Evolution of sex-determination in dioecious plants: From active Y to X/A balance?

Sex chromosomes in plants have been known for a century, but only recently have we begun to understand the mechanisms behind sex determination in dioecious plants. Here, we discuss evolution of sex determination, focusing on Silene latifolia , where evolution of separate sexes is consistent with the classic “two mutations” model—a loss of function male sterility mutation and a gain of function gynoecium suppression mutation, which turned an ancestral hermaphroditic population into separate males and females.Interestingly,thegynoeciumsuppressionfunctionin S.latifolia evolvedvialoss offunctioninatleasttwosex-linkedgenesandworksviagenedosagebalancebetween sex-linked, and autosomal genes. This system resembles X/A-ratio-based sex determination systems in Drosophila and Rumex , and could represent a steppingstone in the evolution of X/A-ratio-based sex determination from an active Y system.


INTRODUCTION
[3] S. latifolia became a de facto model system for studies of sex chromosomes in plants, perhaps due to large and strongly heteromorphic sex chromosomes, making their identification easier under the microscope.Early work in this species [4] has inspired the development of the classical "two mutations" (aka "two genes" or "two-factor") model stating that dioecy evolves via the gynodioecy pathway, in which a mutation responsible for the abortion of stamen or pollen maturation followed by another mutation causing carpel suppression. [5,6]This model has been favored by many plant sex chromosome researchers, probably because it naturally explains the evolution of dioecy and sex chromosomes at the same time.
The alternative "monoecy pathway" model states that dioecy evolves from monoecy via the evolution of a single sex-determining This is an open access article under the terms of the Creative Commons Attribution License, which permits use, distribution and reproduction in any medium, provided the original work is properly cited.© 2023 The Authors.BioEssays published by Wiley Periodicals LLC.
9][10] Single-factor sex-determination systems were indeed found in some dioecious species, such as Diospyros lotus [11] and Poplus spp. [12,13]However, it is worth noting that even in single-factor sexdetermining systems more than one mutation may be necessary for the emergence of dioecy; for example, the evolution of both sexdetermining genes, MeGI and OGI would be required for dioecy in

D. lotus.
Early genetic work in S. latifolia has revealed the presence of two sex-determining genes on the Y-chromosome, the gynoecium suppressing factor (GSF) and stamen promoting factor (SPF). [4] The deletion of GSF or SPF sex-determining genes leads to the development of hermaphroditic or asexual flowers, respectively, [14,15] which is consistent with the "two-factor" model for sex determination and sex chromosome evolution. [5,6]Similarly, in asparagus and kiwifruit sex determination is controlled by two genes.In kiwifruit, Shy Girl (SyGl) and Friendly Boy (FrBy) act as the pistil-suppressing and stamenpromoting genes, respectively. [16,17]

FEMALE FUNCTION (SOFF) and DEFECTIVE IN TAPETAL DEVELOPMENT
[20] However, the sexdetermining region on the Y-chromosome is hemizygous so that the X chromosome has no gametologs of these sex determining genes, [18] making it difficult to reconstruct the origin of sex-determining genes during sex chromosome evolution.Identification of sex-determining genes in S. latifolia [21] offers an opportunity to reveal how the sexdetermination and large heteromorphic sex chromosomes evolve from scratch.

EVOLUTION OF GYNOECIUM SUPPRESSION IN S. LATIFOLIA VIA TWO GENE LOSSES
Detailed deletion mapping of the S. latifolia Y-chromosome, [22,23] combined with genomic sequencing of the hermaphroditic mutants and functional molecular genetic analyses, revealed the most likely candidate for the GSF gene. [21]This gene is homologous to the Arabidopsis CLAVATA3 (CLV3) and it has functional Y-linked (GSFY) and dysfunctional X-linked (GSFX) gametologs that diverged around the time when S. latifolia sex chromosomes originated. [21]CLV3 gene regulates the size of the shoot apical meristems (SAMs) and flower bud primordia in Arabidopsis. [24]The clv-3 mutants in Arabidopsis often show enlargement of carpels, while overexpression of CLV3 leads to pistilless flowers. [24]These phenotypes are similar to those of female and male flowers of S. latifolia, respectively (Figure 1A), suggesting that this gene is a likely candidate for GSF.This was also confirmed by transgenic analyses (Figure 1B) and bioassays in which A. thaliana and S. latifolia were treated with synthetic peptides derived from GSFX/Y, CLV3, and the mutant allele of CLV3 (CLV3m). [21]sed on the available evidence, the GSF function in S. latifolia likely works via the WUSCHEL-CLAVATA (WUS-CLV) feedback loop [21] where the CLV3 mRNA production is activated by WUS protein, while the expression of WUS is repressed by CLV3.[27] The S. latifolia gene encoding a counterpart of CLV3 in this feedback loop, SlWUS1, is linked to the X chromosomes, but absent from the Y chromosome. [28]In Arabidopsis, wus mutants show the loss of pistil [29] and in S. latifolia the X-linked SlWUS1 appears to control the size of gynoecium.This was supported by the observation of smaller gynoecium size and the smaller numbers of seeds in plants with fewer copies of SlWUS1. [30,31]A possible effect of the copy number of SlWUS1 on gynoecium size can also be found in the classic analyses of colchicinetreated polyploid plants. [32]As the ratio of the X chromosomes to the Y chromosome increased, hermaphroditic flowers were produced more frequently and in the case of XXXXY plant, almost all flowers were hermaphrodites.It is interesting that SlWUS1 is not dosage compensated (i.e., not upregulated in males) [28] despite the location in the oldest stratum 1 [33] where we may expect dosage compensation to be prevalent, particularly given the recent models predicting that dosage compensation is responsible for evolution of X-Y recombination repression. [34,35]This likely reflects the dosage sensitivity of SlWUS1 that should not be altered by evolving dosage compensation system.
Taken together, the S. latifolia X chromosome is apparently involved in sex determination with X/Y dosage balance mechanism, that likely operates via the WUS/CLV feedback loop. [21]cording to the canonical two-factor model, the transition from ancestral hermaphroditism to dioecy via gynodioecy involves a recessive loss-of-function male sterility mutation on the proto-X and a dominant gain-of-function female sterility mutation on the proto-Y-chromosome. [5]Thus, the finding of an X-linked loss-of-function mutation in GSFX [21] may seem surprising.However, the presence of functional GSFY on the Y-chromosome effectively plays the role of a dominant gain-of-function Y-linked female sterility gene that is predicted by the canonical two-factor model.This scenario suggests that F I G U R E 2 Gynoecium suppression mechanism in S. latifolia.(A) Ideogram for the evolution of gynoecium suppression in S. latifolia.The protoX/Y chromosomes would have had ancestral GSF and SlWUS1 genes.GSF was dysfunctionalized on the X, leaving the active GSFY gene on the Y chromosome, while SlWUS1 disappeared from the Y chromosome, but remains functional on the X-chromosome.(B) Correlation between WUS/CLV3 ratio and sexual phenotypes.The WUS/CLV3 ratio well explains sexual phenotypes of S. latifolia; 2 is female, 1.5 is hermaphrodite, and 1 is male.This ratio threshold has likely evolved upwards from the original hermaphroditic state seen in S. vulgaris and A. thaliana.not just two but at least three mutations, one for maleness (not yet identified) and two for femaleness (dysfunctionalizing GSFX on the X and SlWUS1 on the Y) might have been involved in the transition from hermaphroditism to full dioecy in S. latifolia.The evolution of female suppression via more than one mutation is not surprising, given the loss of female function is costly in terms of fitness and must be offset by at least a two-fold gain in reproductive success via male function. [36]us, splitting this costly change into several steps, each of which reduces female function, may be a more realistic way to complete the shift from gynodioecy to dioecy.

DOES THE WUS/CLV3 BALANCE REPRESENT A STEPPINGSTONE IN EVOLUTION OF X/A-RATIO-BASED SEX DETERMINATION?
If gynoecium suppression in S. latifolia is controlled via the WUS/CLV feedback loop, [21] we would expect the ratio of WUS and CLV3 copy numbers in a plant to predict the size of its gynoecium.To test this expectation, we also have to take into account the presence of expressed autosomal paralogs (Kazama et al., [28] and unpublished) of these genes in S. latifolia genome-SlWUS2 and SlCLV3.Consistent with the above expectation, the copy number difference of WUS and CLV3 genes between males and females explains the sex of S. latifolia flowers well (Figure 2B).In male, female, and hermaphroditic mutant plants, the WUS/CLV3 ratios are 1, 2, and 1.5, respectively.These ratios are consistent with chromosome numbers in the synthetic polyploid S. latifolia plants. [32]The XXXY plants (with WUS/CLV3 ratio = 1.4) are males with occasional hermaphroditic flowers, while the XXXXY plants (WUS/CLV3 = 1.6) are hermaphrodites with occasional male flowers (Figure 2B).Therefore, the WUS/CLV3 ratio appears to explain the mechanism of gynoecium suppressing functions in S. latifolia.Furthermore, it was reported that treatment of males with hypomethylating chemical 5-azacytidine results in occasional development of hermaphroditic flowers, either due to suppression of a Y-linked female suppressor gene, or due to activation of autosomal femalepromoting gene, [37] which can be explained by the gene balance model in the WUS-CLV feedback loop controlling the development of female organs with X-and autosomal-linked WUS genes in S. latifolia flowers (Figure 2B).
In the case of hermaphroditic species, such as S. vulgaris and A. thaliana, the WUS/CLV3 ratio is 1, indicating that the threshold for gynoecium suppression in the WUS/CLV3 ratio in the dioecious S. latifolia has evolved during dioecy evolution in S. latifolia ancestor.
It is interesting to speculate that this system could evolve further to eventually become an X/A-ratio system when the Y chromosome degenerates and the Y-linked GSFY gene is lost or moved to an autosome.Such X/A-ratio sex determination is observed in R. acetosa [38,39] and Drosophila melanogaster. [40,41]Interestingly, genus Rumex includes dioecious species with the X/A-ratio as well as with the active-Y sexdetermining mechanisms, [42] which is consistent with the suggestion of Westergaard that the X/A-ratio systems evolve secondarily from the active-Y system. [5]e molecular bases of X/A-ratio-based sex determination in D. melanogaster are well-studied. [43]Sex in D. melanogaster is controlled by the ratio of the products of X-linked "numerator" (sisA, sisB, sisC, and runt) and autosomal "denominator" (deadpan, Dpn) genes in early development. [43]The products of the numerator genes, collectively called X-linked signal elements (XSE), are transcription factors that activate expression of the key sex-determining gene sexlethal (Sxl), while the denominator Dpn acts as a repressor of Slx.The ratio of Xlinked nominators and autosomal denominator proteins is sufficient to activate Sxl in females (XX: AA), but not in males (X:AA). [43]Once activated, the expression of Sxl is self-maintained throughout the female body and it controls the splicing of downstream genes in sex determination pathway.The functional SXL in females plays a role in splicing a splicing regulator transformer (tra) into a functional isoform, whereas lack of the functional SXL in males leads to nonfunctional splicing of tra, [44,45] resulting in male specific splicing of the well-conserved sexdetermination related gene, doublesex (dsx). [46]The male-specific dsx promotes the development of male morphology.This X:A ratio system is thought to evolve from the ancestral XY system via sex chromosome turnover with recruitment of Slx to the X chromosomes. [47]milar to sex-determination in D. melanogaster, the femalesuppressing function in S. latifolia is controlled by gene dosage balance, with the X-linked positive regulator (SlWUS1), but unlike D. melanogaster, the negative regulator (GSFY) is Y-linked, resulting in an active-Y rather than X/A-ratio system.However, ongoing Ydegeneration [48,49] may lead to eventual GSFY loss and evolution of X/A-ratio system controlled by the number of X-linked WUS genes.
Interestingly, WUS is also involved in the sex expression pathway in kiwifruits, [16] and Cucumis, [50] implying that this mechanism possibly plays a role in gynoecium suppression of many plant groups with independently evolved dioecy and sex chromosomes.Further analyses of sub-functionalization of GSFY or SlWUS1 and the effect of their copy numbers will allow one to test whether WUS/CLV3 ratio control the gynoecium suppression.

CONCLUSIONS
The gene dosage balance-based gynoecium suppression in S. latifolia may be regarded as a steppingstone in the evolution of an X/A balance sex determination system, such as found in Drosophila and Rumex.The tendency of Y-linked genes to undergo genetic degenera-tion may mean that such shifts from an active-Y to X/A balance-based sex determination is a general phenomenon.The mechanism of sex determination in S. latifolia provides an illustration of how an ancestral active-Y sex determination could be modified to turn it into an X/A balance-based system.The eventual loss of GSFY due progressing Ydegeneration in S. latifolia may be compensated by the modulation of the WUS gene expression level, giving rise to the development of the X/A-ratio system.Further investigations into the molecular bases of this system in S. latifolia will provide valuable insights into the dynamic process of sex-chromosome evolution and enhance our understanding of sex-determination mechanisms in plants generally.

F I G U R E 1
Function of GSFY gene in gynoecium suppression.(A) Expression of GSFY at the early stages of flower development causes the suppression of the gynoecium.(B) Phenotypes of Arabidopsis transformants in which GSFY was introduced under the control of the native promoter and the CaMV35S promoter.Bars = 1 mm.
In asparagus, SUPPRESSOR OF