Yeast sexes: mating types do not determine the sexes in Metschnikowia species

Abstract Although filamentous Ascomycetes may produce structures that are interpreted as male and female gametangia, ascomycetous yeasts are generally not considered to possess male and female sexes. In haplontic yeasts of the genus Metschnikowia, the sexual cycle begins with the fusion of two morphologically identical cells of complementary mating types. Soon after conjugation, a protuberance emerges from one of the conjugants, eventually maturing into an ascus. The originating cell can be regarded as an ascus mother cell, hence as female. We tested the hypothesis that the sexes, female or male, are determined by the mating types. There were good reasons to hypothesize further that mating type α cells are male. In a conceptually simple experiment, we observed the early stages of the mating reaction of mating types differentially labeled with fluorescent concanavalin A conjugates. Three large-spored Metschnikowia species, M. amazonensis, M. continentalis, and M. matae, were examined. In all three, the sexes were found to be independent of mating type, cautioning that the two terms should not be used interchangeably.


Introduction
Yeasts ar e gener all y not thought to exhibit sexual dimorphism, where male and female structures might easily be recognized (Phaff et al. 1978 ).Ho w e v er, as with dimor phic species, wher e sexual offspring only arises from male-female pairs, yeast mating is not random.Compatibility in ascomycetous yeasts is controlled by a single locus for which two alleles, termed idiomorphs, determine the activation of genes that code for either the a -or the αmating type-specific pheromones and their receptors (Wolfe and Butler 2022 ).In some filamentous Ascomycetes (Pezizomycotina), it is possible to distinguish male and female structures .T heir differentiation is generally independent of mating types (Bennett and Turgeon 2016 ), although an association between the two has been reported in Neurospor a tetr aspema (Samils et al. 2013 ).In that species, se v er al genes are co-expressed with the mating types.Mating type A strains favour sexual r epr oduction and a female bias, whereas mating type α strains show a tendency for asexual r epr oduction and a male bias.
Yeasts seldom differentiate into identifiable sexes.Exceptionally, McClelland et al. ( 2004 ) regarded the mating types α and a of the basidiomycetous yeast Cryptococcus neoformans as male and female, r espectiv el y, in vie w of the unidir ectional mov ement of α-nuclei to w ar ds the dikary on follo wing conjugation.Heitman et al. ( 2013 ) also reported a tendency of a -cells to increase in size when exposed to the α-pher omone, whic h may allow recognition of the two mating types under the microscope .T he larger mating type a cells r epr esent cognates to the larger female gametes of oogamous species.In addition, consistent with the notion of sexes in C. neoformans is the report of uniparental inheritance of mitochondria by the presumed female (mating type a ) cells (Xu et al. 2000 ).This notwithstanding, it is not common practice to designate yeast gametes as male or female, and some authors have gone as far as to assume that the notion of sexes is irr ele v ant to y easts (Nieuw enhuis and Aanen 2012 ) or e v en to fungi in general (James 2023 ).
As alwa ys , clear definitions are k e y to fruitful discussion.Whether biological sexes are strictly binary as opposed to part of a spectrum is curr entl y the object of considerable debate in some cir cles (Go ymann et al. 2023, McLaughlin et al. 2023 ).The controversy is often fuelled by ambiguous terminology, where one fails to parse out secondary from primary sexual c har acteristics or to distinguish biological sexes fr om sociall y constructed genders.Reports of non-binary sexes in micr oor ganisms, used as e vidence for the non-binary nature of the sexes, are clear instances of equivocation, deliberate or otherwise.Specifically, the supposition that some fungi feature thousands of sexes (Kothe 1996 ), that a cellular slime mould has three sexes (Douglas et al. 2016 ), or that Tetrahymena thermophila is a se v en-sex species (Yan et al. 2024 ) is in e v ery case a patent misuse of "sexes" as a synonym of mating types, perhaps motivated by the wish to enhance the allure of journal article titles.Other examples exist (Iwasa andSasaki 1987 , Hurst 1996 ).Likewise, the suggestion that "the sexual identity of a strain is established by the mating-type locus" (Kim and Bork ovic h 2006 ) is not borne out by the source r efer ence (Kr onstad and Staben 1997 ) and is undoubtedly a case of imprecise use of the expression "sexual identity".Similar criticisms a ppl y to the use of "sex determination" by Ni et al. ( 2011 ) or "sex genes" by Idnurm et al. ( 2008 ).
It is of no help that the term sex is also used as a shorthand for sexual r epr oduction, r egardless of whether it entails recognizable sexes.Equivocal terminology may be a sign of the times, as evidenced by a Google Ngram depicting a two-fold decrease in the use of the w or d "sexes" from its 20 th century maximum in 1977 to the year 2019, in contrast to the w or d "gender", which experienced a 20-fold increase over the same period.
The frequent inclusion of motility in the definition of male gametes reflects a zoocentric bias that is pervasive in biology.Gamete size (anisogamy) is also a widely applied criterion, but similarly should be seen as ov erl y r estrictiv e, as it fails to account for clear cases of male-female polarity in isogamous organisms, as exemplified for Spirogyra species by Saunders ( 1931 ) or Takano et al. ( 2019 ).T here , the label "female" is clearly understood to designate the conjugant that r eceiv es a nucleus donated by her male counterpart.We see the direction of movement of nuclei as a clear and widely applicable means of identifying male and female sexes.Recognition of the male gamete as that which donates its nucleus to a female gamete was explicit for Cryptococcus neoformans (McClelland et al. 2004 ).
Conjugation in some ascomycetous yeasts proceeds between a bud and its mother cell, a form of heterogamy.In those cases, bud formation, coupled with mitosis, culminates with the separation of two cytoplasms by a septum that is later dissolved, allowing the return of the bud nucleus to the mother cell, which then matures into an ascus containing meiotic spores .T his was elegantl y demonstr ated by electr on micr oscopy in Debaryom yces hansenii (Kr eger-v an Rij and Veenhuis 1975 ).The directionality of nuclear transfer from bud to mother cell thus makes the bud the male gamete, and the mother cell, the female gamete .T hese heterogamous buds are typically small, such that the notion of male-female polarity is further reinforced by a size dimorphism.Genetically, mother-bud conjugation implies that a species is homothallic, where all sister (mother cell and bud) nuclei are compatible while having identical genomes.Wolfe and Butler ( 2022 ) explained that homothallism arises either when e v ery cell of a cultur e carries activ e forms of both mating type alleles, or when cells are capable of mating type s witching.T he former, termed primary homothallism, has been documented in se v er al species, including D. hansenii , and so we m ust r egard this as an instance wher e differ entiation of the sexes is ontogenetic and independent of mating types.It is not inconceivable that sexes and mating types might be coupled in secondarily homothallic species, but this remains to be explored.In filamentous Ascomycetes, the formation of male and female structures (e.g.antheridia, ascogonia) appears to be governed by different sets of genes, although interactions between these and mating type genes are evident in some cases (Wilson et al. 2021 ).As already mentioned, some coupling has been noted in Neurospora tetrasperma (Samils et al. 2013 ).
The male-female polarity of bud-mother cell conjugation in Debaryomyces species is analogous to the sexes encountered in complex fungi or, for that matter, animals or plants, in the sense that the two sexes can be identified mor phologicall y by size, the female conjugant being larger.Ho w ever, conjugation in yeasts is often isogamous, pr oceeding between mor phologicall y identical cells, as in the archetypal Saccharomyces cerevisiae (Herskowitz 1988 ).In suc h species, the r ecognition of an individual as male or female is pr oblematic, suc h that one may argue that the concept of sexes simply does not apply.Absence of sexes does not preclude mate c hoice among str ains of compatible mating types.In Sacc harom yces cerevisiae , for example, mate choice is thought to affect evolutionary fitness (Pilpel et al. preprint ).
The yeast genus Metschnikowia contains both diplontic and haplontic species that can engage in isogamous conjugation, prior to the transition from the haplophase to the diplophase.To the extent of the r ele v ant genome data r e vie w ed b y Lee et al. ( 2018 ), all Metschnikowia species are heterothallic, each haploid strain harboring a single mating locus with either an α or an a idiomorph.In diplontic species, whic h ar e by necessity heter ozygous at the mating locus, ascus formation is envir onmentall y trigger ed (Lac hance 2011 ); conjugation of haploid mating types is rarely observed but is presumed to occur in nature soon after ascospore germination.This is the case for the better-known species M. bicuspidata , a pathogen of crustaceans, M. pulcherrima , utilized in winemaking and biological control of fruit spoilage, and M. reukaufii , a major component of the nectar microbiota of many plants (Lachance 2016 ).Conjugation between germinating ascospores is of all e vidence suppr essed in ha plontic Metsc hnikowia species, as they are invariably recovered in nature as haploid cultures, although they r eadil y conjugate when mixed in the laboratory.Like all isogamous yeasts, haplontic Metschnikowia species initiate the sexual cycle by the fusion of mor phologicall y identical cells.Importantl y, they shar e the additional feature that the ascus clearly arises from one of the two conjugating cells , a wa y from the point of fusion (Fig. 1 ).A male-female polarity thus exists, but unlike what is seen in anisogamous species, the two sexes become perceptible only after conjugation has taken place between identical gametes .T he vestigial zygote consists of a clearly male, donor cell and a clearly female, ascus mother cell.Such a male-female polarity is also found in a few other genera, notably in a small clade of Wickerhamiella species typified by W. occidentalis (Lachance et al. 1998 ).Whether mating type alleles determine sexes in these yeasts is not known.Although the mating types of ha plontic Metsc hnikowia species ar e easil y determined by cr ossing a culture with reference strains of known mating types and observing zygote formation, which takes place within a few hours, the mating types of unconjugated cells in a mixed culture are indistinguishable.
Because of the asymmetry of the de v eloping ascus, we hypothesized that male and female sexes in haplontic Metschnikowia species ar e contr olled by the mating type locus.We further conjectured that α-cells would be male and a -cells female, given certain known differences between α and a mating pheromones.In Metschnikowia (Lee et al. 2018 ) and other yeast genera (Rogers et al. 2015 ), mating specificity is governed to a higher degree by the αpher omone, whic h tends to div er ge faster in evolutionary time.It follows that a -cells share a higher proportion of the fitness burden of mate selection, which is typically a female characteristic.Furthermor e, a -pher omones of Ascomycetes differ by the presence of a farnesyl residue that reduces pheromone diffusibility.The farnesyl group does not appear to contribute to mate recognition itself (Gonçalves-Sá and Murray 2011 ), but it reduces the reach of the apheromone, and in so doing, limits the mating response of α-cells to those that are in close physical proximity to a -cells (at least in Sacc harom yces cerevisiae ; Anders et al. 2021 ), thus pr e v enting a runa wa y feedback loop that would engender a r a pid buildup of satur ating concentr ations of pher omones.Asymmetry in mating r esponses is pr er equisite to sexual selection in isogamous species.Such asymmetries have been reported for a small clade of haplontic Metschnikowia species that are endemic to the Hawaiian arc hipela go, wher e they live in association with endemic nitidulid beetles (Lee et al. 2018 ).In those species, for example M. hawaiiensis , hybridization with moder ate r elativ es (e.g.M. continentalis ) only occurs in crosses between α-strains of the former and a -strains of the latter, and not in the r ecipr ocal cr oss .T hese species ha ve F igure 1. Sexual c ycle of lar ge-spor ed, ha plontic Metsc hnikowia species.Ha ploid budding cells, conjugation (1-2 h), diploid zygote (2-3 h), and de v eloping asci (5-72 h) of M. hawaiiensis traced from fluorescence micrographs of DAPI-stained material (see supplementary data Fig.S1 ).The conjugants are identifiable as male and female from the start of ascus development, where they differentiate, respectively, into a moiety devoid of its nucleus and a mother cell that produces a pr otuber ance in which meiosis and ascospor e de v elopment will take place.Germinating spor es tr aced fr om a phase contr ast micr ogr a ph of (unstained) M. continentalis (Marinoni and Lachance 2004 ).Here, the mature ascus, 18 h after transfer to fresh medium, sho w ed cleav a ge of ascospor es into budding cells.identical a -pher omones, but differ ent α-pher omones.In pr epar ation for the experiments to follow, a similar phenomenon was observ ed wher e hybrid (sterile) asci arise in crosses between α-strains of M. amazonensis and a -strains of the other two species, but not in the r ecipr ocal cr osses.
In Sacc harom yces cerevisiae , the differ ences in diffusibility of the two pheromones, combined with the production by a -cells of protease Bar1, which specifically hydrolyses the α pheromone, results in differences in the mating behaviour of the two cell types (Banderas et al. 2016 ).Mating type a cells respond primarily to the ratio of αover a -cells, whereas α-cells respond to the absolute density of a -cells.Additionally, Anders et al. ( 2021 ) identified two behaviour al differ ences between the two cell types, namely, budding orientation and pr efer ential conjugation tube de v elopment in α-cells .T he latter might be construed as a male feature, although the authors r efr ained fr om doing so.By analogy to the formation of trichogynes by females of filamentous Ascomycetes (Kim and Bork ovic h 2006 ), caution is called for in assigning a sex polarity to conjugation tube de v elopment.
The polarity observed in Cryptococcus neoformans (McClelland et al. 2004 ), although creating a precedent for an association between mating types and sexes, does not entir el y buttr ess our α-male/ a -female hypothesis, as both pheromones of that species ar e farnesylated.Furthermor e, the lac k of detectable sequence identity observed in the peptide portion of ascomycete and basidiomycete pheromones casts doubt on whether they should be regarded as orthologues.Finally, at variance with what was observed in C. neoformans , an association between mating type and mitochondrial inheritance has not been observed in Metschnikowia species, including M. continentalis (Marinoni and Lachance 2004 ).
Alternatives to the α-male/ a -female hypothesis in Metschnikowia species are (1) that α-cells are female or (2) that the sexes arise independently of mating types.An adaptationist explanation for either outcome would be more difficult to form ulate.An y association between sexes and mating types, regar dless of polarity, w ould ho w e v er begin to shed light on the so-called non-sex genes found in the two mating type alleles of all species of the Metsc hnik owiaceae (Riley et al. 2016, Lee et al. 2018 ).In that family, mating loci contain, in addition to the mating genes mat a 1 and mat a 2 , or mat α1 , three genes that do not appear to bear on mating compatibility.Two of these non-collinear genes, PIK1 (phosphatidylinositol kinase) and PAP1 (pol y(A) pol ymer ase) are essential in several species including Candida albicans (like Metschnikowia , a member of the order Serinales, but in the stem famil y Debaryomycetaceae; Gr oene wald et al. 2023), but deletion of a third gene, OBP1 (oxysterol binding protein), is not lethal (Srikantha et al. 2012 ).The genes are suspected to play a role in infectivity of C. albicans by affecting the permeability of biofilms and their susceptibility to antibiotics such as fluconazole, but the mechanism is not known.
Non-sex genes have been more appropriately named nonmating genes in a discussion of Candida auris (Muñoz et al. 2018 ), a member of the family Metschnikowiaceae.Were it demonstrated that PIK1 , PAP1 , and OBP1 do indeed determine sex polarity, the expression "sex-determining genes" would indeed be a more accur ate c har acterisation, emphasizing their distinct but complementary nature with respect to the mat a 1 , mat a 2 , and mat α1 transcription factor-encoding genes known to regulate the expression of mating genes.
To shed light on these matters, we labeled mating types of three Metschnikowia species with fluorescent concanavalin A (conA) conjugates and observed their arrangement during mating.Our hypotheses allow three mutually exclusive predictions: (1) that all cells identified as male will be of mating type α, (2) that all females will have mating type α, or (3) that the joint distribution of sexes and mating types will be random (50:50).

Yeast cultures
The strains were obtained from the yeast culture collection of the Department of Biology (formerly Plant Sciences) of the University of Western Ontario (UWOPS), where they are k e pt frozen in liquid nitrogen.They were selected to r epr esent the two mating types of three closely or moderately related species for which draft genome sequences have been deposited.

Mating and microscopy
Mating experiments were prepared from cultures grown on YM agar (Glucose 1%, peptone 0.5%, malt extract 0.3%, yeast extr act 0.3%, a gar 2%).Similar r esults wer e obtained fr om ov ernight cultures k e pt at 25 • C or with up to tw o-w eek-old cultures k e pt at 15 • C. For DAPI staining, crosses were performed by mixing small, equal amounts of compatible cells on corn meal agar (Difco, Detroit, MI, USA).Mixtures were observed periodically to identify the various stages of ascogenesis, from zygote formation (1-3 h), to ascus de v elopment (4-6 h) to maturation of the two ascospores (1-3 d).Suitably diluted, formaldehyde-fixed cells wer e a pplied to an a gar bloc k and tr ansferr ed to a conAcoated coverslip, following the procedure of Streiblová and Hašek ( 1996 ).They were then stained with 2 μg.mL −1 4',6-diamidino-2-phenylindole (DAPI, Sigma Aldrich, St. Louis, MO, USA) as described by Balasubramanian et al. ( 1997 ) and observed under a Zeiss Axioimager A.1 fluorescence microscope with the DAPI filter cube (358/463 nm) and a 100x objective .T he images were taken in both fluorescence and bright field modes with a high-resolution monoc hr ome XCD-X700 CCD camer a (Son y) using Northern Eclipse 8.0 software from Empix Imaging Inc. (Mississauga, ON, Canada).
Mating type-specific staining follo w ed a modification of the pr otocol of Loc khart et al. ( 2003 ).Cells of each mating type were suspended in sterile water, centrifuged, and resuspended in 0.5 mg.mL −1 Alexa Fluor 488-conjugated conA (green/c y an) or Alexa Fluor 647-conjugated conA (r ed/ma genta), both obtained from T hermoFisher Scientific , London, ON, Canada, and dissolved in 0.1 M NaHCO 3 .These pr ocedur es wer e carried out in the dark with a red headlamp.After 1 h at room temperature, the cells were twice centrifuged and resuspended in w ater.Tw enty micr olitr es of the stained suspensions were mixed and inoculated onto Yeast Carbon Base (Difco) agar with 0.01% yeast extract (YCBY) and incubated at room temperature in the dark.Observations took place 5 to 7 h after mixing.Agar blocks carrying the mating mixtures wer e tr ansferr ed, cells down, onto co verslips , in a drop of Citifluor antifadent mountant (Electr on Micr oscopy Science) containing 100 μM calcofluor (Sigma).The pr epar ations wer e observ ed in a Nikon Eclipse Ti2-E microscope (Nikon Canada Inc., Mississauga, ON, Canada) with a 60x oil immersion objective and a 1.5x intermediate magnification, at the Microscopy Centre of the Biotr on, Univ ersity of Western Ontario.For eac h ima ge, a focal series was recorded at each appropriate excitation/emission wavelength for the three fluorophores (calcofluor: 378/432 nm; Alexa 488: 474/515 nm; Alexa 647: 635/680 nm).Images were deconvolved using NIS Elements imaging software (Nikon) version 5.42.03 and extended depth of focus images were created to yield two-dimensional images with all parts of the cells in focus .T he thr ee fluor escent colours wer e optimized manuall y to giv e the clear est inter pr etation of the location of the two conjugants relative to nascent asci.Although the observation of two different conjugation figures would be sufficient to reject hypotheses 1 or 2, the minimum of 21 conjugation figures observed for each species ensures a good fit to the binomial distribution, were hypothesis 3 to pr e v ail.Recipr ocal experiments with both combinations of Alexa Fluor conA conjugates and mating types were conducted to rule out an y differ ential effects of the dyes on the mating outcome.

Results
Figure 2 shows examples of the ima ges r ecorded for the three species examined.Single cells or members of conjugated pairs in the zygotes clearly stained green or red.Fluorescent labeling of either mating type with either Alexa Fluor gav e equiv alent r esults.De v eloping asci, whic h ar ose after mating type-specific staining, acquired the blue fluorescence characteristic of calcofluor .Importantly, in the three species examined, asci grew out of either mating type in zygotes (Table 1 ).The numbers lie within the 95% confidence limits of a 50:50 binomial distribution, the largest spread, seen in M. amazonensis , sitting at 1.95 standard de viations fr om the expected mean of 0.5.

Discussion
Inter pr etation of the numbers in Table 1 is straightforw ar d.Were mating loci responsible for determining sex polarity, one would hav e obtained exclusiv el y zygotes wher e all male conjugants were of the same mating type, be it α or a .Ho w e v er a ppealing, the hypothesis of interdependence of the sexes and the mating types m ust be r ejected.Instead, the r esults support the notion that alleles of the mating locus have no direct bearing on which half of the zygote gives rise to the ascus.It follows that sexual polarity is independent, not only of the mating genes proper, but also of the socalled non-sex genes found in the mating loci of metsc hnik owiaceous yeasts, such that their renaming to sex-determining genes is not justified, although the term non-mating genes remains mor e a ppr opriate.Sex determination acr oss the living world is extr emel y div erse, but a lar ge pr oportion of the r esearc h on this topic is focused on animals or plants (Tree of Sex Consortium 2014 ).The experiment described here, ho w ever simple , pro vides an important element to w ar ds our understanding of sex determination in yeasts, namely that sex polarity can be the result of the chance movement of the nascent diploid nucleus to one or the other parent cell, regardless of mating type.Our observations were limited to species of a single genus.It will be worthwhile to examine other cases where ascogenesis is asymmetrical, for example species of the distantl y r elated genus Wic kerhamiella , to obtain a better sense of gener ality.Importantl y, the r esult highlights the need to exercise utter care in the choice of terminology surrounding sex, the sexes, sexual identity, sex determination, sex genes, or non-sex genes in discussions of these and other related topics such as mating compatibility or gender.

Figure 2 .
Figure 2. Random joint distribution of mating types and sexes in three species of Metschnikowia , M. amazonensis (A,B), M. continentalis (C), and M. matae (D).Alexa Fluor-labeled mating types ( α = r ed/ma genta, A,C or green/c y an B,D).Ascus mother cells (female) are those proximal to the developing asci (blue fluorescence of calcofluor) and the conjugated cells distal of the nascent asci are regarded as male.A complete set of images can be found in the supplementary data as Figs S2 to S5 .

Table 1 .
Female and male sexes are independent of mating types .Strains of Metschnikowia species and r esults of cr osses .T he number of zygotes where mating types a or α became the female or males sexes is shown.Raw data can be found in the supplementary data, Figs S2 to S5 .Superscripts identify type strains and allotypes , as applicable .UFMG are accession numbers in the culture collection of the Univ ersidade Feder al de Minas Ger ais , Belo Horizonte , MG, Brazil.