individual basis after extracting the corresponding sites using the module VariantsToTable from GATK. Due to the high variance in heterogeneity across the genomes of progeny from the different genetic crosses and unequal numbers of progeny analyzed from each different genetic cross, the non-parametric Kruskal-Wallis test followed by Dunn’s test were performed using JMP v15 (SAS Institute).
To predict variants in the JEC20a msh2Δ-1 mutant, we employed the variant calling procedure described above, using the JEC21 genome (GCA_000091045.1) as reference. The effect of a variant (SNPs and INDELs) was predicted and annotated using SnpEff . Only variants of moderate and high impact were considered (i.e. excluding synonymous and non-coding variants), and variants in common between the mutant and the JEC20a parent were flagged as background mutations and excluded. Given the inability of the msh2Δ-1 mutant to produce 5-FOA resistant colonies, we specifically focused on mutations in genes involved in the de novo and salvage pathways of pyrimidine biosynthesis. A single base-pair deletion was identified in the gene FUR1: deletion of a thymine at position 570229 on Chr5 (AE017345.1), GeneID: CNE02100 (AE017345.1:569,042–570,826). Mutations in components of the pyrimidine de novo biosynthesis pathway, such as URA3, have been demonstrated to be synthetically lethal with FUR1 mutations in yeast [97,98]. Independent overnight cultures of each of the independent JEC20a msh2Δ mutants, a KN99α fur1Δ mutant, the JEC20a progenitor strain (negative control), and a KN99α msh2Δ mutant (positive control) were swabbed on YNB plates supplemented with 5-FOA to illustrate that both JEC20a msh2Δ-1 and KN99α fur1 Δ mutants are incapable of producing 5-FOA resistant colonies (S2B Fig). Furthermore, fur1 Δ mutants are also known to be resistant to the antifungal drug 5-fluorouridine (5FU) and the clinically relevant antifungal drug 5-fluorocytosine (5FC) . Papillation assays with the same strains used on 5-FOA were also conducted on YNB plates supplemented with 5FC or 5FU. As anticipated, the JEC20a msh2Δ-1 and the KN99α fur1Δ mutant displayed congruent growth phenotypes and were resistant to both 5FC and 5FU, unlike the parental JEC20a, the KN99α msh2Δ strain, or any of the other JEC20a msh2 Δ mutants (S2C and S2D Fig).
To detect possible recombination events in C. neoformans x C. deneoformans hybrids, the respective reference genomes were retrieved from NCBI (accession: ASM301198v1 for H99 and ASM9104v1 for JEC21 as no genome assembly for JEC20 was available). As the H99 strain was not sequenced in this project, the respective Illumina paired-end reads were retrieved from SRA (SRR7042283). All Illumina paired-end libraries were filtered with the default parameters of Trimmomatic v0.36 . Filtered reads of each strain (including the parental strains) were aligned on the combined reference (containing both H99 and JEC21 genome assemblies) with HaploTypo pipeline , using BWA-MEM v0.7015 , samtools v1.9 (Li 2011) and GATK v126.96.36.199 .
To detect recombination events, we assumed that at recombinant sites it would be possible to detect read-pairs with the two reads aligning to different chromosomes. Therefore, for each strain (including the parental strains), we obtained the coordinates of all the pairs with each read aligning to a chromosome of a different parental species. Only reads with at least 75 bp and a maximum of 3 mismatches were considered. As there were some differences between the parental JEC20 and the reference JEC21, variant calling was performed with HaploTypo pipeline , using GATK v188.8.131.52 . Bedtools v2.26  intersect was used to determine how many SNPs overlapped each read, and in cases in which more than 3 SNPs would overlap, the filter of mismatches was adjusted to that value. The final results were filtered based on mapping quality (MQ) and number of pairs supporting a given recombination event (Bedtools v2.26  merge with a distance of 100bp). Four different analyses were performed using more or less stringent filters: i) MQ = 60 and at least 15 reads supporting the event (stricter); ii) MQ = 60 and at least five reads supporting the event; iii) MQ > = 0 and at least 15 reads supporting the event; iv) MQ > = 0 and at least five reads supporting the event (less strict). The script that automates this pipeline is available on GitHub (https://github.com/Gabaldonlab/detect_recombination/). As it was possible to detect recombination in the parental strains (false positives), Bedtools v2.26  subtract was used to remove all the regions detected in hybrid progeny that were also detected in the parental strains.
Paired-end reads of the progeny resulting from intraspecific crosses of C. neoformans H99α × KN99a and C. deneoformans JEC21α × JEC20a, or from intraspecific crosses involving their derived msh2Δ mutants, were mapped to the C. neoformans H99α and C. deneoformans JEC21α reference genomes, respectively, using the procedures described above. Gross aneuploidy of chromosomes was inferred from read counts collected in 1-kb non-overlapping intervals across the genome using the module “count_dna” from the Alfred package (v0.1.7) (https://github.com/tobiasrausch/alfred) and subjected to median normalization and log2 transformation. The resulting data was converted to binary tiled data (.tdf) using “igvtools toTDF” and plotted as a heatmap in IGV viewer. Chromosome breaks and de novo telomere addition were inferred by abrupt changes in read coverage within a chromosome and visually confirmed by read mapping in IGV.
Recombination breakpoints in hybrid progeny could potentially occur at regions with higher sequence identity between the two parental genomes (i.e. regions with lower SNP densities). If no SNPs could be scored between the two species in genomic tracts smaller than 300 bp (corresponding to the read size) this would potentially lead to an underestimation of the recombination events. Therefore, to determine the distribution of SNPs differing between C. neoformans and C. deneoformans, 150-bp paired-reads of C. deneoformans JEC20a were mapped to a C. neoformans H99α reference genome using the methods described above. SNP density was calculated from the resulting VCF file in 300-bp bins using VCFtools with the option “—SNPdensity 300”, parsed into a BED file format, and visualized as a density heatmap in IGV. Regions without mapped reads were identified by extracting intervals with no coverage using the output of “bedtools genomcov -bga”. No regions of 300 bp were found without any SNP, including the ~40-kb nearly identical region between the two parental genomes (~98.5% sequence identity) that was introgressed from C. neoformans to C. deneoformans .
Next, to examine whether high-confidence recombination events occur in regions with significantly lower or higher SNP densities, we compared the SNP densities within 1-kb on each side of each of the inferred recombination breakpoints, with other 1kb-binned genomic regions (excluding centromeres and the mating-type locus region), and plotted their density and distribution. Three sets of high-confidence recombination events were chosen for this analysis: (a) recombination events supported by mapping quality of 60 and by both 5 and 15 read-pairs (MQ60-Cov15 and MQ60-Cov5), but that could not be inferred directly from the read depth plots; (b) events supported by MQ60-Cov15 and the read depth plots, but not called when using MQ60-Cov5 as a filtering criteria; and (c) events supported by both MQ60-Cov15 and MQ60-Cov5 which were also directly inferred from the read depth plots.
Finally, to inspect if any of these recombination breakpoints were associated with genomic regions enriched in repeats or transposable elements, RepeatMasker (RepeatMasker Open-4.0 2013–2015; http://www.repeatmasker.org) was run with a library of previously characterized C. neoformans transposable elements  and de novo-identified repeat consensus sequences generated by RepeatModeler2 (https://github.com/Dfam-consortium/RepeatModeler). “Bedtools intersect” was employed to determine if any of the genomic regions associated with repeats or transposable elements overlapped with any of the recombination breakpoints.
Hybrid progeny were considered aneuploid if they were missing or had an extra chromosome compared to the theoretical expected number given their ploidy level as measured by FACS analyses, and irrespective of whether they inherited copies from each parent or from only one of the parents. For example, progeny YX4 was scored as euploid because it was determined to be 2n by FACS and has two copies of each chromosome, even though the two copies of chromosomes 4 and 11 were both inherited from C. neoformans. Another example, YX6, which by FACS is close to 1n , was scored as aneuploid for chromosomes 9, 11, 12, 13 and 14 (see S4 Table). The results of these analyses were plotted as the number of hybrid progeny with aneuploidies for each chromosome and genetic cross, and the total aneuploidy observed for all strains was plotted by chromosome length, which showed no correlation between chromosome size and aneuploidy.
We thank Tom Petes, Sue Jinks-Robertson, and Paul Magwene for critical reading and comments on the manuscript. We thank Kevin Zhu for critical reading, comments on the manuscript, and generating figures. We thank Jay Jawahar for his assistance in the initial phases of constructing the JEC20a msh2Δ mutants. We also thank the Madhani Laboratory for the KN99α msh2Δ and fur1Δ deletion strains.
8 Jun 2020
Dear Dr Priest,
Thank you very much for submitting your Research Article entitled 'Factors enforcing the species boundary between the human pathogens Cryptococcus neoformans and Cryptococcus deneoformans' to PLOS Genetics. Your manuscript was fully evaluated at the editorial level and by three independent peer reviewers with expertise in the field. The reviewers appreciated the attention to an important problem, but raised some substantial concerns about the current manuscript. Based on the reviews, we will not be able to accept this version of the manuscript, but we would be willing to review a revised version. We cannot, of course, promise publication at that time.
Should you decide to revise the manuscript for further consideration here, your revisions should address the specific points made by each reviewer. We will also require a detailed list of your responses to the review comments and a description of the changes you have made in the manuscript.
If you decide to revise the manuscript for further consideration at PLOS Genetics, please aim to resubmit within the next 60 days, unless it will take extra time to address the concerns of the reviewers, in which case we would appreciate an expected resubmission date by email to firstname.lastname@example.org.
If present, accompanying reviewer attachments are included with this email; please notify the journal office if any appear to be missing. They will also be available for download from the link below. You can use this link to log into the system when you are ready to submit a revised version, having first consulted our Submission Checklist.
To enhance the reproducibility of your results, we recommend that you deposit your laboratory protocols in protocols.io, where a protocol can be assigned its own identifier (DOI) such that it can be cited independently in the future. For instructions see our guidelines.
Please be aware that our data availability policy requires that all numerical data underlying graphs or summary statistics are included with the submission, and you will need to provide this upon resubmission if not already present. In addition, we do not permit the inclusion of phrases such as "data not shown" or "unpublished results" in manuscripts. All points should be backed up by data provided with the submission.
While revising your submission, please upload your figure files to the Preflight Analysis and Conversion Engine (PACE) digital diagnostic tool. PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email us at email@example.com.
PLOS has incorporated Similarity Check, powered by iThenticate, into its journal-wide submission system in order to screen submitted content for originality before publication. Each PLOS journal undertakes screening on a proportion of submitted articles. You will be contacted if needed following the screening process.
To resubmit, use the link below and 'Revise Submission' in the 'Submissions Needing Revision' folder.
We are sorry that we cannot be more positive about your manuscript at this stage. Please do not hesitate to contact us if you have any concerns or questions.
Gregory P. Copenhaver
Reviewer's Responses to Questions
Comments to the Authors:
Please note here if the review is uploaded as an attachment.
Reviewer #1: In this study the authors performed crosses between two species of Cryptococcus (C. neoformans and C. deneoformans, ~15% divergent), and then analyzed progeny that were derived from wild-type or mismatch repair defective (msh2 null, unlateral or bilateral) parents. They found that progeny derived from msh2null bilateral crosses showed higher viability and levels of aneuploidy. Recombination events were seen at low frequency in the progeny (1 to 1.5 events per cross), with similar frequencies in wild-type and msh2null crosses. Based on these observations, the authors suggest that the mismatch repair system is not playing a role in limiting meiotic recombination between divergent Cryptococcus species.
In general, the work is well done and clearly explained and shows a role for the mismatch repair factor Msh2 in enforcing species boundaries in Cryptococcus. I have two major comments indicated below.
1. The authors do a nice job describing the previous work in bacteria (Hfr crosses between E. coli and Salmonella typhimurium) and yeast (meiotic recombination involving interspecific hybrid of S. cerevisiae and S. paradoxus) indicating that the mismatch repair system acts to maintain species barriers. However, in addition to these studies, it’s also worth thinking about the following:
Elliott and Jasin (MCB 21:2671) measured I-SceI induced recombination between divergent substrates in mammalian cells. They observed that substrates with 1.5% divergence displayed 10-fold higher levels of recombination in Msh2-/- compared to wild-type cells. In addition, the Msh2-/- recombinants displayed a much higher level of uncorrected tracts, indicating that many recombination events involved the formation of heteroduplex DNA.
Anand et al. (Nature 544:377) examined break-induced replication in yeast using templates that contained different densities of DNA polymorphisms. A key observation of their work was that “When recombination occurs without a protruding nonhomologous 3' tail, the mismatch repair protein Msh2 does not discourage homeologous recombination. However, when the DSB end contains a 3' protruding nonhomologous tail, Msh2 promotes the rejection of mismatched substrates.” In addition, they showed that “Nearly all mismatch correction depends on the proofreading activity of DNA polymerase-δ, although the repair proteins Msh2, Mlh1 and Exo1 influence the extent of correction.”
Peterson et al. (Mol Cell doi: 10.1016/j.molcel.2020.04.009) examined meiotic recombination in F1 hybrid mice. They analyzed events at two recombination hotspots loci that in the hybrid displayed a high (~1%) polymorphism density. Interestingly, they observed no difference in the frequency of crossover events in wild-type and Msh2-/-, but still observed high levels of heteroduplex DNA in Msh2-/-.
I summarized the above studies because they involved systems where a DSB was induced/occurred at a specific site/region and polymorphisms were present that allowed them to monitor the presence of heteroduplex DNA. In addition, the studies illustrate the different outcomes that can be obtained when mismatch repair is ablated. In some cases, mismatch repair had no effect on the overall frequency, but did affect the repair of heteroduplex intermediates. With this said I have the following concerns with the work presented by Priest et al.:
A. Crosses between the two species yielded very low numbers of events (1 to 1.5 per cross), and as the authors stated on line 479, it is “difficult to determine if these recombination events occurred during meiosis or subsequent mitotic growth during the culturing required for whole genome sequencing.”
B. Their analysis did not map the exact breakpoints or determine whether there was a significant polymorphism density in the regions undergoing recombination, or if there was any evidence of uncorrected repair tracts. It was intriguing that the authors observed a fair number of de novo telomere addition events in msh2 bilateral crosses, perhaps indicative of an increase in breakage under these conditions. Because it’s not clear where the breakpoints occurred, it’s hard to determine if the breaks occurred in repetitive regions where one might not expect to see effects of mismatch repair. These issues make it difficult for me conclude that the authors can provide mechanistic evidence for factors enforcing the species boundary between different Cryptococcus species.
2. The authors report low rates of loss of heterozygosity and high rates of aneuploidy in hybrid progeny from msh2null mutant crosses. This is an interesting observation but, at least for this reviewer, not a lot of mechanistic insights were provided. It would be interesting to perform crosses within the same species (C. neoformans x C. neoformans, C. deneoformans x C. deneoformans), analyzing progeny derived from both wild-type and msh2null isolates. This would provide information on whether some of the phenotypes that were observed were the result of sequence divergence or other activities promoted by Msh2.
1. Line 70: Define what you mean by “transgressive. A violation of accepted dogma?
2. Lines 109-121. The authors provide many details about how many hybrids exist in nature, but they don’t address how often they are fertile. This paragraph is nicely written, but feels like a distraction from the topic they are addressing (factors that enforce species boundaries).
3. Line 197. It would be worth explaining here what the targets are in the rapamycin +FK506 (FRR1 is a common target) and 5-FOA (URA3 or URA5) assays that lead to resistance to these drugs. This would give the reader an immediate context in terms of the kinds of mutations that can be identified.
4. The detailed analysis of the msh2null-1 mutant feels like a distraction. It’s somewhat interesting that the authors could track down why msh2null-1 failed to produce 5-FOA resistant isolates, but the extended analysis of the mutant does not provide much to the overall story. I see no reason why it can’t be presented in the Materials and Methods or supplement, rather than being emphasized in the Results and Discussion.
5. Line 253 and elsewhere. It would be valuable in the text to provide some of the statistical analysis presented in the Figures that indicated that the differences seen between wild-type and msh2null were significant.
6. Line 491. Typo-do you mean 15% divergent sequences?
Reviewer #2: Priest et al. aim to identify the role of the gene MSH2 in hybrid sterility in the pathogenic fungi Cryptococcus. Previous work in Saccharomyces has shown that anti-recombination due to sequence divergence is responsible for sterility between Saccharomyces species. Knocking out MSH2 can rescue part of this sterility by increasing the number of meiotic events between divergent sequence and thus promoting proper chromosome segregation. Here, the researchers show that MSH2 does function in DNA MMR in Cryptococcus similar to other organisms, but does not increase successful recombination in hybrids. This is at odds with their results that germination frequency is greatly increased, and it is unclear what mechanism is leading to higher viability. They also demonstrate other interesting results, like high rates of diploidy in the viable hybrids, and potentially an ability for hybrids to self fertilize. They thus show support that MSH2 has diverged in function in Cryptococcus, surprising as its role in anti-recombination is conserved from bacteria to eukaryotes.
I think the topic of hybridization in fungal pathogens is really interesting, and a better understanding of the mechanisms and phenotypes associated with hybrid viability is needed. I particularly enjoyed the Discussion section where this was addressed. I have several larger questions and comments, below, and some clarifying questions I think would lead to a more clear manuscript.
Throughout the manuscript there is a lot of reference to phenotypes “unique to hybrid progeny,” but it’s unclear if these are unique to hybrid progeny because non-hybrid crosses are not analyzed (or at least not discussed). For example, is the ability to undergo sexual reproduction after 3 weeks under abnormal conditions unique to msh2 hybrids? What do wild-type hybrid and non-hybrid cross controls look like? What are the difference between hybrids and non-hybrids under the same conditions after this time period? Statistics are needed to identify if differences are significant. Similarly, the observation of repairs of double strand breaks with telomeric repeat sequences (Lines 341-345) - is this unique to hybrid progeny with msh2 or does this happen in msh2 knockouts in non-hybrid progeny as well?
I’m not sure I agree with the statement that loss of Msh2 did not relax species boundaries in Cryptococcus. It seems from their results of germination frequency, that MSH2 does indeed play a role in maintaining species boundaries (when msh2 is deleted in both species, germination increases), in line with results from other organisms. The statement that alternative pathways limit recombination is also unclear, as only MSH2 was tested in this capacity and the MMR pathway still appears to be functioning in many aspects of chromosome segregation/meiosis.
Could you quantify the aneuploidy you see? It would be helpful to see proportions of crosses with aneuploidy across non-hybrids, non-hybrid msh2, msh2 knockouts in wild type hybrids, unilateral and bilateral msh2 hybrids. Is the aneuploidy random or are there particular chromosomes that are lost or gained from one species?
Is recombination required for successful completion of meiosis in Cryptococcus? What is the expected meiotic recombination frequency? It’s particularly interesting that the few recombination events identified are in the wild-type hybrid crosses instead of the msh2 knockouts. What might this suggest?
Other comments and questions
What was the intention behind creating a series of msh2 knockouts and why do the independent msh2 knockouts have variable mutation rate and other phenotypes? Is this likely because of mutations at other loci due to hypermutator phenotype?
Can you describe what you mean by self-fertility? To a naïve Cryptococcus person, I don’t know what the typical reproductive cycle is. Can a (non-hybrid) diploid not self-fertilize normally? Can they persist asexually?
On a related note, I wasn’t clear what the expected ploidy of the hybrid progeny was until the discussion.
Are the hybrids isolated from environmental/clinical samples typically diploid?
What is wild type germination frequency between strains? (non hybrid cross) What about in msh2 knockouts in non-hybrids?
Figure 2A and text: Do bald basidia = sterility? Hard to tell because the PDF version of the figures is pretty blurry, but I can’t distinguish differences between panels. Could you point out the relevant mating structures (or lack there of) that I should be looking for? Is the relevant point that even though there are more bald basidia that there is higher germination?
Did you look at the germination frequency of the F1 hybrid progeny that produced sexual structures after incubation on YPD for 3 weeks?
Some other literature to include:
Bozdag et al. bioRXiv https://doi.org/10.1101/755165
Rogers et al. 2018 https://doi.org/10.1371/journal.pbio.2005066
Li et al. 2019 https://doi.org/10.1038/s41467-019-12927-7
Reviewer #3: The manuscript by Priest et al. explores the role of MSH2 and mismatch repair (MMR) in the post-zygotic reproduction barrier in hybrids between two Cryptococcus species. Hybrids between C. neoformans and C. deneoformans are readily found in the environment and in higher frequency in clinical cases. Like the Saccharomyces yeasts, there appears to be little or no pre-zygotic reproductive isolation, however these hybrids are generally sterile with relatively low germination frequencies of meiotic products, a post-zygotic barrier. The manuscript tests the role of MMR in this low fertility by knocking out MSH2 which has been shown to improve meiotic spore viability in Saccharomyces hybrids by increasing crossing over and improving segregation of chromosomes. They find that knocking out MSH2 in both species parent of hybrids increases fertility significantly, however this does not appear to be via increased crossing over during meiosis. The authors conclude that other factors must be involved in repressing meiotic recombination in these Cryptococcus hybrids.
In general, the experiments are well done and the conclusions reasonably sound. I think the authors are underselling the role of MSH2 which clearly is involved in some way in the reproductive isolation barrier and some further discussion of this is merited. The presumed role of MMR in reproductive isolation in hybrids, based on experiments in bacteria (E. coli and S. typhimurium) and Saccharomyces (mostly S. cerevisiae and S. paradoxus with a similar level of diversity as seen here in the Cryptococcus species), is to be saturated for mismatches during strand invasion at the beginning of the recombination process, resulting in abortion of the initial recombination intermediates. This has the consequence of reducing meiotic recombination and increasing meiosis I non-disjunction due to the lack of tension on the spindle. It’s clear from the Saccharomyces studies that different MMR components may have independent roles in this barrier – pms1 and msh2 mutants in CHR III homeologous XO (Chambers et al. 1996) – despite their shared role in MMR in vegetative cells. MSH2 has also been shown to have a mismatch independent role in telomere repeat recombination which may or may not be relevant here but indicates additional roles independent of sequence divergence. As the role in Saccharomyces and bacteria is binding mismatches in strand invasion structures and perhaps aborting them, and not in XO resolution, then it might be that in Cryptococcus the lack of MSH2 is allowing meiosis to proceed far enough to generate viable products without the resolution of interactions into XOs. Perhaps these invasions are resolved as non-crossovers, even after meiosis is completed. Given that the progeny of the bilateral msh2 hybrids are near diploid hybrids themselves, meiosis appears to result in a skipping of the reductional division which might occur when all the homeologous chromosomes are linked by strand invasions. They might not be resolvable as XOs or if they are, they may be inviable (an explanation of the reduced frequency of meiotic structures in these mutant hybrids?) such that only those that do not complete recombination are viable. This is just speculation of course and not easily addressed experimentally here for this paper, but the authors should address the issue that MSH2 does have a role in the post-zygotic barrier, but it is an unknown or unexpected role. I think this is more exciting than the conclusion that there are other repressors of meiotic recombination.
More specific comments:
The near haploid and near diploid nature of all the progeny assessed makes sense from the assessment of the tolerance of aneuploids in S. cerevisiae (Parry and Cox 1970 Genet Res 16) where viable segregants from a triploid cross were all close to euploidy. The data presented here looks similar with the wt cross exhibiting both, the heterozygous msh2 crosses exhibiting various levels of heterozygosity in the near diploid progeny with lower levels consistent with aneuploidy (2n-1). The progeny of the msh2 homozygous cross appear to be near diploid with whole complements of each parental species which is different than random segregation of chromosomes and look more like Meiosis I non-disjunction. One question – in the 3 near haploid progeny of the wt cross are the chromosomes present a mixture of each parental species?
The lack of detections of any possible increased recombination could be due to various factors. Given that the homozygous msh2 progeny were all near diploid and heterozygous across the genome, the authors did the right thing to look for cryptic XOs (in just the sequence) by looking for reads that crossed breakpoints. In general, when XOs are resolved they will be in the longer tracts of homology. Depending on the distribution of the SNPs in the two species and the range of homologous tract lengths (are there any of 200 bp long for example?), it is possible that most XOs will be in tracts longer than the reads could cover and therefore such recombination events would be missed. Alternatively, it is possible that the recombination events lead to death as is the case in S. cerevisiae where one partner of two recombining chromosomes is unresolved (Chambers 1996) resulting in a dead spore while the other partner survives. This isn’t consistent with the results presented here but all viable progeny here appear to be MI non-disjunctions which may hide such events by resolving them after meiosis.
Line 430 – If there is little or no recombination then QTL mapping will be difficult.
Line 491 – it’s 15% divergent not 85% divergent
Line 514 – loss of MSH2 clearly did reduce the boundary between these species as there was an increase in viability of meiotic products – many of these displayed transgressive sexual behaviour as well, meaning that it may be heritable. Question – were basidiospores dissected from these progeny? How viable were they?
Have all data underlying the figures and results presented in the manuscript been provided?
Large-scale datasets should be made available via a public repository as described in the PLOS Genetics data availability policy, and numerical data that underlies graphs or summary statistics should be provided in spreadsheet form as supporting information.
Reviewer #1: Yes
Reviewer #2: Yes
Reviewer #3: Yes
PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.
If you choose “no”, your identity will remain anonymous but your review may still be made public.
Reviewer #1: No
Reviewer #2: No
Reviewer #3: No
4 Dec 2020
Dear Dr Priest,
We are pleased to inform you that your manuscript entitled "Factors enforcing the species boundary between the human pathogens Cryptococcus neoformans and Cryptococcus deneoformans" has been editorially accepted for publication in PLOS Genetics. Congratulations!
The reviewers were particularly complimentary of the careful revisions you made - bravo! As you will see (below) two of the reviewers had some final small suggestions that you should consider as you prepare your final draft for the production team (the editorial team will not need to re-evaluate).
Before your submission can be formally accepted and sent to production you will need to complete our formatting changes, which you will receive in a follow up email. Please be aware that it may take several days for you to receive this email; during this time no action is required by you. Please note: the accept date on your published article will reflect the date of this provisional acceptance, but your manuscript will not be scheduled for publication until the required changes have been made.
Once your paper is formally accepted, an uncorrected proof of your manuscript will be published online ahead of the final version, unless you’ve already opted out via the online submission form. If, for any reason, you do not want an earlier version of your manuscript published online or are unsure if you have already indicated as such, please let the journal staff know immediately at firstname.lastname@example.org.
In the meantime, please log into Editorial Manager at https://www.editorialmanager.com/pgenetics/, click the "Update My Information" link at the top of the page, and update your user information to ensure an efficient production and billing process. Note that PLOS requires an ORCID iD for all corresponding authors. Therefore, please ensure that you have an ORCID iD and that it is validated in Editorial Manager. To do this, go to ‘Update my Information’ (in the upper left-hand corner of the main menu), and click on the Fetch/Validate link next to the ORCID field. This will take you to the ORCID site and allow you to create a new iD or authenticate a pre-existing iD in Editorial Manager.
If you have a press-related query, or would like to know about making your underlying data available (as you will be aware, this is required for publication), please see the end of this email. If your institution or institutions have a press office, please notify them about your upcoming article at this point, to enable them to help maximise its impact. Inform journal staff as soon as possible if you are preparing a press release for your article and need a publication date.
Thank you again for supporting open-access publishing; we are looking forward to publishing your work in PLOS Genetics!
Gregory P. Copenhaver
Comments from the reviewers (if applicable):
Reviewer's Responses to Questions
Comments to the Authors:
Please note here if the review is uploaded as an attachment.
Reviewer #1: I am impressed by the efforts made by the authors to address my concerns, especially their analysis of intraspecific crosses. I feel that this new work provides a better understanding of the phenotypes seen in the msh2 hybrid crosses. It also helped me better understand the Discussion (beginning on line 489) where the authors discussed why the msh2null mutant hybrid progeny displayed higher ploidy levels compared to wild-type. The possibility of divergent chromosomes being linked by strand invasions is an interesting idea, but then it requires some interesting explanations-they are resolved either through non crossovers or crossovers, with crossovers resulting in inviability. This will clearly require some sorting out in the future. With that said, I am curious what might be found in future experiments in sgs1 and mph1 mutant backgrounds.
Lastly, I find it curious that the hybrid progeny in msh2 remain aneuploid and as shown in S4 Table it appears that they show 2N or 4N content rather than 1N. If I understand this correctly it would make it difficult to identify heteroduplex DNA in the progeny of the random spores dissection. Perhaps adding a short sentence to this effect would be valuable (unless of course I interpreted this incorrectly!).
Reviewer #2: I thought Priest et al did a fantastic job responding to my previous comments and questions. The added experiments, analyses, figures, and tables greatly improve the manuscript. I think their strengthened results regarding the production of hyphae by hybrids under abnormal conditions (Figure 3B and TableS3) are really exciting and interesting. I really appreciated the thoughtful responses and clear effort that went into addressing reviewer feedback.
Lines 210-214: Include statistics
Lines 264-278: Any idea why the bilateral intra-species germination frequencies are so different between the species? (a bilateral msh2Δ C. neoformans intraspecific cross germinated on average only 71% of the time; progeny from bilateral msh2Δ C. deneoformans intraspecific crosses germinated only 36% of the time)
Reviewer #3: I am satisfied with the responses to my and other reviews.
Have all data underlying the figures and results presented in the manuscript been provided?
Large-scale datasets should be made available via a public repository as described in the PLOS Genetics data availability policy, and numerical data that underlies graphs or summary statistics should be provided in spreadsheet form as supporting information.
Reviewer #1: Yes
Reviewer #2: Yes
Reviewer #3: Yes
PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.
If you choose “no”, your identity will remain anonymous but your review may still be made public.
Reviewer #1: No
Reviewer #2: No
Reviewer #3: No
If you have submitted a Research Article or Front Matter that has associated data that are not suitable for deposition in a subject-specific public repository (such as GenBank or ArrayExpress), one way to make that data available is to deposit it in the Dryad Digital Repository. As you may recall, we ask all authors to agree to make data available; this is one way to achieve that. A full list of recommended repositories can be found on our website.
The following link will take you to the Dryad record for your article, so you won't have to re‐enter its bibliographic information, and can upload your files directly:
More information about depositing data in Dryad is available at http://www.datadryad.org/depositing. If you experience any difficulties in submitting your data, please contact email@example.com for support.
Additionally, please be aware that our data availability policy requires that all numerical data underlying display items are included with the submission, and you will need to provide this before we can formally accept your manuscript, if not already present.
If you or your institution will be preparing press materials for this manuscript, or if you need to know your paper's publication date for media purposes, please inform the journal staff as soon as possible so that your submission can be scheduled accordingly. Your manuscript will remain under a strict press embargo until the publication date and time. This means an early version of your manuscript will not be published ahead of your final version. PLOS Genetics may also choose to issue a press release for your article. If there's anything the journal should know or you'd like more information, please get in touch via firstname.lastname@example.org.
14 Jan 2021
Factors enforcing the species boundary between the human pathogens Cryptococcus neoformans and Cryptococcus deneoformans
Dear Dr Priest,
We are pleased to inform you that your manuscript entitled "Factors enforcing the species boundary between the human pathogens Cryptococcus neoformans and Cryptococcus deneoformans" has been formally accepted for publication in PLOS Genetics! Your manuscript is now with our production department and you will be notified of the publication date in due course.
The corresponding author will soon be receiving a typeset proof for review, to ensure errors have not been introduced during production. Please review the PDF proof of your manuscript carefully, as this is the last chance to correct any errors. Please note that major changes, or those which affect the scientific understanding of the work, will likely cause delays to the publication date of your manuscript.
Soon after your final files are uploaded, unless you have opted out or your manuscript is a front-matter piece, the early version of your manuscript will be published online. The date of the early version will be your article's publication date. The final article will be published to the same URL, and all versions of the paper will be accessible to readers.
Thank you again for supporting PLOS Genetics and open-access publishing. We are looking forward to publishing your work!
With kind regards,
On behalf of:
The PLOS Genetics Team
Carlyle House, Carlyle Road, Cambridge CB4 3DN | United Kingdom
email@example.com | +44 (0) 1223-442823
plosgenetics.org | Twitter: @PLOSGenetics