Unisexual reproduction promotes competition for mating partners in the global human fungal pathogen Cryptococcus deneoformans

Courtship is pivotal for successful mating. However, courtship is challenging for the Cryptococcus neoformans species complex, comprised of opportunistic fungal pathogens, as the majority of isolates are α mating type. In the absence of mating partners of the opposite mating type, C. deneoformans can undergo unisexual reproduction, during which a yeast-to-hyphal morphological transition occurs. Hyphal growth during unisexual reproduction is a quantitative trait, which reflects a strain’s ability to undergo unisexual reproduction. In this study, we determined whether unisexual reproduction confers an ecological benefit by promoting foraging for mating partners. Through competitive mating assays using strains with different abilities to produce hyphae, we showed that unisexual reproduction potential did not enhance competition for mating partners of the same mating type, but when cells of the opposite mating type were present, cells with enhanced hyphal growth were more competitive for mating partners of either the same or opposite mating type. Enhanced mating competition was also observed in a strain with increased hyphal production that lacks the mating repressor gene GPA3, which contributes to the pheromone response. Hyphal growth in unisexual strains also enables contact between adjacent colonies and enhances mating efficiency during mating confrontation assays. The pheromone response pathway activation positively correlated with unisexual reproduction hyphal growth during bisexual mating and exogenous pheromone promoted bisexual cell fusion. Despite the benefit in competing for mating partners, unisexual reproduction conferred a fitness cost. Taken together, these findings suggest C. deneoformans employs hyphal growth to facilitate contact between colonies at long distances and utilizes pheromone sensing to enhance mating competition.


Introduction
Successful courtship is key to the evolutionary success of sexual organisms, and many species have evolved distinct strategies to locate and choose a mating partner. For example, primates and humans utilize aggression to secure a mating partner [1]; male hummingbirds apply acoustic control using tail feathers during high-speed dives to court females [2]; male Drosophila vibrate their wings to generate different songs to trigger mating responses in females [3]; male tree-hole frogs also adopt acoustic strategies taking advantage of tree trunk cavities to attract females [4]; and female pipefish display a temporal striped pattern ornament to woo male partners [5]. These examples demonstrate that complex eukaryotic organisms can employ visual, vocal, or mechanical tactics to secure a mate and transmit their genetic traits to the next generation.
In eukaryotic fungal systems, mating often involves a morphological transition. Saccharomyces cerevisiae yeast cells undergo polarized growth and form shmoo projections in preparation for cell fusion during mating [6]. In filamentous fungi, including both ascomycetes and basidiomycetes, sexual reproduction involves the formation of a fruiting body (perithecium or basidium, respectively) [7]. Candida albicans, an ascomycete, undergoes a white-opaque switch to initiate mating [8]. Despite their divergent sexual strategies, these morphological transitions are all controlled by the pheromone response pathway [9]. During yeast mating, physical agglutination of yeast cells does not promote courtship, but rather a gradient of pheromone signals is crucial for successful cell-cell fusion during early mating [10,11]. Similarly, in Schizosaccharomyces pombe, local pheromone signals and a spatially focal pheromone response dictate cell-cell pairing and fusion position during early mating processes [12,13]. In C. albicans, overexpression of the pheromone response MAP kinase pathway components can enhance mating efficiency [14]. These studies establish that the pheromone response pathway plays a critical role in promoting fungal mating efficiency.
The opportunistic human fungal pathogen Cryptococcus deneoformans undergoes a yeastto-hyphal morphological transition upon mating induction [15]. This species has two modes of sexual reproduction: bisexual reproduction between cells of opposite mating types and unisexual reproduction involving cells of only one mating type [15][16][17]. Cell fusion between MATa and MATα cells during bisexual reproduction, and between two MATα cells during unisexual reproduction, triggers hyphal development [18]. This morphological transition is orchestrated by the pheromone response pathway [18,19]. However, recent studies have shown that hyphal growth during unisexual reproduction can also occur independent of cell fusion and the pheromone response pathway [20][21][22][23], and that pheromone-independent hyphal development is dependent upon the calcineurin pathway [20,24].
Because the majority of identified natural and clinical C. neoformans isolates are of the α mating type, unisexual reproduction likely has significant ecological impacts on the Cryptococcus species complex population structure and diversity [25][26][27]. The limited abundance of MATa cells in natural environments restricts outcrossing and in the absence of a-α mating, unisexual reproduction has been shown to reverse Muller's rachet and offset the low abundance of MATa cells to avoid an evolutionary dead end [28]. Unisexual reproduction can also generate genotypic and phenotypic diversity de novo [29]. Interestingly, population genetics studies have revealed that genome recombination occurs frequently among environmental isolates [30][31][32], even those that are exclusively α mating type, providing evidence that unisexual reproduction involving fusion of MATα cells of distinct genotypes allows meiotic recombination in nature. Despite these evolutionary benefits, cell fusion-independent solo-unisexual reproduction also occurs and because this pathway involves genetically identical genomes, it does not contribute to genome reshuffling or recombination. Similar to pseudohyphal differentiation in S. cerevisiae, C. deneoformans hyphal growth during unisexual reproduction has an ecological benefit in promoting foraging for nutrients and habitat exploration in the surrounding environments [33,34]. In this study, we address whether the ability to undergo unisexual reproduction has an additional ecological benefit in promoting foraging for mating partners to facilitate outcrossing and enable recombination in nature.

Results and discussion
Strains with enhanced unisexual reproduction potential are more competitive for mating partners of the opposite mating type During C. deneoformans solo-unisexual reproduction, cells undergo the yeast-to-hyphal morphological transition independent of cell fusion and nuclei diploidized through endoreplication [16,23]. The hyphal growth is a quantitative trait associated with unisexual reproduction that can be used to determine a strain's ability to undergo unisexual reproduction [35]. Although solo-unisexual reproduction occurs independently of cell-cell fusion, cells can fuse with partners of both the same or opposite mating type at varying frequencies [16,23]. To test whether the ability to undergo unisexual reproduction impacts competition for mating partners during outcrossing, we performed mating competition experiments employing three MATα and three MATa C. deneoformans strains with different degrees of unisexual reproduction potential based on their abilities to produce hyphae ( Fig 1A) [35]. Among these strains, several were F2 progeny derived from crosses between the environmental MATa isolate NIH433 and the clinical MATα isolate NIH12 including a high hyphal (HH) strain XL190α, an intermediate hyphal (MH) strain XL280α, a low hyphal strain (LH) XL187a, and a no hyphal (NH) strain JEC20a [15,16,[36][37][38]. LH strain JEC21α and MH strain XL280a are congenic strains of JEC20a and XL280α, respectively, derived through 10 rounds of backcrossing (S1 Fig) [36,38,39]. For each mating competition experiment, cells of three strains with different hyphal growth carrying dominant, selectable drug resistance markers were mixed, spotinoculated, and incubated on V8 agar media for 4 days (Fig 1B). Cells were recovered on YPD medium to obtain colony forming units (CFU), and on YPD medium supplemented with different two-drug combinations to determine the cell fusion frequencies. Cell fusion frequencies were compared between different pairs of strains within the same competition mating mixture to determine whether the ability to undergo unisexual reproduction confers benefits in competition for mating partners to facilitate outcrossing ( Fig 1B).
Prior to the mating competition experiments, cell fusion frequencies were compared between different hyphal strains. During α-α cell fusion, the MATα MH strain displayed a significantly higher cell fusion frequency (5 cell fusion events per million CFU) compared to the HH and LH strains (0.013 and 0.019 cell fusion events per million CFU, respectively), in which cell fusion rarely occurred (Fig 1C). This suggests that the ability to undergo more robust hyphal growth is not strictly correlated with α-α cell fusion efficiency. In contrast, during a-α cell fusion, hyphal growth positively correlated with a-α cell fusion efficiency. MH-HH strains had a cell fusion frequency (53 cell fusion events per thousand CFU) about 109 times higher than LH-MH strains, which in turn had a cell fusion frequency (0.49 cell fusion events per Foraging for mating in Cryptococcus deneoformans thousand CFU) about 26 times higher than NH-LH strains (0.019 cell fusion events per thousand CFU) (Fig 1D). In all of the strains tested, a-α cell fusion occurred at a much higher level compared to α-α cell fusion, similar to previous findings [16,23]. Cell fusion has been previously shown to be dispensable for solo-unisexual reproduction [19,23], which can account for the observed poor correlation between hyphal growth and α-α cell fusion frequency. Thus, we hypothesize that increased hyphal growth may not provide an advantage in competing for mating partners of the same mating type. Indeed, when we performed the unisexual mating competition assay mixing the HH, MH, and LH cells, we observed that HH and LH cells yielded the most fusion products with a cell fusion frequency of 1.3 cell fusion events per million CFU that is not significantly different from cell fusion frequencies involving MH cells (Fig 1E) that exhibited the highest cell fusion frequency (Fig 1C). These findings indicate that neither α-α cell fusion frequency nor hyphal growth can be used to predict mating partner preference during unisexual reproduction, which supports the hypothesis that the ability to undergo unisexual reproduction does not promote competition for mating partners of the same mating type.
To test whether the propensity for unisexual reproduction plays a role in competing for mating partners of the opposite mating type, mating competition assays were conducted for a given MATa isolate between two MATα strains of different hyphal growth phenotypes ( Fig  2A). Interestingly, cells capable of producing more hyphae always had a significantly higher cell fusion frequency with MATa cells compared to cells with lower hyphal growth potential (Fig 2A and S1 Table). For example, in the presence of MH MATa cells, HH MATα cells fused with MATa cells 24 times more efficiently than LH MATα cells (green bar, circles vs squares) and 8.1 times more efficiently than MH MATα cells (red bar, circles vs triangles), and MH MATα cells fused with MATa cells 5.8 times more efficiently than LH MATα cells (yellow bar, triangles vs squares; S1 Table). These results suggest that increased hyphal growth correlates with competition for mating partners of the opposite mating type during bisexual reproduction. It was also noted that the mating competition advantage decreased for each competition pair (24, 14.5, and 8.9 fold differences for HH vs LH, 8.1, 6.5, and 5.3 fold differences for HH vs MH, and 5.8, 4.6, and 1.7 fold differences for MH vs LH) with the decreasing hyphal phenotype of the MATa cells (Fig 2A and S1 Table), suggesting that increased hyphal growth of MATa cells can also promote cell fusion.
Besides the observation that hyphal growth enhanced competition for mating partners of the opposite mating type, the presence of higher hyphal MATα cells also stimulated a-α cell fusion.  Table). Similar trends were observed during competition for LH MATa cells in that the presence of MH MATα cells increased LH MATa and HH MATα cell fusion frequency by 4.5-fold compared to the presence of LH MATα cells (circles in red bar vs circles in green bar in Fig 2A a LH group; medium yellow-shaded cells in S1 Table), and the presence of  Table). Interestingly, no enhancement of cell fusion frequency by high hyphal MATα cells was observed during competition for NH MATa cells (Fig 2A a NH group; light colorshaded cells in S1 Table). Notably, the enhancement of cell fusion frequency by higher hyphal MATα cells did not occur when either MATa NH or MATα LH cells were involved in a-α cell fusion, suggesting that strains with poor unisexual reproduction potential have a disadvantage in competing for mating partners of the opposite mating type.
The ability to undergo unisexual reproduction also correlated with cell fusion between cells of the same mating type when a MATa partner is present. In the presence of MH or LH MATa cells, HH and MH MATα cells fused at higher frequencies (26 and 27 cell fusion events per Foraging for mating in Cryptococcus deneoformans million CFU, respectively) compared to HH and LH MATα cells (8.1 and 3.4 cell fusion events per million CFU, respectively), and in the presence of MH, or LH, or NH MATa cells, HH and LH MATα cells fused at higher frequencies (8.1, 3.4, and 2.8 cell fusion events per million CFU, respectively) compared to MH and LH MATα cells (0.71, 0.28, and 0.14 cell fusion events per million CFU, respectively) ( Fig 2B), suggesting that increased hyphal growth correlated with enhanced α-α cell fusion frequency in the presence of MATa cells. We also observed a trend where α-α cell fusion frequencies (HH and MH, HH and LH, and MH and LH) decreased with reduced hyphal MATa cells (Fig 2B), suggesting that the presence of more robust hyphal MATa cells can further enhance α-α cell fusion. In summary, strains with robust hyphal production have an advantage in competing for mating partners of the opposite mating type, and also for mating partners of the same mating type when cells of the opposite mating type are present.

gpa3Δ mutation enhances competition for mating partners of the same or opposite mating type
The pheromone response pathway plays an important role in the yeast-to-hyphal morphological transition during C. deneoformans sexual reproduction. This signaling cascade is controlled by G proteins and RGS proteins, including the Gα protein Gpa3 which represses hyphal growth during mating [40][41][42][43]. To further examine the impact of the ability to undergo unisexual reproduction during mating competition, we generated strains enhanced for hyphal production by deleting the GPA3 gene in the LH strain JEC21α. gpa3Δ mutants exhibited significantly increased hyphal growth during both unisexual and bisexual reproduction compared to the parental strain ( Fig 3A). Next, mating competition assays were conducted using the enhanced hyphal (EH) strain JEC21α gpa3Δ to test its ability to compete for mating partners.
Similar to the observation in HH, MH, and LH strains, enhanced hyphal production did not increase cell fusion between MATα cells but did increase cell fusion frequency by 4-fold between MATa and MATα cells compared to the parental LH strain (Fig 3B and 3C). However, the increase is not statistically significant due to the low cell fusion frequencies between strains of low hyphal background. Unisexual mating competition assays were performed to compare the abilities of LH and EH MATα cells to fuse with MH MATα cells. In the control assay, cell fusion frequencies were comparable between cells of all three strain combinations (MH with LH-NAT, MH with LH-NEO, and LH-NAT with LH-NEO) ( Fig 3D). In the assay mixing LH, MH, and EH cells, EH cells fused with MH cells at a significantly higher frequency of 2.8 cell fusion events per million CFU compared to LH cells (85-fold) (Fig 3E), suggesting that deletion of GPA3 increases competitiveness for mating partners of the same mating type. In the mating competition during bisexual reproduction, no advantage was observed in the fusion of NH MATa cells with either EH or the parental LH MATα cells (Fig 3G). However, a significant 2.9-fold increase was observed in total a-α cell fusion events during mating competitions for NH MATa cells between EH and LH MATα cells compared to control mating competitions for the same MATa cells between LH-NAT and LH-NEO MATα cells (Fig 3F and  3G), indicating that presence of cells with enhanced ability to undergo unisexual reproduction allows both EH and LH MATα cells to fuse with MATa mating partners more efficiently during bisexual reproduction. A significant 2.9-fold increase was observed in α-α cell fusion between EH and LH MATα cells in the presence of NH MATa cells compared to cell fusion between LH-NAT and LH-NEO MATα cells (Fig 3H), suggesting that in the presence of MATa cells, GPA3 deletion also enhances competition for mating partners of the same mating type. Overall, this analysis of the enhanced hyphal growth strain JEC21α gpa3Δ provides additional support for models in which increased unisexual reproduction potential enhances competition for mating partners.

Hyphal growth promotes foraging for mating partners
Unisexual reproduction provides evolutionary and ecological benefits for C. deneoformans by generating aneuploid progeny with phenotypic diversity and by promoting habitat exploration through hyphal growth [29,34]. Here we further show that unisexual cells have an advantage in competing for mating partners within the same colony. We tested whether hyphal growth during unisexual reproduction confers benefits in foraging for mating partners.
Both long-term and short-term foraging for mating partner experiments suggested that hyphal growth promoted foraging for mating. In a six-week mating confrontation experiment, hyphae of different MATα unisexual reproduction backgrounds marked with NEO grew towards the same MATa cells marked with HYG that were 4 mm apart (Fig 4A and 4B). When competing for either the same MATa or MATα cells (except LH MATα cells), although not all pairwise comparisons by t-test were significant due to the lack of contact when strains of no or low hyphal growth were involved, a significant trend by one-way ANOVA was observed that isolates with more hyphal growth yielded more double drug resistant colonies than isolates with reduced hyphal growth ( Fig 4C and S2 Table, seven group one-way ANOVA analyses are listed above pairwise Welch's t-test analyses). In a seven-day mini-colony mating experiment, colonies derived from single cells produced hyphae that allowed contact with adjacent colonies of the opposite mating type (S2A Fig). Similar to the long-term mating confrontation experiment, a significant trend by one-way ANOVA was observed in which isolates with more hyphal growth had an advantage in forming double drug resistant colonies (S2B Fig and S3  Table). Although pairwise comparisons by t-test showed significant differences between crosses involved NH and LH cells that were not observed in the confrontation experiment, this discrepancy is due to differences in the experimental setup where cells were inoculated at 0.4 cm apart during confrontation, whereas randomly plated on agar media during mini-colony mating experiment, where colony contact is enabled by both hyphal growth and chance. Overall, these results suggest that hyphal growth during unisexual reproduction can facilitate contact between mating partners in adjacent environments.

Pheromone response pathway activation is correlated with hyphal growth phenotype during bisexual reproduction
Elevated pheromone response pathway activation and increased response to pheromones are critical to successful courtship during mating in S. cerevisiae and C. albicans [10,11,14]. S. cerevisiae utilizes the α-factor protease Bar1 and a-factor barrier Afb1 to discriminate mating partners with different pheromone levels and drive evolution towards higher pheromone production for efficient mating [44][45][46][47]. Pheromones also stimulate mating and the yeast-tohyphal morphological transition during C. neoformans bisexual reproduction [48]. To determine the role of the pheromone response pathway during C. deneoformans mating competition, expression levels were examined for the genes encoding the pheromones MFα and MFa, the pheromone receptors Ste3α and Ste3a, the MAP kinase Cpk1, the transcription factors Mat2 and Znf2, and the plasma membrane fusion protein Prm1 [23,40,49].
Pheromone response pathway activation did not correlate with the hyphal growth phenotype in MATa or MATα strains. In MATa strains, all of the pheromone response pathway genes were significantly upregulated in the MH strain XL280a compared to the LH and NH strains, but only MFa and PRM1 were significantly upregulated in the LH strain XL187a compared to the NH strain JEC20a (S3 Fig). In MATα strains, all of the pheromone response pathway genes were significantly upregulated in the HH and MH strains compared to the LH strain JEC21α, but the MH strain XL280α had a significantly higher pheromone pathway activation compared to the HH strain XL190α (S3 Fig). The pheromone pathway was significantly upregulated in the EH strain JEC21α gpa3Δ compared to the parental LH strain JEC21α (S3 Fig). These expression analyses suggest that pheromone response pathway activation per se is not sufficient to explain the ability to undergo unisexual reproduction and its association with competition for mating partners. It was previously shown that the cell fusion protein Prm1 is not required for unisexual reproduction [23], and certain environmental factors, such as copper and glucosamine, can induce hyphal growth independently of the pheromone response Foraging for mating in Cryptococcus deneoformans pathway [20,21]. In a recent study on the quorum sensing peptide Qsp1, deletion of pheromone and pheromone receptor genes had little impact on hyphal growth during unisexual reproduction [22], further indicating the polygenic nature of unisexual reproduction.
Despite the incongruent association of the pheromone response pathway and the ability to undergo unisexual reproduction, pheromone response pathway activation was positively correlated with the hyphal growth during bisexual reproduction. The α pheromone gene MFα, both a and α pheromone receptor genes STE3α and STE3a, and the plasma membrane fusion gene PRM1 showed significant correlation with the hyphal growth phenotype (Fig 5A). Although MFa expression was lower in the cross between a MH and α HH strains compared to a LH and α MH strains, the significant upregulation of the MFα, STE3α, and STE3a genes may compensate for the overall pheromone response activation (Fig 5A). The gene expression patterns of two transcription factors Mat2 and Znf2 that regulate yeast-to-hyphal morphological transition and mating significantly correlated with the hyphal growth except for the cross between a MH and α HH strains. Nonetheless, these two genes were expressed at higher levels compared to the cross between a LH and α MH strains (Fig 5A). The expression of the MAP Kinase CPK1 gene poorly correlated with hyphal growth (Fig 5A), which is likely due to posttranslational control of the MAP kinase through phosphorylation, which can relieve a requirement for expression level upregulation during pathway activation. Overall, the pheromone response pathway activation is largely congruent with the hyphal growth phenotype suggesting that in the presence of cells of the opposite mating type, unisexual cells are capable of upregulating the pheromone response pathway in both MATa and MATα cells to compete for mating partners.
To validate that pheromone contributes to mating competitiveness, we tested whether synthetic α pheromone promotes a-α cell fusion. Indeed, exogenous α pheromone promoted cell fusion between a NH and α LH cells in a dose-dependent manner (Fig 5B). However, the enhancement of cell fusion frequency by pheromone is limited, compared to the 2788-fold increase of cell fusion frequency between a MH and α HH cells over a NH and α LH cells ( Fig  1D) that coincided with a 34.5-fold increase in α pheromone expression (Fig 5A). Interestingly, the mild increase in cell fusion by exogenous pheromone is not observed in the cross between a NH and α EH cells (S4A Fig), which is likely due to the saturation of Ste3a by the 1184-fold higher α pheromone expressed by α EH cells (S3 Fig). The less than two-fold increase in cell fusion provided by 5 μM exogenous pheromone suggested that changes in α pheromone alone are not able to tip the balance and drive the entire pheromone response pathway towards a stronger increase in mating efficiency. In support, exogenous supply of 500 nM pheromone provided limited impact in cell fusion between cells of higher hyphal growth phenotypes (S4A Fig); and when hyphal growth was suppressed under nutrient rich conditions, exogenous pheromone had little to no impact on cell fusion (S4B Fig). In response to the pheromone signal, S. cerevisiae undergoes filamentous growth to enhance the probabilities of cells finding a mating partner [50]. Here we observed that the a EH colonies responded to 5 μM α pheromone peptide and produced abundant hyphae (Fig 5C), similar to previous report [51], suggesting that pheromone can promote hyphal growth and increase the contact opportunities between adjacent colonies within the same environment.

gpa3Δ mutation resulted in a fitness cost
Upregulation of the pheromone response pathway enhances mating efficiency; however, this upregulation can result in a fitness cost in S. cerevisiae and C. albicans [14,52]. In yeast, a short-term experimental evolution experiment showed that mutations abrogating expression Foraging for mating in Cryptococcus deneoformans of 23 genes involved in mating conferred a fitness benefit during yeast growth when functions of these genes are not required [52]. In C. albicans, cells undergo a white-opaque switch and upregulate the pheromone response pathway, which results in a fitness cost for the opaque cells [14]. Given that the pheromone response pathway is activated at a higher level in C. deneoformans strains with more hyphal growth during bisexual reproduction (Fig 5A), we investigated whether unisexual reproduction confers a fitness cost.
Growth curve analyses in YPD liquid medium were performed using an automated Tecan Sunrise absorbance reader to determine the fitness of different strains. The cultures were agitated with vigorous shaking for one minute bihourly before each OD 600 measurement. The minimum agitation allows for differentiation of growth curve kinetics among the strains tested, which would otherwise be indistinguishable when grown on solid YPD medium. Compared to the low hyphal growth strains JEC21α and JEC20a, gpa3Δ mutants exhibited a growth defect in nutrient rich media, suggesting that deletion of GPA3 resulted in a fitness cost (S5 Fig). However, this fitness cost was not due to the yeast-to-hyphal morphological transition as hyphal growth was not present. Interestingly, deletion of the GPA3 gene in the sister species C. neoformans did not induce hyphal growth or cause a growth defect compared to the nonhyphal strain KN99a (S5 Fig). It has been shown that the pheromone response pathway activation by GPA3 gene deletion is lower in KN99a than in JEC21 or JEC20, indicating that deletion of GPA3 is not sufficient to rewire cellular responses to induce unisexual reproduction in C. neoformans [40].
We next performed fitness competition experiments. After 10 days of incubation of equally mixed cells on YPD and V8 agar medium, cells were collected and plated on selection media to determine colony forming units for each competition strain. Hyphal growth was observed on both YPD and V8 media when a MH cells were present, and both yeast cells and hyphae were collected to compare fitness (S6 Fig). During competition between two LH strains, cells were recovered at about 1 to 1 ratio after 10-day incubation on both YPD and V8 media both in the absence or in the presence of a cells (Fig 6A). In contrast, LH strain outcompeted EH strain when cells were incubated on V8 medium or when a cells were present (Fig 6B). On YPD medium in the absence of a cells, the EH strain that displayed poor growth in liquid media (S5 Fig) outcompeted the parental LH strain (Fig 6B), which is likely due to differential cellular responses under different growth conditions. However, this competition advantage was reversed when a cells were present or when mixed cells were incubated on V8 medium (Fig 6B), suggesting that the presence of cells of the opposite mating type or the mating inducing environment can elicit a fitness cost. This competition disadvantage for the EH strain on V8 medium was further exacerbated when a cells were present during competition (Fig 6B). These fitness competition assays demonstrate that gpa3Δ mutation enhances mating competition at a cost of growth fitness, and this fitness cost is likely due to the energy required in the expression of the pheromone response pathway genes. It has been reported that long-term passage on rich media in the lab often diminishes the hyphal growth phenotype of MATa strains, which further suggests that there is a fitness cost associated with the ability to undergo unisexual reproduction [53]. Interestingly, in S. cerevisiae, the [SWI + ] prion state promotes outcrossing efficiency due to a defect in HO expression and mating type switching, which also leads to a fitness cost [54]. Taken together, fungal species have evolved different strategies in promoting mating in nature accompanied with a fitness tradeoff.

Conclusion
Sexual reproduction plays a pivotal role in shaping fungal population structure and diversity in nature. However, studies on how fungi secure a mating partner in nature for successful mating are limited. In this study, we aimed to characterize the ecological and evolutionary benefits of unisexual reproduction in C. deneoformans. Similar to the landmark study by Jackson and Hartwell showing that higher pheromone production promotes courtship in S. cerevisiae [11], we showed that strains with higher potential for unisexual reproduction are more competitive for mating partners of both the same and the opposite mating types when cells of the opposite mating type are present, and the pheromone response pathway activation is positively correlated with the hyphal growth phenotype (Fig 7). More interestingly, in addition to  pheromone sensing, unisexual cells employ hyphal growth to increase contact opportunities between colonies at relatively long distances. However, this mating competition advantage results in a fitness cost for unisexual cells during mitotic growth under mating-inducing conditions.
The strains involved in this study were all derived from natural and clinical isolates under laboratory conditions, suggesting that the ability to undergo unisexual reproduction is likely to span a broad range in the environment. The majority of natural and clinical isolates of the Cryptococcus species complex are found to be of the α mating type, which accounts for 99% of C. neoformans isolates [25][26][27]. In a survey of C. deneoformans environmental distribution around the Mediterranean basin, 27% of isolates are MATa isolates, which were all recovered in Greece, suggesting certain environmental niches harbor MATa cells [27]. Genomic and genetic evidence also suggest that recombination is prevalent among these environmental isolates, and some isolates are isolated from a single Eucalytus hollow, which underscores that mating occurs in nature [30][31][32]. Although sexual structures of Cryptococcus species have yet to be documented in nature, plant material-based media such as V8, on which we conducted the mating competition assays, can readily induce sexual reproduction under laboratory conditions, suggesting that the mating competition we observed could happen in its environmental niche. We hypothesize that in the presence of MATa cells sparsely distributed in the environment, undirected hyphal growth first enables unisexual MATα cells to forage for mating partners over a much larger surface area than is available to cells within a much more compact budding yeast colony. Next, as MATα hyphae come into the proximity of rare MATa cells, pheromone response pathway activation in both MATα and MATa cells can further enhance mating competition. This mating competition advantage could promote outcrossing and provide an evolutionary advantage by facilitating genome reshuffling via meiotic recombination in a pathogenic yeast species.

Drug-resistant marker strain generation and gene deletion
To generate deletion mutants for GPA3, a deletion construct consisting of the 5' upstream and 3' downstream regions of GPA3 gene flanking the NEO cassette was generated by overlap PCR as previously described [58]. The GPA3 deletion construct was introduced into the strain JEC21α via biolistic transformation. Transformants were selected on YPD medium supplemented with G418, and gene replacement by homologous recombination was confirmed by PCR. Primers used to generate these deletion constructs are listed in S5 Table. Microscopy Cells were grown on V8 agar medium for seven days or three weeks in the dark at room temperature to allow hyphal formation. Hyphal growth on the edge of mating patches was imaged using a Nikon Eclipse E400 microscope equipped with a Nikon DXM1200F camera.

Competitive mating assays
For each competitive mating assay, cells were grown overnight in YPD liquid medium at 30˚C and adjusted to OD 600 = 0.5 in sterile H 2 O, and then equal volumes of cells marked with different dominant drug resistant markers were mixed and spot inoculated (50 μl) on V8 agar medium. The mating plates were incubated for four days in the dark at room temperature. The cells were harvested and plated in serial dilution on YPD medium and YPD medium supplemented with different two drug combinations (NAT and NEO, NAT and HYG, or NEO and HYG). The cells were incubated for three to five days at room temperature and colony forming units were counted. Cell fusion frequencies were determined as double drug resistant CFU/total CFU. The complete competitive mating experimental design is listed in S6 Table. Each mating competition was performed in biological triplicate.

Foraging for mating assays
To investigate whether hyphal growth enables cells foraging for mating partners, we performed long-term confrontation and short-term mini-colony mating experiments. For the confrontation mating experiment, HYG resistant XL280a, XL187a, and JEC20a were streaked and grown on V8 medium to form a line of cells, and then NEO resistant XL190α, XL280α, and JEC21α, and JEC21α gpa3Δ::NEO were spot inoculated 4 mm apart in parallel along the MATa cells. Unisexual hyphae grew towards cells of the opposite mating type for six weeks. Then the cells were collected and plated on YPD medium supplemented with HYG and NEO. After incubation at room temperature for three to five days, total double drug resistant colony forming units were counted to determine whether hyphal growth conferred an advantage in foraging for mating partners. Each confrontation mating experiment was performed in biological quintuplicate. For the mini-colony mating experiment, the aforementioned HYG resistant MATa strains and NEO resistant MATα strains were grown overnight in YPD liquid medium and adjusted to OD 600 = 0.008. For each mating pair, 100 μl of MATa cells and 100 μl of MATα cells were mixed and plated on V8 agar medium to form evenly spaced mini colonies. Unisexual hyphae facilitate contact between adjacent colonies after growing for seven days. Then the cells were collected and plated on YPD medium and YPD medium supplemented with HYG and NEO. The cells were incubated for three to five days at room temperature and colony forming units were counted. Cell fusion frequencies were determined as double drug resistant CFU/total CFU. Each mating was performed in biological triplicate.

RNA extraction and qRT-PCR
To examine pheromone response pathway activation, qRT PCR experiments were performed on RNA extracted from cells incubated on V8 agar medium for 36 hours as previously described [23]. In brief, XL190α, XL280α, JEC21α, JEC21α gpa3Δ::NEO, XL280a, XL187a, and JEC20a were grown overnight in YPD liquid medium and adjusted to OD 600 = 2 in sterile H 2 O. Then cells of individual strains and an equal-volume mixtures of cells for crosses between XL190α and XL280a, XL280α and XL187a, JEC21α and JEC20a, and JEC21α gpa3Δ:: NEO and JEC20a were spotted (250 μl) on V8 medium and incubated for 36 hours. Cell patches of individual strains and of mixture of a and α strains were scraped off the medium and transferred into Eppendorf tubes then flash frozen in liquid nitrogen. RNA was extracted using TRIzol reagent (Thermo) following the manufacturer's instructions. RNA was treated with Turbo DNAse (Ambion), and single-stranded cDNA was synthesized by AffinityScript RT-RNAse (Stratagene). cDNA synthesized without the RT/RNAse block enzyme mixture was used to control for genomic DNA contamination. The relative expression levels of targeted genes were measured by qRT PCR using Brilliant III ultra-fast SYBR green QPCR mix (Stratagene) in an Applied Biosystems 7500 Real-Time PCR system. A "no template control" was used to analyze the resulting melting curves to exclude primer artifacts for each target gene. Gene expression levels were normalized to the endogenous reference gene GPD1 using the comparative ΔΔCt method. Primers used for qRT-PCR are listed in S5 Table. For each target gene and each sample, technical triplicate and biological triplicate were performed.

Mating assays with exogenous α pheromone peptide
To address whether pheromone promotes mating competition and hyphal growth, carboxyl farnesylated and methylated α pheromone peptide (QEAHPGGMTLC) (synthesized at Gen-Script, USA) was tested for its impact on mating and hyphal growth. α pheromone peptide was dissolved in methanol at the concentration of 50 μM and 10-fold serial dilutions in methanol were prepared as stock solutions. For the mating assay, HYG marked MATa and NEO marked MATα cells were prepared and mixed as mentioned above for the mating competition assays. α pheromone peptide was added to mixed a NH (CF926) and α LH (CF759) cells at the concentrations of 0, 500 pM, 5 nM, 50 nM, 500 nM, and 5 μM, and the mixed cells were spotinoculated on the V8 media and incubated in the dark at room temperature for four days. Cells were then harvested and plated on both YPD and YPD media supplemented with NEO and HYG to determine cell fusion frequency. Same mating assays were carried out for crosses between a MH (CF978) and α HH (CF914), a LH (CF931) and α MH (CF752), a NH (CF926) and α LH (CF759), and a NH (CF926) and α EH (CF1314) both in the absence and in the presence of 500 nM α pheromone peptide on both YPD and V8 media.
To test the impact of α pheromone peptide on hyphal growth, 5 μM α pheromone peptide and methanol were dropped onto MS media and allowed to dry. a EH (YPH86) cells were grown overnight and washed with H 2 O twice, and then inoculated onto the MS plate with dried α pheromone peptide and methanol droplets at a different spot. Cells were then microscopically manipulated and transferred to the α pheromone peptide and methanol spots. Colony hyphal growth was monitored daily and imaged after incubation in the dark at room temperature for 72 hours.

Fitness competition and growth curve assays
Competition experiments were performed to compare fitness between two low hyphal strains and between low and enhanced hyphal strains both in the absence and in the presence of the a MH cells. Overnight cultures of JEC21α NAT, JEC21α NEO, JEC21α gpa3Δ::NEO, and XL280a were washed with H 2 O twice and cell densities were determined with a hemocytometer. For each competition experiment, 10 μl of H 2 O containing 100,000 cells of each strain was spotted on either YPD or V8 agar medium and incubated in the dark at room temperature for 10 days. Cells were collected and plated on YPD medium supplemented with NAT or G418 to determine colony forming units. Cell mixtures were plated before incubation to control for equal mixing. Each competition was performed in triplicate. Fitness was determined by calculating the percentile of the recovered CFU of each strain out of the total recovered dominant drug resistant CFU.
To determine the growth fitness of different unisexual strains, KN99a, KN99a gpa3Δ::NEO, JEC20a, JEC20a gpa3Δ::NEO, JEC21α, and JEC21α gpa3Δ::NEO were grown overnight in YPD liquid medium and washed twice in H 2 O. 10,000 cells for each strain were resuspended in 200 μl YPD liquid medium and incubated in a 96-well plate (Corning) at 30˚C with vigorous shaking for 1 min bihourly. OD 600 readings were measured bi-hourly after shaking using an automated Tecan Sunrise absorbance reader. Each sample was tested in quintuplicate.

Statistical analysis
All statistical analyses were performed using the Graphpad Prism 7 program. Welch's t-test was performed for each pairwise comparison, and one-way ANOVA was performed for each group analysis with a p value lower than 0.05 considered statistically significant ( � indicates 0.01<p�0.05, �� indicates 0.001<p�0.01, ��� indicates 0.0001<p�0.001, and ���� indicates p�0.0001).
Supporting information S1 Fig. Derivation of the strains used in this study. B4478 MATα, JEC20a, XL187a, XL190α, and XL280α are F2 progeny from the cross between F1 progeny B3502 MATa and B3501 MATα, which were derived from a cross between the environmental isolate NIH433 MATa and the clinical isolate NIH12 MATα. JEC20a was then crossed with B4478 MATα, and an α progeny was backcrossed with JEC20a. This process was repeated 9 times to yield the congenic partner B4500 JEC21α of B4476 JEC20a. XL280α was crossed with JEC20a, and a MATa progeny was backcrossed with XL280α. This process was repeated 9 times to yield the congenic partner XL280a.  Table. p-Values of one-way ANOVA analyses and Welch's t-test for each pairwise comparison for the foraging for mating assay during mating confrontation. (DOCX) S3 Table. p-Values of one-way ANOVA analyses and Welch's t-test for each pairwise comparison for the foraging for mating assay during mating among mini-colonies.