Evolution of sexual reproduction: a view from the Fungal Kingdom supports an evolutionary epoch with sex before sexes

Abstract

Sexual reproduction is conserved throughout each supergroup within the eukaryotic tree of life, and therefore thought to have evolved once and to have been present in the last eukaryotic common ancestor (LECA). Given the antiquity of sex, there are features of sexual reproduction that are ancient and ancestral, and thus shared in diverse extant organisms. On the other hand, the vast evolutionary distance that separates any given extant species from the LECA necessarily implies that other features of sex will be derived. While most types of sex we are familiar with involve two opposite sexes or mating types, recent studies in the fungal kingdom have revealed novel and unusual patterns of sexual reproduction, including unisexual reproduction. In this mode of reproduction a single mating type can on its own undergo self-fertile/homothallic reproduction, either with itself or with other members of the population of the same mating type. Unisexual reproduction has arisen independently as a derived feature in several different lineages. That a myriad of different types of sex determination and sex determinants abound in animals, plants, protists, and fungi suggests that sex specification itself may not be ancestral and instead may be a derived trait. If so, then the original form of sexual reproduction may have been unisexual, onto which sexes were superimposed as a later feature. In this model, unisexual reproduction is both an ancestral and a derived trait. In this review, we consider what is new and what is old about sexual reproduction from the unique vantage point of the fungal kingdom.

Introduction

No one knows the exact nature of the LECA, but we think that this ancestor was a unicellular, aquatic, motile creature with one or two flagella. Thus, in some respects the LECA was simple. But in other ways, it was already quite complex, with a nucleus, mitochondria, secretory apparatus, RNAi, and reproducing both asexually and sexually. Thus, when we think of where sex first evolved, it was in the water, involving swimming cells (Levin and King, 2013; Umen and Heitman, 2013). And when we think of how sex first evolved, this involved changes in ploidy and the process of meiosis, given their conserved nature throughout eukaryotes. And while cell-cell and nuclear-nuclear fusion play prominent roles in sexual reproduction today, there may have been an era in which endoreplication cycles followed by meiosis drove the processes of ploidy change during ancestral modes of sexual reproduction. In this view, cell-cell fusion may be ancient, but perhaps not as ancient as other features of sexual reproduction.

Why sex is so pervasive is thought to result from potential benefits conferred by sexual reproduction. These include purging the genome of deleterious mutations and shuffling the genome via independent chromosomal assortment and recombination to give rise to a diverse repertoire of meiotic progeny. Sex may also enable organisms to keep pace with or outrun pathogens, including those both external and those internal (such as transposons). There is sound experimental evidence from studies in Caenorhabditis elegans and in naturally occurring snails in New Zealand for this last hypothesis in which sex allows species to keep pace with their pathogens (King et al., 2009; King et al., 2011; Morran et al., 2011; Vergara et al., 2013). However, these potential benefits of sex are pitted against well-known costs of sexual reproduction: that only 50% of a parental genome is transmitted to any given progeny, the time and energy required to locate mates, and the breaking apart of well adapted genomic configurations.

The core features of sexual reproduction are conserved in organisms as diverse as the model budding yeast Saccharomyces cerevisiae and humans, despite a billion years or more of evolution separating us from our last common shared ancestor. These conserved features include: 1) ploidy changes from haploid to diploid to haploid (or diploid to haploid to diploid), 2) the process of meiosis that enables meiotic recombination and halves the ploidy of the genome, and 3) cell-cell fusion between mating partners (a and α cells) or gametes (the sperm and the egg). This ubiquity of the conserved features of sex again speaks to the antiquity of the process.

Beyond the commonalities in the mechanisms of sex, there are also shared features to the modes of sexual reproduction. This includes outbreeding between genetically divergent members of the population, but also types of inbreeding that can involve the ability of the yeast S. cerevisiae to undergo mating type switching that allows mother cells to mate with their daughter cells. And in humans there are the examples of consanguineous marriages, resulting for example from cousin-cousin pairings, which lead to considerable inbreeding with the risk of exposure of recessive alleles in a homozygous configuration. We will return to this theme of the balance between outbreeding and inbreeding.

Mechanisms of sex determination

Sex in humans and many other animals is determined by the X and Y sex chromosomes, in which individuals with XX karotype are female and those with XY are male. The two sex chromosomes are dramatically different in size, and are referred to as heteromorphic sex chromosomes. A single gene resident on the Y chromosome, SRY, is sufficient to direct male fate and transferring this single gene from the Y to the X chromosome suffices to cause sex reversal in both humans and in mice. But in other plants and animals, there are different mechanisms of sex determination. Some species, such as the plant Papaya and the fish Medaka, have sex chromosomes in which the sex specific region is small and the two sex chromosomes are the same size, so called homomorphic sex chromosomes (Kondo et al., 2004; Liu et al., 2004; Myosho et al., 2012). Chickens and other birds have a completely different type of sex chromosome, called Z and W, and in these lineages it is the heterogametic ZW pattern that specifies female and the homogametic ZZ the male (Zhou et al., 2014). In some animals, including turtles and crocodiles, the temperature at which an egg hatches determines sexual identity and this is called Environmental Sex Determination (ESD) to distinguish it from Chromosomal Sex Determination (CSD) (Barske and Capel, 2008). Yet other species appear to be hybrids of the two with features of both environmental and chromosomal sex determination. Finally in some lines of the zebra fish Danio rerio sex appears to be a quantitative trait, in which genes on multiple different chromosomes come together in allelic combinations that favor either female or male fate (Anderson et al., 2012; Bradley et al., 2011; Liew et al., 2012; Liew and Orban, 2014). This quantitative sex determining system has been termed polygenic sex determination (PSD). Recent studies of wild D. rerio reveal a sex determining region on one end of chromosome 4 that may be consistent with a WZ/ZZ sex chromosome system, suggesting loss of a sex determinant or recent origin of a novel one during domestication (Wilson et al., 2014). To summarize, in simple terms the ways in which sex is determined are plastic and diverse.

What about fungi? Relatively few fungi have large size dimorphic sex chromosomes, but there are a few well studied examples such as Neurospora tetrasperma, Podospora, and Microbotryum (Ellison et al., 2011; Fraser et al., 2004; Fraser and Heitman, 2004; Grognet et al., 2014; Hood et al., 2013; Menkis et al., 2008; Whittle et al., 2015). Most fungi have relatively smaller regions of their genome, called mating-type loci, or MAT for short, that dictate their mating type or sex (Fraser and Heitman, 2003). The paradigmatic example is S. cerevisiae in which a relatively small region of the genome, less than a thousand base pairs, expresses in the alternate mating types one or two key cell fate determinants, all of which are transcription factors responsible for orchestrating both haploid cell type specificity (a or α) and the diploid zygote fate (a/α). Two are homeodomain proteins of the HD1 and HD2 class that form a heterodimer, a1/α2, which is necessary for the diploid zygote fate. The other factor, α1 from MATα, encodes an alpha domain transcription factor necessary for turning on genes required for the α cell fate, while α2 represses a genes to further enforce the α cell haploid fate. The a haploid cell type is the default, and is not actively specified by the MAT locus. This type of mating-type system is called bipolar to reflect the two mating types, a and α. When the two mating types are in balance in the population, bipolar mating systems enable 50% outcrossing and 50% inbreeding.

But other fungi have much more exotic sex lives, and have a more complex mating-type determining system in which there are literally thousands and thousands and thousands of different mating types (Brown and Casselton, 2001; Casselton, 2002, 2008; Heitman et al., 2007; Raper, 1966). In these species there are two loci that lie unlinked on different chromosomes that specify mating type. These are called the A and B MAT loci, and one encodes the homeodomain factors, and the other encodes pheromones and pheromone receptors (which locus encodes which genes depends on the species, because these loci were named historically as they were discovered genetically long before their molecular basis was elucidated). In many species, both loci are multiallelic, and as a result there are many different mating types. Both loci must differ for productive mating and thus an isolate of A1B1 mating type can mate with an isolate of A2B2 mating type, but not with isolates that are A1B2 or A2B1. Because there are thousands of different mating types, most encounters between isolates in nature will be fertile, and this drives the frequency of outbreeding to >99%. On the other hand, from any given cross, say A1B1 by A2B2, because two MAT loci segregate independently four different types of progeny are produced (A1B1, A2B2, A1B2, and A2B1) and these systems are therefore called tetrapolar mating systems. Because any given progeny is only interfertile with 25% of its siblings (A1B2 can mate with A2B1 but not with A1B1 or A2B2), the tetrapolar configuration not only promotes outcrossing but also leads to inbreeding depression. It is thought to be these differences in the frequency of outcrossing and inbreeding that provided the evolutionary pressure for transitions between bipolar and tetrapolar mating systems.

Phylogenetic reconstructions across the fungal kingdom support the conclusion that bipolar mating type is an ancestral state and the tetrapolar configuration is a derived state. If we examine species in the Ascomycota and Zygomycota/Mucorales, all species with known mating type systems are bipolar. This includes a myriad of species throughout the Ascomycota including S. cerevisiae (a and α), Schizosaccharomyces pombe (P and M), Candida albicans (a/a and α/α), Neurospora crassa (A and a), and Aspergillus fumigatus (MAT1-1 and MAT1-2) just to name a few (Glass et al., 1988; Hull et al., 2000; Magee and Magee, 2000; O'Gorman et al., 2009). In the Mucorales, we again find just two mating types, typically called P and M for plus (+) and minus (−), in Phycomyces blakesleeanus, Mucor circinelloides, and Rhizopus oryzae (Gryganskyi et al., 2010; Idnurm et al., 2008; Lee et al., 2008). This is in marked contrast to the Basidiomycota branch of the dikarya in which a majority of species have the tetrapolar mating system configuration. But thus far, no species with the tetrapolar mating system have been found outside of the Basidiomycota. Thus, we can root the phylogenetic tree with multiple outgroups that are all bipolar, and conclude that the tetrapolar system is a derived state, possibly with a single origin at the base of the Basidiomycota. In essence, an ancestral system with just one MAT locus encoding homeodomain factors evolved into a system with a second sex determinant on another chromosome encoding pheromones and pheromone receptors. How such a transition might have occurred could have first involved an auto-stimulated pheromone, pheromone receptor gene pair, which underwent genetic drift to form more than one self-activating allele, followed by recombination between the two to generate two self-sterile/cross-fertile gene combinations that became a second MAT locus. Some aspects of this have been modeled, conceptualizing how such an event might have transpired (Fraser et al., 2007).

Interestingly, there are examples of species with bipolar mating type within the Basidiomycota phylum. Are these remnants of the bipolar ancestral state, like the microwave echoes of the Big Bang that formed our universe, or rather derived from the tetrapolar configuration? This question has been addressed in some detail in the Cryptococcus pathogenic species complex (Heitman et al., 2011). All of the pathogenic species and lineages, including C. neoformans var. grubii (serotype A), C. neoformans var. neoformans (serotype D), and C. gattii (VGI, VGII, VGIII, and VGIV) have just two mating types and are bipolar species. Analysis of a series of closely aligned and related species reveals that multiple outgroup species have tetrapolar mating systems. This includes C. amylolentus and C. heveanensis, for which our group discovered extant sexual cycles, cloned and characterized their mating- type loci, and showed definitively that both are tetrapolar (Findley et al., 2012; Metin et al., 2010). Related studies for Kwoniella mangrovensis characterized its mating type locus and clarified that it is tetrapolar (Guerreiro et al., 2013), in contrast to a previous report from others that had concluded it was bipolar (Statzell-Tallman et al., 2008). Taken from this vantage point of multiple tetrapolar outgroup closely aligned species, we can conclude that the bipolar configuration of the pathogenic species complex is a derived state and has a monophyletic origin. E pluribus duo! The longer evolutionary pathway was thus bipolar to tetrapolar and then a return to bipolar. In this view then the bipolar state is both an ancestral state and a derived one.

The transition from tetrapolar to bipolar has occurred several times independently in the Basidiomycota, and other examples include Ustilago hordei and Malassezia species (Bakkeren et al., 2006; Bakkeren and Kronstad, 1994; Gioti et al., 2013; Hsueh and Heitman, 2008; Lee et al., 1999). That the cases that are known involve pathogens of either humans or plants suggests there may be a causal association between restricted outbreeding and enhanced inbreeding and pathogenesis. Perhaps as microbial pathogens become well adapted to their hosts, less rather than more genetic exchange is adaptive. They clearly do conserve the ability to undergo sex, and thus this may ensure their longer term evolutionary success, but restricting genetic exchange may allow them to vary just enough to keep pace with their hosts.

We return then to a more general consideration of MAT loci and several recent poignant examples prove illuminating. Recent studies have defined the MAT loci for two ciliates in the Alveolate supergroup (Tetrahymena and Paramecium), for two representative algae in the Planta supergroup (Chlamydomonas and Volvox), and for the slime mold Dictyostelium in the Amoebozoa supergroup (Bloomfield et al., 2010; Cervantes et al., 2013; Ferris et al., 2010; Singh et al., 2014). While there are fascinating features, such as the three mating types of the slime mold, and the dramatic expansion of MAT loci into true sex chromosomes from Chlamydomonas to Volvox, and the roulette type mating-type switching of Tetrahymena which has seven mating types (one for each day of the week, or seven brides for seven brothers), there is essentially no underlying conserved feature. Taken from the perspective of diverse and plastic sex chromosomes and mating-type loci from at least 5 of the 8 supergroups of the eukaryotes, there seems to be nothing ancestral or static about sex determination.

These considerations lead us to conclude that mating type and sex determination are not ancestral, but rather derived features. If so, then what was the ancestral pattern of sexual reproduction? Put another way, we know that the LECA was sexy but how she sexuality was manifested is the central question. As we will consider in the section that follows, our central premise is that LECA was unisexual.

Modes of sexual reproduction

We now turn our considerations from mechanisms of sexual reproduction to modes of sexual reproduction. Consideration of the sexual nature of the Cryptococcus pathogenic species complex provides further illumination. We have discussed how they have undergone a transition from an enhanced outcrossing/restricted inbreeding ancestral tetrapolar state with thousands of mating types to a derived bipolar state with just two mating types with enhanced inbreeding and restricted outbreeding. We will now consider how they have taken this transition one step further from a bipolar to a unipolar sexual state with just one mating type that reproduces unisexually, dispensing with the need for a partner of opposite mating type. E pluribus unum!

A central conundrum in the field was that while there was a defined bipolar mating-type system with two mating types that could occur under lab conditions (Kwon-Chung, 1976a), the vast majority of clinical and environmental isolates were all of just one mating type, α. In many populations no isolates of a mating type could be identified. This led to the conclusion that the organism might be largely clonal and asexual. Similar reasoning was applied to the vast majority of fungal pathogens and eukaryotic parasite pathogens of humans, and as recently as a decade ago it was thought that the majority were asexual and clonal. We now appreciate that they are sexual, but in unusual ways, including unisexual and parasexual (Heitman, 2006, 2010).

The discovery of unisexual reproduction of Cryptococcus came from revisiting older observations (Erke, 1976; Wickes et al., 1996) about the production of hyphae, basidia, and spores by certain isolates cultured solo on a variety of media that support mating (V8, MS, SLAD, FA). Several lines of evidence converged to show that this developmental process, which had been called haploid or monokaryotic fruiting and was thought to be strictly asexual, is in fact an unusual form of self-fertile homothallic sexual reproduction (Feretzaki and Heitman, 2013a, b; Lin et al., 2005). The process involves ploidy changes which, in some cases, can involve cell-cell fusion, but likely also occur via endoreplication. Key meiotic genes, including SPO11 and DMC1, which encode the proteins that make and repair the DSB DNA breaks that provoke meiotic recombination, are required for this developmental sexual process. Our group has recently reported a detailed analysis of meiotic recombination features occurring during bisexual and unisexual reproduction, further underscoring that core features of sexual reproduction occur during both modes of sexual reproduction (Sun et al., 2014). Thus, under laboratory conditions unisexual reproduction has all of the hallmark features associated with bisexual reproduction, with the exception that it dispenses with the obligate need for a partner of opposite mating type. There are situations, so called ménage à trois matings, in which three partners are present (two α and one a) and a small number of limiting a cells serve as pheromone donors to stimulate the fusion and unisexual reproduction of two α partners present as the majority in the population. In fact in mixed populations, both a-α bisexual and α-α unisexual reproduction can occur concomitantly and the balance depends upon the ratio of the mating types and who encounters whom.

While early reports focused on unisexual reproduction/monokaryotic fruiting of isolates of α mating type (Wickes et al., 1996), we now appreciate that isolates of a mating type can also undergo unisexual reproduction (Hull and Heitman, 2002; Tscharke et al., 2003). Subsequently, by detailed genetic analysis of progeny from a cross of a more and a less fertile parental strain, it was found that unisexual reproduction is a quantitative trait, to which multiple quantitative trait loci (QTL) scattered around the genome contribute (Lin et al., 2006). The MAT locus was found to be the most significant of the QTLs mapped, with the α allele promoting unisexual hyphal development to a greater extent than the a allele (Lin et al., 2006). To put this another way, isolates with the α allele are skewed towards being of higher fecundity compared to isolates with the a allele, but if the other QTLs are those that promote unisexual reproduction they can elicit development in an a mating type background. Why α isolates predominate in nature is not known, and could result from a bottleneck when Cryptococcus neoformans emerged from Africa to become globalized, or may result from enhanced fitness under conditions that have not yet been found.

Population genetic evidence also supports that unisexual reproduction may be a predominant mode of sexual reproduction in nature. This includes detailed analyses of αADα hybrids produced via unisex (Lin et al., 2007), and also αAAα diploids that may be intermediates or products of the unisexual cycle (Lin et al., 2009). Finally a series of studies have revealed evidence of recombination, even in populations that are exclusively of one mating type (Bui et al., 2008; Hiremath et al., 2008; Saul et al., 2008).

The analysis of αAAα diploids reveals two distinct types. One class is clearly the progeny of two genetically different parents, and thus appears to have been generated via α–α cell-cell fusion. The second class has two seemingly identical genomes, and thus may have arisen from either endoreplication or cell-cell fusion between a mother and her daughter. Our group is currently identifying and studying genes involved in cell-cell fusion and nuclear fusion to test these and other models.

The fact that some unisexual reproduction appears to involve two genetically identical genomes or isolates challenges conventional models for the evolution and impact of sexual reproduction. Put another way, if there is no pre-existing genetic diversity to admix, why go to the trouble of expending so much energy undergoing sexual reproduction? Are these just teenagers spinning their wheels in a parking lot, or is there some potential benefit? It is important to consider that the null hypothesis may be that this mode of sexual reproduction is not of benefit, rather that it may have little cost and therefore is just tolerated (Lynch, 2007). But this said, our group has been exploring experimentally if and how it might confer some type of evolutionary benefit, and have published three studies thus far supporting the contention that it may.

First, our group has presented evidence that unisexual reproduction allows a transition from yeast to hyphae and thereby promotes more efficient foraging for nutrients and also the generation of spores that can be disseminated to more distant locales by wind currents (Phadke et al., 2013). Second, our research has shown that unisexual reproduction can generate genetic diversity de novo, and that much of this diversity results from aneuploidy as a consequence of meiosis, and that this occurs during both unisexual and bisexual reproduction (Ni et al., 2013). Third, our studies recently reported indicate that unisex has the capacity to turn back the hands of time and reverse Muller's ratchet and prevent the otherwise inexorable accumulation of deleterious mutations (Roach and Heitman, 2014), the process that is thought to be that which dooms asexual species to inevitable extinction. There are a variety of possible benefits that are in the process of being tested experimentally, some of which have been reviewed (Roach et al., 2014).

Things that are found once in biology are interesting, but those that are found more than once are more interesting because they may be generalizable. Thus, the discovery that C. albicans has the capacity to undergo both bisexual mating and unisexual mating as a prelude to its parasexual cycle was an important advance (Alby et al., 2009). Thus, we now know that two of the most common systemic human fungal pathogens have extant sexual/parasexual cycles and the capacity for unisexual reproduction. Interestingly, there are a group of four Neurospora species that have extant homothallic (self-fertile) sexual cycles but only one mating type, and thus likely another independent origin and example of unisexual reproduction (Gioti et al., 2012; Glass et al., 1990; Glass and Smith, 1994). Similarly, a group of Stemphylium isolates representing a novel self-fertile group of isolates with only the MAT1-1 mating type information at the MAT locus and no isolates with the MAT1-2 information represents another independent example and origin of unisexual reproduction (Inderbitzin et al., 2005). A very recent study reports that Pneumocystis species, which are obligate fungal pathogens that infect and only live in animal lungs, have a fusion of the two opposite mating types related to the S. pombe MAT loci, suggesting that Pneumocystis is homothallic/self-fertile and may be undergoing its sexual cycle in the lungs of infected animals (Almeida et al., 2015). This is not strictly a case of unisexual reproduction because both MAT loci are present in a fused configuration rather than just one MAT locus, but it does serve to further illustrate how commonly pathogenic microbes have undergone the transition from outcrossing heterothallic sexual reproduction to homothallic self-fertile inbreeding. It is important to note that these types of fused mating-type loci leading to self-fertile homothallic fungal species had been described more than a decade earlier by Gillian Turgeon and colleagues in now classic studies of plant fungal pathogens, including Cochliobolus sp. and Gibberella sp., illustrating convergent evolution to homothallism in diverse fungal lineages (Yun et al., 2000; Yun et al., 1999).

This review is dedicated to considering sexual reproduction in the fungal kingdom, so we will just note in passing that it is also the case that multiple eukaryotic parasitic pathogens including Toxoplasma, Giardia, Trypanosomes, and Leishmania appear to undergo selfing forms of sexual reproduction (Heitman, 2010; Poxleitner et al., 2008; Wendte et al., 2010). Nothing is as yet known about whether they have mating types and if so how these are configured, but it does serve to illustrate common features shared between fungal and parasite eukaryotic pathogens.

I wish to return then to our question of what the sexuality of the LECA ancestor of all eukaryotes looked like and propose that there was an evolutionary epoche in which there was sex before sexes and in which unisexual reproduction was the mode of sexual reproduction. In this view, sexes and mating types were then added later, superimposed on an earlier ancestral system that was unisexual. In this model, the unisexual state is then both an ancestral state and a derived one. And in this view, the pathogenic microbes have not invented an entirely new mode of sexual reproduction out of whole cloth. Rather they have returned to an ancestral model of reproduction in an evolutionary pathway involving transitions from unisexual to bisexual and back again to unisexual.

It seems fitting to close this section with a quote:

“We shall not cease from exploration

And the end of all our exploring

Will be to arrive where we started

And know the place for the first time.”

T.S. Eliot, from “Little Gidding”, Four Quartets

What questions remain to be addressed by studies of fungal sex?

I. Asexuality?

Given that there are an estimated 5 million species of fungi, and maybe more, much remains to be learned about their sexual nature (Blackwell 2011). Many fungi are currently classified as asexual, and yet many fungi long thought to be asexual have been subsequently found to be sexual, or in some cases at least parasexual (i.e. C. albicans). Are there any truly asexual fungi? Candida glabrata is one potential candidate, but also a mystery in that all known isolates are haploids of one or the other of two opposite mating types and much of the machinery associated with mating and meiosis are present in the genome (Muller et al., 2008; Ramirez-Zavaleta et al., 2010; Srikantha et al., 2003; Wong et al., 2003). There is even evidence for recombination in the population. Despite prodigious effort, no extant sexual cycle has as yet been found, but it seems likely that some clever genetic trick, or specific environmental condition, or combination of fertile strains, remains to be discovered.

If a fungus were asexual, how would we recognize it? It is often posited that when a pathway is lost that the genes associated with it rapidly decay into pseudogenes and then are lost. And yet recent studies from Greg Lang and David Botstein reveal that sterile variants of S. cerevisiae rapidly arise during asexual vegetative growth (Lang et al., 2011; Lang et al., 2009). They discovered that 23 genes activated by the mating pheromone response pathway are expressed at some level even during asexual growth. Mutants that inactivate any step in the pheromone response pathway grow ~2% faster than their compatriots because they no longer express these 23 messages and the encoded proteins. And relieved of this gene expression deadweight, they are more fit as asexual competitors. Sequence of their genome reveals a variety of simple mutations in many steps in the pheromone response pathway. But the pathway has not yet had time to decay away, and once the first mutation has occurred to relieve the burden of futile gene expression, there is very little additional benefit conferred by loss of other genes in the pathway. Thus after a very quick coup de grâce, the decline of the rest of the pathway is likely to be anything but swift. Put another way, the rate of decay of the pathway is not constant, and there is a very rapid first step followed by a series of much slower steps. This is interesting for several reasons. First, it appears to be very easy and even beneficial to evolve asexuality, at least for some clones in a population. Second, genome inspection of an unknown species may lull us into complacency when we find the mating and meiotic genes seem to be largely intact; it seems likely that a single nucleotide variant (SNV) in one gene in either pathway might be easily missed. Finally, evolutionary paths are often thought to be unidirectional. Losing sex and becoming asexual is thought to be irreversible. But if the loss of sex is due to one SNV in one gene, this is likely to be fully reversible given sufficient selective pressure for mating or recombination.

Other ways in which an asexual fungal species might manifest could involve gross rearrangements of the genome. Regular repeated rounds of meiosis serve to cull from the population those members whose genomes have strayed from the canonical karyotype. But if a species were to be asexual, one might find a highly variable karyotype from isolate to isolate in the population. And examples of just this pattern have been observed in Verticillium species and suggested to reflect asexuality (de Jonge et al., 2013).

What other asexual hallmarks might we search for? It has been hypothesized that to survive a transition to asexuality, species might need to have shed their entire transposon load in order to survive. So searching for fungal species whose genomes lack transposons might be a starting place. Alternatively, we might explore for fungi that lack one or more key meiotic genes, such as SPO11 or other genes in the meiotic toolkit (Schurko and Logsdon, 2008). Thus far, the only known examples of eukaryotes lacking SPO11 are the Dictyostelium species. In those cases, meiosis and meiotic recombination are known to occur and it seems plausible and even likely that some other enzyme has stepped into the shoes of SPO11 and serves to produce the lesions that provoke meiotic recombination (Goodenough and Heitman, 2014).

II. Why two sexes?

Why are there most commonly two sexes? We do not know the answer to this seemingly simple question, but hypotheses abound (Haag, 2007; Schaffer 2007). 1) It may be to increase the efficiency of mate recognition and fusion. 2) It may serve to restrict outbreeding compared to populations or species in which there is one universal mating type that can mate with anyone. 3) Having two sexes may be also to restrict inbreeding compared to one universal mater that can mate with itself as well as with anyone else. 4) Having two mating types or sexes may have evolved as a response to conflicts between nuclear and mitochondrial genomes, or between mitochondrial genomes, and thereby given rise to uniparental inheritance of mitochondria and chloroplasts. 5) Having two defined mating types or sexes may have enabled specialization of gametes and differential contribution of resources, giving rise to certain development outcomes that we observe such as the transition from isogametic species to anisogametic ones. 6) Two sexes might have served to enable the development of sex specific and sex antagonistic traits. 7) Finally, having two mating types or sexes may serve to enable organisms to know when they are diploid, as they will be heterozygous rather than homozygous for sex determinants contributed by each haploid parent. These possible functions are also not mutually exclusive, either with each other or with other possible functions not considered here.

In the context of considering why there are most commonly two sexes, it is worth remembering that there are examples in which there are more than two mating types or sexes. These include the basidiomycete fungi we have discussed that have literally thousands and thousands and thousands of different mating types (multisexual) including Coprinopsis cinerea, the slime mold Dictyostelium with its three mating types (trisexual), and the ciliate Tetrahymena with seven (septasexual). These may seem exceptions, but considering what is shared and what is distinct between closely aligned species with two sexes vs. those with more, or those with only one (unisexual, to mean “uni” as in one sex, or “universal” as in pansexual and capable of mating with any other member of the population as well as itself), may prove enlightening.

III. Why obligate sexuality?

Many animals are obligately sexual. Why? This is in marked and striking contrast to the vast majority of microbial eukaryotes which are facultatively sexual, and happily capable of vegetative mitotic clonal reproduction interspersed with bouts of sexual reproduction. Why this is not so for humans, mice, rabbits, is an interesting conundrum. At the root it suggests that recombinant progeny may be always better than the parental. If so, what selective force might have driven the transition from a facultatively sexual last common ancestor to a derived obligately sexual state? One driving force may have been a need to outrun pathogens, either exogenous (bacteria, viruses, parasites, fungi, prions) or endogenous ones (transposons).

Another way to frame this key evolutionary transition and question about the facultative to obligately sexual state is to consider a simple question? Are there any fungi who are obligately sexual? One might make the case that some fungi such as basidiomycetous mushrooms spend a significant fraction of their lifecycle as a dikaryotic mycelium and fruiting bodies, and these might be examples of either very frequently sexual or even obligately sexual fungi.

One interesting species that our group has been studying as one of the outgroups to the Cryptococcus pathogenic species complex is Filobasidiella depauperata (Rodriguez-Carres et al., 2010). It is one of three species that are most closely aligned with the pathogenic species (Findley et al., 2009), but it itself is not a pathogen of animals, but it may be a mycoparasite (a parasite of other fungi). This is a particularly fascinating species. It grows remarkably slowly, is easily contaminated, and is quite heat intolerant and must be cultured at temperatures of 24°C or lower. But what is fascinating is that this species has no yeast form, and it is strictly hyphal. Microscopic examination of its hyphae reveals that it shares features with the hyphae produced during the sexual cycle of Cryptococcus, and the hyphae are decorated with basidia and abundant, long, glorious chains of spores. Thus, it appears that F. depauperata may be a close relative of Cryptococcus that has become locked in a permanent sexual state. If so, it may be a case of an obligately sexual fungus. That there are only two isolates known, and their growth is limited and restricted under lab conditions, does not provide a ready explanation for why an obligate sexual state would be more beneficial than a facultative one, but given further study, explanations may emerge. It is at least one vantage point from which to begin to address this question.

Coda

Some things are new, and others old, and sometimes what is new is also old.

Figure 1.

Unisexual reproduction as an ancestral and derived trait. A phylogenetic depiction of the major supergroups of the Eukaryotic tree of life depicting the LECA as sexual/unisexual in the ancestral state, with a diverse pattern of derived sexual states including predominantly bisexual extant species but also those that are unisexual, trisexual, septasexual, and multisexual. Modified from (Baldauf, 2003) with permission of the author.