Plant Sex Determination PDF

Current Biology Minireview Plant Sex Determination John R. Pannell Department of Ecology and Evolution, University of Lausanne, 1015 Lausanne, Switzerland Correspondence: [email protected] http://dx.doi.org/10.1016/j.cub.2017.01.052 Sex determination is as important for the fitness of plants a...

Current Biology Minireview Plant Sex Determination John R. Pannell Department of Ecology and Evolution, University of Lausanne, 1015 Lausanne, Switzerland Correspondence: [email protected] http://dx.doi.org/10.1016/j.cub.2017.01.052 Sex determination is as important for the fitness of plants as it is for animals, but its mechanisms appear to vary much more among plants than among animals, and the expression of gender in plants differs in impor- tant respects from that in most animals. In this Minireview, I provide an overview of the broad variety of ways in which plants determine sex. I suggest that several important peculiarities of plant sex determination can be understood by recognising that: plants show an alternation of generations between sporophytic and gametophytic phases (either of which may take control of sex determination); plants are modular in structure and lack a germ line (allowing for a quantitative expression of gender that is not common in animals); and separate sexes in plants have ultimately evolved from hermaphroditic ancestors. Most theorising about sex determination in plants has focused on dioecious species, but we have much to learn from monecious or hermaphroditic species, where sex is determined at the level of modules, tissues or cells. Because of the fundamental modularity of plant development and potentially important evolutionary links between monoecy and dioecy, it may be useful to relax the distinction often made between ‘developmental sex deter- mination’ (which underpins the development of male versus female flowers in monoecious species) and ‘ge- netic sex determination’ (which underpins the separation of males and females in dioecious species, often mediated by a genetic polymorphism and sex chromosomes). I also argue for relaxing the distinction be- tween sex determination involving a genetic polymorphism and that involving responses to environmental or hormonal cues, because non-genetic cues might easily be converted into genetic switches. Introduction and female flowers develop in monoecious plants such as maize Land plants are remarkably diverse in their sexual systems , or melon. In contrast, geneticists tend to think of sex determina- but the rich variety of ways in which they express their sexuality tion as the result of a genetic polymorphism (e.g., in species with has largely been built on a conserved foundation that determines sex chromosomes) that gives rise to the development of males whether reproductive tissues produce egg cells or sperm. In this or females in dioecious species, often initially through male- or sense, plants are similar to most other eukaryotes. But when female-sterility mutations. In this genetic sense, the term sex should the cell cycle end in the ultimate production of eggs determination would not apply to monoecious species at all, versus sperm, which individuals or flowers should make this because all individuals in a population could be genetically iden- decision, and how should the decision be acted upon in develop- tical at loci determining sex-organ development. mental genetic terms? The first of these questions (when?) rec- Although it is useful to remain alert to a possible distinction ognises that different parts of a plant, at different times, might in meaning between genetic versus developmental sex determi- be channelled towards being male versus female, and implies nation, a key point is that the two meanings are not, in fact, as that selection in a given lineage will often optimise a distribution distinct as they might appear to be. As I discuss below, evolu- of decisions among modules in the plant [3,4]. The second ques- tionary transitions have frequently taken place between dioecy tion (who?) is prompted by the fact that the life cycle of all land and monoecy, with the possibility that genes involved in plants involves an alteration of generations between haploid ga- developmental sex determination in monoecious populations metophytes and diploid sporophytes, and that one of these may play a role in genetic sex determination in dioecious popu- stages might take full or partial control of sex determination lations — although this is yet to be established. Moreover, fully (Figure 1). The third question (how?) refers to the developmental monoecious populations can simultaneously be genetically genetic means by which sex is determined in a given lineage, dimorphic, with a genetic class of hermaphrodites that empha- with the subtext that there might be some sort of general answer sise their male function (without being exclusively male), and a we might hope to find from a survey among different lineages second class of hermaphrodites that emphasise their female [5,6]. In this Minireview, I address these questions by taking an function. For these reasons, I blur the distinction between evolutionary perspective — we should expect sex to be deter- ‘genetic’ and ‘developmental’ definitions of sex determination, mined in ways influenced both by a lineage’s phylogenetic his- and consider them here together. tory, and by the action of selection on its populations’ sex ratio The topic of plant sex determination has been reviewed in sub- or sex-allocation strategies, ultimately by modifying suitable stantial detail in recent years [2,5–10]. An important take-home gene networks. message is that sex determination in plants is eclectic and varies For conceptual reasons, I take a broad view of what we mean a great deal among lineages [5,6]. The contrast with animals is by ‘sex determination’. Developmental biologists tend to use the striking. Whereas much the same gene networks are de- term ‘sex determination’ when referring to the way in which male ployed among highly divergent animals [2,11], the gene networks Current Biology 27, R191–R197, March 6, 2017 ª 2017 Elsevier Ltd. R191  Current Biology Minireview Figure 1. Relationships between main Sex chromosomes clades and grades of land plants. Control over sex Heterospory Illustrated are the key features of variation in: present present whether sex is controlled by the gametophytes, by sporophytes or, depending on the lineage, both; whether hetorospory is present or entirely absent Non-vascular plants from the group; and whether or not sex is known to be determined by sex chromosomes. Top image Bryophytes courtesy of Robert Kipps; centre and bottom im- ages by the author. Gametophyte No Yes Hepatophytes Vascular non-seed plants plant Anthocerophytes phytic phase. However, which phase has control over the decision, and how the Lycophytes Gametophyte Yes decision is enacted, varies. Sporophyte No No For an overview of sex determination in Monilophytes land plants, it is convenient to categorise their diversity into three groups — a non- Seed plants vascular group, including mosses, liver- Gymnosperms worts and hornworts (‘bryophytes’ in the Sporophyte Yes Yes broad sense); the vascular non-seed pro- ducing group, including lycophytes and Angiosperms monilophytes; and the clade of (vascular) Current Biology seed plants, including extant gymno- sperms and flowering plants (angio- sperms; Figure 1). In non-seed plants, ga- determining sex even among angiosperms, which have diverged metophytes may be male or female (in ‘dioicous’) species, with much more recently, appear to be very diverse indeed [5,6]. separate individuals producing either antheridia or archegonia, Consideration of the nature of plant sexuality, as well as the his- or they may be hermaphroditic (in ‘monoicous’ species), with tory of transitions between sexual systems, can help to compre- both antheridia and archegonia present in the same individuals. hend this diversity. Here, sex determination is a matter of the differentiation of cells In this Minireview, I begin with an overview of a few salient fea- and tissues in different parts of the same individuals. In dioicous tures of sex determination in the major land-plant lineages species, separate male and female gametophytes germinate (Figure 1). I then consider what we might infer about plant sex from spores of the same size. Sexual dimorphism nevertheless determination from the fact that so many plants are hermaphro- can be extreme. In some cases, dwarf male gametophytes ditic. Functional hermaphroditism in flowering plants takes a grow on larger female gametophytes. In many bryophytes, sex number of forms, ranging from the case where all flowers are is determined in gametophytes by (U and V) sex chromosomes ‘perfect’, with both male and female functions in each flower, (e.g., ). to ‘monoecy’, where individuals are hermaphroditic but each In ferns and lycophytes, the role and target of sex determina- flower only develops functional male (stamens) or female organs tion differ between ‘homosporous’ species, in which sporo- (carpels). Much can be learned about sex determination in plants phytes all produce spores of the same size (as in bryophytes), by comparing monoecy and dioecy. In both of these sexual sys- and ‘heterosporous’ species, in which sporophytes produce tems, flowers are either male (staminate) or female (pistillate), but both small ‘microspores’ and larger ‘megaspores’. In some ho- they differ in whether both sexes can be found in the same mosporous ferns, gametophytes are all functionally hermaphro- genetic individuals (monoecy) or only in separate genetic individ- ditic, with both antheridia and archegonia. In these cases, sex uals (dioecy). Finally, I reflect on the implications for our under- determination is a question of cellular and tissue differentiation standing of sex determination of the fact that plant gender is into different male and female ‘gametangia’. In other homospo- very often a highly plastic, quantitative trait. rous species, gametophytes may develop a unisexual function (producing either only antheridia or only archegonia). Here, the Sex Determination in the Major Land-plant Lineages sex-determination pathway targets separate, individual, game- The evolution of sexual reproduction constituted a major transi- tophytes. However, the gametophytes differentiate in response tion in the evolution of life, and occurred well before plants first to environmental cues perceived soon after spore germination ventured onto land some 470 million years ago. By that time, (Figure 2A), rather than in response to a genetic polymorphism, the ancestors of land plants already had differentiated male as in some mosses mentioned above. Indeed, to my knowledge, and female gametes (spermatozoids and eggs), and different sex chromosomes have not been described for any ferns or sex organs to produce them (archegonia and antheridia). The lycophytes. first land plants had also already evolved an alteration of haploid In ‘heterosporous’ ferns and lycophytes, sex determination and diploid generations (the gametophytes and sporophytes) acts through the size of the spore from which the gametophytes. In all land plants, the differentiation of sex organs into germinate. The gametophytes do not differ genetically, and the egg- or sperm-producing tissues occurs in the haploid gameto- sporophyte controls the sex of its gametophytes by regulating R192 Current Biology 27, R191–R197, March 6, 2017 Current Biology Minireview A the spore-producing ‘sporangia’ to produce either small ‘micro- HERs + – HERs spores’, which develop into male (micro-) gametophytes, or larger ‘megaspores’, which develop into female (mega-) gameto- TRAs MAN1 NOT1 TRAs phytes. Heterospory has evolved several times in vascular land plants, in ferns and lycophytes, as well as independently in the NOT1 FEM1 FEM1 MAN1 lineage that gave rise to seed plants. In angiosperms (flowering plants) and gymnosperms, the Male Exogenous cue: Female gametophytic phase has become even more highly modified gametophyte antheridiogen gametophyte and reduced, and differentiation between the male and female structures takes place in sporophytic structures in which the ga- metophytes form. In extant seed plants, the role of sex determi- nation is thus passed from the gametophytes entirely to their diploid sporophyte ‘hosts’. Seed-plant sporophytes may ulti- mately be functionally hermaphroditic, producing both pollen (modified and highly reduced male gametophytes) and ovules B (which contain modified female gametophytes), or they may be OGI XY XX male or female, producing only pollen or ovules. In contrast to MeGI MeGI non-seed plants, seed-plant gametophytes are always either male or female, and it is the sporophyte that determines their sex. The developmental machinery of sex determination in Stamen Stamen development seed plants therefore governs which gametophytes a given development diploid sporophyte individual produces, where on the plant they will occur, and in which proportions (i.e., the sex allocation; see below). Implications of Widespread Hermaphroditism for Sex Determination in Plants Male flower Female flower The great majority of plants are hermaphroditic angiosperms [1,15,16]. Most of these have ‘perfect flowers’, with a remarkably C conserved arrangement of their four floral whorls, the sepals, ACS11 ON OFF ACS11 petals, stamens, and the central carpels. The genetics of floral development, which also implies sex differentiation in specific WIP1 ASC-7 WIP1 ASC-7 tissues, is now reasonably well characterised in terms of the ABCE model [17,18], which envisages organ identity as the Carpels Stamens Carpels Stamens outcome of (partially overlapping) expression of genes in four gene classes. In particular, sex is determined by expression of Endogenous cue: ethylene B- and/or C-class genes that initiate the development of sta- mens or carpels. Although the action of these gene classes for organ differentiation in flowers has been investigated in detail in only a handful of species, it appears to apply quite generally across flowering plants, and probably evolved with the origin Female flower Male flower Current Biology of flowers. Given the highly conserved role of a number of genes in sex determination in animals , one might have ex- Figure 2. Three contrasting sex-determining pathways in land pected the ABCE genes to play a key role in determining sex in plants. dioecious species. However, although they do play a key role Black and grey text and symbols indicate genes and functions switched on in some lineages, they are rarely the genes that decide whether and off, respectively. Arrows and ‘flat-ended’ arrows indicate promotion or a plant becomes male or female [5,6]. suppression of activity, respectively. (A) Environmental determination of sex in the homosporous fern Ceratopteris richardii. Sex determination depends on The wide diversity of genetic mechanisms for sex determina- the outcome of two antagonistically interacting genes, FEM1 and TRA: in the tion in dioecious plants is attributable to the fact that separate absence of an exogenous signal from the hormone antheridiogen, the gene sexes have evolved repeatedly in different lineages, often TRA is expressed; this both elicits the development of a female (or her- maphrodite) gametophyte, and suppresses the expression of FEM1 through recently [15,20], and that genetic switches involved in sex an intermediary MAN1. In the presence of exogenous antheridiogen, the gene HER is expressed; HER suppresses the expression of TRA (and, in turn, of MAN1), allowing the expression of FEM1. FEM1 promotes both the develop- ment of a male gametophyte and the expression of the gene NOT1, which in The model invokes three interacting loci, with gene ASC11 suppressing the turn suppresses TRA. Images courtesy of Jody Banks. (B) Chromosomal carpel suppressor WIP1, and WIP1 also suppressing the stamen suppressor determination of sex in persimmons. In XY individuals (males), expression of ACS-7, allowing flowers to develop carpels and to be fully female; expression OGI, a small RNA, targets and suppresses a second autosomal gene MeGI, a of WIP1 thus both prevents the production of carpels and suppresses ACS-7, transcription factor that suppresses stamen development. In XX individuals, allowing flowers to develop stamens and thus to be fully male. An upstream which lack OGI, MeGI is expressed, suppressing stamen development. Im- regulatory cue (perhaps the hormone ethylene) turns ASC11 on or off in ages courtesy of Jeff Pippen. (C) Sex determination in monoecious melons. different parts of the plant. Images courtesy of Marcel Dorken. Current Biology 27, R191–R197, March 6, 2017 R193  Current Biology Minireview A Figure 3. The quantitative (and potentially 1.0 plastic) nature of sex determination in 1.0 plants. (A) Plots of the proportion of female flowers in two dioecious species, taken from the work of D. Lloyd (republished with permission from John Wiley and Sons, Inc., ), who first emphasized the impor- Relative rank tance of quantitative gender in plants. In both plots, the distribution of gender is plotted as the 0.5 0.5 relative rank gender of individuals sampled in the population, with closed and open circles repre- senting male and female plants, respectively. Left panel: Gingidia montana; right panel: Leptinella dendyi. (B) Chromosomal sex determination in the plant Mercurialis annua, showing an XX female (left), an XY males (centre), and an XY male with 0 0 ‘leaky’ gender expression, i.e., a male producing 0 0.5 1.0 0 0.5 1.0 both male and female flowers and fruits. Blue ar- rows indicate male (staminate) flowers; red arrows Proportion of female flowers in inflorescence indicate female (pistillate) flowers. Images cour- tesy of John Baker (left and centre right), Guillaume B Cossard (centre left) and Xinji Li (right). male sterility is recessive, as is most probable for a loss-of-function mutation [22,25], females should be homozygous (and thus XX in the case of a sex chromo- some), and hermaphrodites or males should be heterozygous (and thus XY, with female sterility or male enhancement Female (XX) ‘Leaky’ female (XX) Male (XY) ‘Leaky’ male (XY) genes being located on the Y). Little is known about the genes that Current Biology cause male and female sterility in dioe- cious plants, but it seems that genetic linkage between them on a sex chromo- determination have thus evolved many times independently some is common , as exemplified by recent work on white [5,21]. But generalisation is nonetheless possible if we consider campion (Silene latifolia) , wild strawberry (Fragaria virginiana) how transitions to dioecy have likely occurred. First, dioecy must and broadleaf arrowhead (Sagittaria latifolia). Papaya evolve through the positive selection of both male and female (Carica papaya) also has an XY sex-chromosome system, but sterility mutations in what are destined to become the sex-deter- a modified Yh chromosome is also present, and that fails to sup- mination genes. Theory predicts that male-sterility mutations are press the individual’s female function, allowing hermaphrodites most likely to spread in hermaphrodite populations first, largely to develop, and suggesting that the normal Y in papaya males because they prevent self-fertilizing, and thus from producing indeed harbours a female suppressor. Modified Y chromo- progeny that might suffer inbreeding depression, but also somes are similarly known in Silene latifolia and Vitis vinifera because it is difficult for female-sterility mutations to spread. In the annual mercury (Mercurialis annua), females have no in partially selfing populations [22,23]. The ‘gynodioecious’ pop- Y chromosome, and have acquired a male function in some pop- ulation that results may then evolve towards full dioecy, for ulations (Figure 3B), suggesting either that male-sterility example, in response to selection for sexual specialisation, mutations on an X chromosome have incomplete expression in either with the subsequent spread of a female-sterility mutation, these individuals, or that there are loci elsewhere in the genome or through the gradual shift in sex allocation of non-male-sterile that can partially restore male function in females. hermaphrodites towards greater maleness (the production of The two-locus model for the evolution of sex determination fewer seeds and more male flowers or pollen) [23,24]. and dioecy makes intuitive sense and has received some sup- The two likely alternative paths from gynodioecy to dioecy port , but sex is determined in many species by a single switch have important implications for how sex might be determined. gene. In Caucasian persimmon (Diospyros lotus), for In the case of the spread of a female-sterility mutation in a gyno- example, the anther fertility is determined by transcription of a dioecious population, the gene affected must become tightly Y-linked sequence encoding a small RNA that targets and sup- linked to the male-sterility mutation on a single (sex) chromo- presses expression of an unlinked autosomal feminizing gene some, because otherwise segregation or recombination could (Figure 2B). In other words, although two loci are involved create genotypes that carry and express both male and female at the top of the sex-determination cascade, they are unlinked sterility at the same time. Alternatively, the female-sterility to one another, and only one of them now acts as the switch mutation could show epistatic expression, masculinising only gene in the gene network. It is not yet known how such develop- those individuals not expressing the male-sterility mutation. If mental gene networks have been assembled during the R194 Current Biology 27, R191–R197, March 6, 2017 Current Biology Minireview evolution of dioecy. For example, did they evolve through the Such variation allows for multiple pathways to sex differentiation initial spread of (linked) male- and female-sterility mutations, as and potentially sex determination (Figure 3). supposed by theory ? Or might they have evolved through While the distribution of quantitative gender in a population the spread of a single sex-determining gene distinguishing be- may have a strong genetic component, individuals can also tween females and hermaphrodites, say, followed by the fixation modify their sex expression in response to context-dependent of mutations expressed only in hermaphrodites that reduce the opportunities or costs. Plastic sex expression applies not female allocation of the hermaphrodites? only to the quantitative nature of sex allocation in functional her- The plausibility of single-locus sex determination in some spe- maphrodites , but, more rarely, to complete sex change. cies is also suggested by the ontogeny of male and female ste- In jack-in-the-pulpit (Arisaema species), young plants are fully rility. In their review, Diggle et al. noted that stamen abortion in male, whereas, in later seasons, they become fully female or her- a given species tends to occur at approximately the same time maphroditic. Such ‘sexual diphasy’ is expected to evolve as carpel abortion. This pattern is not what we would immedi- both if the marginal benefits of reproducing as male versus fe- ately expect if sexual organ abortion results from the expression male change with plant size , and if there are additional ben- of independent male- and female-sterility mutations, unless it re- efits to sexual specialisation at a particular point in time (e.g., if flects a wide distribution of times since the evolution of dioecy in fitness is compromised by interference between simultaneous the respective lineages. Alternatively, the pattern hints at the male and female functions). Plasticity in sex allocation, and possibility that a single underlying switch acts within a gene sexual diphasy specifically, indicate that sex may be determined network to suppress one or the other sexual function at about not by a genetic dimorphism, but by the differential expression of the same time of floral differentiation. Such a scenario has genes shared by all individuals in a population in response to been elucidated in monoecious melons [35,36], although again physiological and environmental cues. it is by no means clear how it evolved. Sex in jack-in-the-pulpit is likely to be responsive to intrinsic In melons, and in maize, the sexual identity of flowers appears physiological signals reporting the resource status of the plant, to result from the expression of genes in different networks that but also to exogenous cues. Such responses have been docu- are responsive to positional information mediated by hormones. mented for many angiosperm species, although angiosperms In melons, gene expression at one locus causes suppression of do not seem to have evolved fully environmental sex determina- carpel development (female function), whereas expression at a tion (ESD) that might parallel, for example, temperature-depen- different locus suppresses stamen development (the male func- dent sex determination in some reptiles. In contrast, some homo- tion). Expression of an upstream gene that is thought to be sen- sporous ferns do show true ESD [6,48,49]. When the haploid sitive to local ethylene levels determines which of those two spores of the fern Ceratopteris richardii germinate, for example, genes is switched on (Figure 2C). In maize, another monoe- they develop into either male or female gametophytes, depend- cious species, different mutant classes have been identified that ing on the presence of potential mates (Figure 2A). In the absence suppress stamen or carpel development in female (ears) or male of mates, gametophytes produce both antheridia and arche- inflorescences (tassels), respectively. Just as male and fe- gonia. They also release into the water around them the hormone male flowers in melons tend to develop in different parts of the antheridiogen, which causes nearby gametophytes to develop as plant, ears and tassels in maize develop in leaf axils and at the miniature males. Some of the underlying genes involved in the plant apex, respectively; the gender distribution within the plant switch between hermaphroditic and male pathways have been is responsive to gradients in plant hormone levels , and con- identified [6,48]. Mutations at these loci render gametophytes forms to adaptive expectations for a wind-pollinated plant. insensitive to antheridiogen, and their sex is thus determined genetically by the mutations they carry (Figure 2A) [6,48]. In prin- The Quantitative and Plastic Nature of Plant Sex ciple, the spread of such mutations in a natural population could Determination conceivably convert ESD into genetic sex determination (GSD). In Plant gender is fundamentally a quantitative trait. Whereas monoecious melons cited above, or in maize, in which the sex of individuals in most gonochoristic animals are either fully male individual flowers or inflorescences depends on their location in or female, gender in plants is often better viewed on a quantita- the plant, genetic manipulation of mutants can render a popula- tive scale, measured in terms of the relative allocation to each tion effectively dioecious in similar ways [35,36], that is, experi- sexual function, or the proportion of genes transmitted through mental manipulation can convert a type of hormone-sensitive, sperm versus eggs (Figure 3A). Many dioecious species, tissue-specific ESD into GSD. Such examples show that ESD including those with well-developed sex chromosomes, show and GST may be readily interconverted. ‘leaky’ or ‘inconstant’ gender expression, with males and fe- males producing flowers of the other gender. Far from being Conclusions an aberrant feature of plant reproduction, such phenotypes are Much has been learned about how sex is determined in a handful typically fully functional for both sexes (e.g., Figure 3B). The of plant species. Although detailed analysis of a wider taxonomic quantitative distribution of gender in a population is under strong sample will no doubt be revealing, it is already becoming clear selection and can evolve rapidly when circumstances change — that different gene networks probably determine sex in different for example, when males are lost from a population. Such monoecious and dioecious lineages, and that generalisations shifts in sex allocation can occur at any level in the modular hier- are thus unlikely to emerge in terms of the classes of genes archy of plant development, from the number of pollen grains involved. However, the very diversity of sex-determination produced in anthers, to the number of floral or inflorescence mechanisms in plants reflects the important fact that separate primordia that develop with male versus female functions. sexes tend to have evolved independently in different lineages, Current Biology 27, R191–R197, March 6, 2017 R195  Current Biology Minireview and much more recently than in most animal lineages. This diver- 7. Ming, R., Bendahmane, A., and Renner, S.S. (2011). Sex chromosomes in land plants. Annu. Rev. Plant Biol. 62, 485–514. sity should not surprise us. But are there any general patterns in how gene networks come 8. Charlesworth, D. (2016). Plant sex chromosomes. Annu. Rev. Plant Biol. to be assembled that involve the co-ordination of both male- and 67, 397–420. female-sterility mutations, often in different parts of the plant or- 9. Janousek, B., and Mrackova, M. (2010). Sex chromosomes and sex deter- ganism? The two-locus model for the evolution of sex chromo- mination pathway dynamics in plant and animal models. Biol. J. Linn. Soc. 100, 737–752. somes and sex determination is satisfying in its potential gener- ality and logical coherence. In contrast, we have no similarly 10. Juarez, C., and Banks, J.A. (1998). Sex determination in plants. Curr. Opin. general proposition for how single-locus sex switches become Plant. Biol. 1, 68–72. embedded in necessarily complex networks. For instance, do 11. Bachtrog, D., Mank, J.E., Peichel, C.L., Kirkpatrick, M., Otto, S.P., Ash- dioecious species employ the same developmental pathways man, T.L., Hahn, M.W., Kitano, J., Mayrose, I., Ming, R., et al. (2014). Sex determination: why so many ways of doing it? PLoS Biol. 12, as the monoecious species from which they are derived? Ad- e1001899. dressing these questions should be a major goal for future 12. Kenrick, P., and Crane, P.R. (1997). The origin and early evolution of plants research, for example, through a comparative analysis of gene on land. Nature 389, 33–39. networks in related species that express their gender differently. The questions of when and where on a plant which sex is ex- 13. Okada, S., Sone, T., Fujisawa, M., Nakayama, S., Takenaka, M., Ishizaki, K., Kono, K., Shimizu-Ueda, Y., Hanajiri, T., Yamato, K.T., et al. (2001). The pressed has long been accessible through analysis of competing Y chromosome in the liverwort Marchantia polymorpha has accumulated allocation strategies in an evolutionary framework. The theoret- unique repeat sequences harboring a male-specific gene. Proc. Nat. Acad. Sci. USA 98, 9454–9459. ical underpinnings for such an understanding are well-estab- lished , and observed patterns of sex expression and sex allo- 14. Bateman, R.M., and Dimichele, W.A. (1994). Heterospory - the most itera- cation often conform nicely to expectations – though many tive key innovation in the evolutionary history of the plant kingdom. Biol. Rev. Camb. Philo. Soc. 69, 345–417. puzzles remain. Because sex determination directly affects an important currency of fitness (reproductive success), it is not sur- 15. Renner, S.S. (2014). The relative and absolute frequencies of angiosperm sexual systems: dioecy, monoecy, gynodioecy, and an updated online prising that its responses to selection can be rapid and dramatic. database. Am. J. Bot. 101, 1588–1596. An interesting question that deserves further attention here concerns shifts between plasticity and genetic polymorphism. 16. Yampolsky, C., and Yampolsky, H. (1922). Distribution of sex forms in the phanerogamic flora. Bibl. Genet. 3, 4–62. How often plasticity in sex expression acts as a capacitor for the evolution of genetically separate sexes is not known, but 17. Weigel, D., and Meyerowitz, E.M. (1994). The ABCs of floral homeotic genes. Cell. 78, 203–209. the facility with which these modes of sex determination can be interconverted experimentally suggests that they do not 18. Soltis, D.E., Chanderbali, A.S., Kim, S., Buzgo, M., and Soltis, P.S. (2007). The ABC model and its applicability to basal angiosperms. Ann. Bot. 100, reflect fundamental evolutionary constraints. Nonetheless, it 155–163. remains puzzling as to why a chromosomal polymorphism that determines sex in the gametophytes of bryophytes has never re- 19. Chanderbali, A.S., Yoo, M.J., Zahn, L.M., Brockington, S.F., Wall, P.K., Gitzendanner, M.A., Albert, V.A., Leebens-Mack, J., Altman, N.S., Ma, placed environmental sex determination in ferns or lycophytes. It H., et al. (2010). Conservation and canalization of gene expression during is tempting to think that the distinction is linked to the ephemeral angiosperm diversification accompany the origin and evolution of the flower. Proc. Nat. Acad. Sci. USA 107, 22570–22575. nature of fern but not bryophyte gametophytes , but why there should be such a link is not obvious. 20. Kafer, J., Marais, G.A.B., and Pannell, J.R. (2017). On the rarity of dioecy in flowering plants. Mol. Ecol. http://dx.doi.org/10.1111/mec.14020. ACKNOWLEDGEMENTS 21. Charlesworth, D. (2002). Plant sex determination and sex chromosomes. Heredity 88, 94–101. I thank Jörn Gerchen, Paris Veltsos, Wen-Juan Ma and three anonymous re- 22. Charlesworth, D., and Charlesworth, B. (1978). A model for the evolution of viewers for comments on the manuscript, Deborah Charlesworth and mem- dioecy and gynodioecy. Am. Nat. 112, 975–997. bers of my lab group for valuable discussions, and the University of Lausanne and the Swiss National Science Foundation for funding. 23. Dufay, M., Champelovier, P., Kafer, J., Henry, J.P., Mousset, S., and Mar- ais, G.A.B. (2014). An angiosperm-wide analysis of the gynodioecy-dioecy REFERENCES pathway. Ann. Bot. 114, 539–548. 24. Charlesworth, D. (1999). Theories of the evolution of dioecy. In Gender and 1. Barrett, S.C.H. (2002). The evolution of plant sexual diversity. Nat. Rev. Sexual Dimorphism in Flowering Plants, M.A. Geber, T.E. Dawson, and Genet. 3, 274–284. L.F. Delph, eds. (Heidelberg: Springer), pp. 33–60. 2. Beukeboom, L.W., and Perrin, N. (2014). The Evolution of Sex Determina- 25. Wright, S. (1934). Physiological and evolutionary theories of dominance. tion (Oxford: Oxford University Press). Am. Nat. 68, 24–53. 3. Charnov, E.L. (1982). The Theory of Sex Allocation (Princeton, NJ: Prince- 26. Westergaard, M. (1958). The mechanism of sex determination in dioecious ton University Press). plants. Adv. Genet. 9, 217–281. 4. West, S.A. (2009). Sex Allocation (Princeton: Princeton University Press). 27. Fujita, N., Torii, C., Ishii, K., Aonuma, W., Shimizu, Y., Kazama, Y., Abe, T., and Kawano, S. (2012). Narrowing down the mapping of plant sex- 5. Diggle, P.K., Di Stilio, V.S., Gschwend, A.R., Golenberg, E.M., Moore, determination regions using new Y-chromosome-specific markers and R.C., Russell, J.R.W., and Sinclair, J.P. (2011). Multiple developmental heavy-ion beam irradiation-induced Y-deletion mutants in Silene latifolia. processes underlie sex differentiation in angiosperms. Trends Genet. 27, G3-Genes Genomes Genet. 2, 271–278. 368–376. 28. Spigler, R.B., Lewers, K.S., Main, D.S., and Ashman, T.L. (2008). Genetic 6. Tanurdzic, M., and Banks, J.A. (2004). Sex-determining mechanisms in mapping of sex determination in a wild strawberry, Fragaria virginiana, land plants. Plant Cell. 16, S61–S71. reveals earliest form of sex chromosome. Heredity 101, 507–517. R196 Current Biology 27, R191–R197, March 6, 2017 Current Biology Minireview 29. Dorken, M.E., and Barrett, S.C.H. (2004). Sex determination and the evo- 39. Harder, L.D., and Prusinkiewicz, P. (2013). The interplay between inflores- lution of dioecy from monoecy in Sagittaria latifolia (Alismataceae). Proc. cence development and function as the crucible of architectural diversity. Ro. Soc. London. B. 271, 213–219. Ann. Bot 112, 1477–1493. 30. VanBuren, R., Zeng, F.C., Chen, C.X., Zhang, J.S., Wai, C.M., Han, J., 40. Lloyd, D.G., and Bawa, K.S. (1984). Modification of the gender of seed Aryal, R., Gschwend, A.R., Wang, J.P., Na, J.K., et al. (2015). Origin and plants in varying conditions. Evol. Biol. 17, 255–338. domestication of papaya Y-h chromosome. Genome Res. 25, 524–533. 41. Lloyd, D.G. (1980). Sexual strategies in plants. III. A quantitative method 31. Picq, S., Santoni, S., Lacombe, T., Latreille, M., Weber, A., Ardisson, M., for describing the gender of plants. NZ J. Bot. 18, 103–108. Ivorra, S., Maghradze, D., Arroyo-Garcia, R., Chatelet, P., et al. (2014). A small XY chromosomal region explains sex determination in wild dioecious 42. Dorken, M.E., and Pannell, J.R. (2009). Hermaphroditic sex allocation V-vinifera and the reversal to hermaphroditism in domesticated grape- evolves when mating opportunities change. Curr. Biol. 19, 514–517. vines. BMC Plant Biol. 14, 17. 32. Russell, J.R.W., and Pannell, J.R. (2015). Sex determination in dioecious 43. Delph, L.F., and Wolf, D.E. (2005). Evolutionary consequences of gender Mercurialis annua and its close diploid and polyploid relatives. Heredity plasticity in genetically dimorphic breeding systems. New Phytol. 166, 114, 262–271. 119–128. 33. Renner, S.S. (2016). Pathways for making unisexual flowers and unisexual 44. Schlessman, M.A. (1988). Gender diphasy (‘‘sex choice’’). In Plant Repro- plants: Moving beyond the ‘‘two mutations linked on one chromosome’’ ductive Ecology: Patterns and Strategies, J. Lovett Doust, ed. (New York: model. Am. J. Bot. 103, 587–589. Oxford University Press), pp. 139–151. 34. Akagi, T., Henry, I.M., Tao, R., and Comai, L. (2014). A Y-chromosome-en- 45. Bierzychudek, P. (1984). Determinants of gender in jack-in-the-pulpit: the coded small RNA acts as a sex determinant in persimmons. Science 346, influence of plant size and reproductive history. Oecologia 65, 14–18. 646–650. 46. Zhang, D.Y., and Jiang, X.H. (2002). Size-dependent resource allocation 35. Boualem, A., Troadec, C., Camps, C., Lemhemdi, A., Morin, H., Sari, M.A., and sex allocation in herbaceous perennial plants. J. Evol. Biol. 15, 74–83. Fraenkel-Zagouri, R., Kovalski, I., Dogimont, C., Perl-Treves, R., et al. (2015). A cucurbit androecy gene reveals how unisexual flowers develop 47. Charnov, E.L., and Bull, J. (1977). When is sex environmentally deter- and dioecy emerges. Science 350, 688–691. mined? Nature 266, 228–230. 36. Ma, W.J., and Pannell, J.R. (2016). Sex determination: separate sexes are a double turnoff in melons. Curr. Biol. 26, R171–R174. 48. Banks, J.A. (1994). Sex-determining genes in the homosporous fern Cera- topteris. Development 120, 1949–1958. 37. Irish, E.E. (1999). Maize sex determination. In Sex Determination in Plants, G.C. Ainsworth, ed. (Oxford: Bios Scientific), pp. 183–188. 49. Haig, D., and Westoby, M. (1988). Sex expression in homosporous ferns: An evolutionary perspective. Evol. Trends. Plants. 2, 111–120. 38. Rood, S.B., Pharis, R.P., and Major, D.J. (1980). Changes of endogenous giberellin-like substantces with sex reverssal of the apical inflorescence of 50. Lloyd, D.G. (1980). The distribution of gender in four angiosperm species corn. Plant Physiol 66, 793–796. illustrating two evolutionary pathways to dioecy. Evolution 34, 123–134. Current Biology 27, R191–R197, March 6, 2017 R197