environmental sex determination in the genus equisetum

Int. J. Plant Sci. 163(5):825–830. 2002. 2002 by The University of Chicago. All rights reserved.1058-5893/2002/16305-0014$15.00ENVIRONMENTAL SEX DETERMINATION IN THE GENUS EQUISETUM: SUGARSINDUCE MALE SEX EXPRESSION IN CULTURED GAMETOPHYTESJean-Michel Guillon 1, * , † and Christian Raquin**Laboratoire Ecologie, Systématique et Evolution, Unité Propre de Recherche de l’Enseignement Supérieur Associée au Centre National de laRecherche Scientifique 8079, Université Paris-Sud, F-91405 Orsay Cedex, France; and †Conservatoire Botanique National duBassin Parisien, Muséum National d’Histoire Naturelle, 61 rue Buffon, F-75005 Paris, FranceHorsetails (Equisetum, Sphenophyta) are homosporous, and sexual differentiation of Equisetum gametophytesis under the influence of environmental conditions. Still, the environmental cues responsible for sexdetermination of Equisetum gametophytes in vitro and in the wild have remained elusive. Here, we show thatsignificantly different sex ratios are obtained when gametophytes are grown on media with or without sugar.In our experimental conditions, male gametophytes outnumber females in the presence of 60–120 mM sucroseand 120 mM glucose, whereas in the absence of sugar, most gametophytes differentiate as female. A similareffect is also observed on already differentiated female gametophytes, which become hermaphroditic soonerwhen cultured in the presence of sucrose in vitro. Interestingly, these results are reproducible within and acrossspecies representative of the two subgenera Equisetum and Hippochaete, indicating that the entire genus mayshare an identical pattern of environmental sex determination.Keywords: sex determination, Equisetum, environment, gametophyte, in vitro culture, sugars.IntroductionSeedless vascular plants are a paraphyletic group of plantscharacterized by a haplo-diplophasic life cycle. Among these,Isoetaceae, Selaginellaceae, and Marsileaceae are heterosporous,i.e., producing micro- and macrospores that differ inmorphology, whereas club mosses (Lycopodium), horsetails(Equisetum), and most ferns (Filicophyta) are homosporous.In Equisetum, perennial sporophytes (2n) produce sexuallyundifferentiated spores (n) that are dispersed and germinateto form photosynthetic male, female, or sometimes bisexualgametophytes (n). Fertilizations between male and female gametes(n) result in the formation of new sporophytes.The sexual determination of Equisetum gametophytes hasbeen problematic for many years, with various authors emphasizinggenetic sex determination, environmental sex determination,or a mixture of both (Mohan Ram and Chatterjee1970; Laroche et al. 1972, 1978; Hauke 1977). However, theobservations that (i) archegonial (female) gametophytes alwayschange to antheridial (male) lobe production when culturedover a prolonged time, (ii) secondary archegonial branches canbe regenerated from otherwise purely male gametophytes, and(iii) the proportions of male and female gametophytes developingfrom the same initial set of spores are highly sensitiveto growth conditions all indicate that the spores of Equisetumare not individually predetermined for the expression of onesex or the other (Duckett 1970, 1972, 1977, 1979). Instead,1Author for correspondence; address for correspondence: LaboratoireEcologie, Systématique et Evolution, Université Paris-Sud, Bâtiment362, F-91405 Orsay Cedex, France; telephone 33-1-69-15-63-42; fax 33-1-69-15-73-53; e-mail jean-michel.guillon@ese.u-psud.fr.Manuscript received October 2001; revised manuscript received March 2002.in each developing gametophyte, a series of threshold conditionsappears to determine which sex is expressed (Duckett1977).To better understand sex determination in Equisetum gametophytes,several authors have studied their response to anumber of culture conditions in vitro (Mohan Ram and Chatterjee1970; Hauke 1971, 1977). Studied environmental factorsinclude light quality and level, crowding, pH, and sucroseconcentration in the culture media. Although each of thesefactors seems able to affect the sex ratio, in many cases, theeffects are not consistent among different species. For example,sucrose enrichment has been reported (i) to inhibit female sexdifferentiation in Equisetum ramosissimum (Mohan Ram andChatterjee 1970), (ii) to decrease the proportion of males inEquisetum arvense (Hauke 1971, 1977), (iii) to increase theproportion of males in E. ramosissimum (Mohan Ram andChatterjee 1970) and Equisetum hyemale (Hauke 1971), or(iv) to have no effect on sex expression in the same E. hyemale(Hauke 1977). Hauke (1977) concluded: “A number of experimentsutilizing sugar enrichment were attempted, but theresults were so inconsistent that no conclusions can be drawnfrom them” (p. 29). Such discrepancies may be explained byvariations in culture conditions between experiments, leadingto the interaction of other important factors such as light qualityand crowding. Alternatively, there may be variability amongspores collected from different individuals of the same speciesand/or from individuals of different species.In an attempt to characterize the pattern of sex expressionin Equisetum gametophytes from different species, we conductedrepeated and controlled culture experiments in the presenceof various sugar concentrations. Also, we examined thestages during which sex differentiation can be modified bysugar. Here, we present evidence that the male-promoting ef-825

826 INTERNATIONAL JOURNAL OF PLANT SCIENCESfect of sucrose and glucose in the culture medium is significantand can be reproduced within and across species of Equisetum.We also show that the addition of sucrose to the culture mediumcan induce mature female gametophytes to change sex.The significance of this finding is discussed in relation to lifehistory in this ancestral lineage of plants and in the light ofevolutionary theories of sex determination.Material and MethodsSpore Collection and StorageSpores were collected from one cone per plant. For eachEquisetum species, plants from two distinct localities were usedas sources of spores. Equisetum palustre L.: P11, P12, CampusUniversity Paris XI, Orsay, France; P2, northern bank of theGreat Pond, Saint-Mars-La-Brière, France. Equisetum arvenseL.: A1, Campus University Paris XI, Orsay, France; A2, CampusUniversity Paris XI, Bures-sur-Yvette, France. Equisetumtelmateia Ehrh.: T1, Campus University Paris XI, Orsay,France; T2, Moulin de la Mare, Milon-La-Chapelle, France.Equisetum variegatum Schleich.: plants cultivated in the ConservatoireNational Botanique du Bassin Parisien and originatingfrom (V1) Forest of Marly, Marly, France, and (V2)Botanical garden, Samoens, France.For E. variegatum and E. palustre, spores were collectedfrom mature cones and stored in hermetically closed tubes at70C. For E. arvense and E. telmateia, mature but unopenedcones were harvested and sterilized for 5 min in 10 g L 1 ofcalcium hypochlorite to which 50 mL of Teepol (Lever) detergentper 100 mL was added. Cones were rinsed in distilledwater, allowed to dry overnight, and then dissected to releasethe spores, which were stored as above.In Vitro CultureG0 culture medium contained the macronutrients of Estelleand Sommerville (1987): 2.5 mM KH 2 PO 4 replaced by 2 mMKH 2 PO 4 and 0.5 mM K 2 HPO 4 so that the medium was directlybuffered to pH 6.1, the mineral micronutrients of Murashigeand Skoog (1962) diluted by half, and 10 mM Fe EDTA. Phytagel(0.2% w/v; Sigma) and agar (0.4% w/v; Biofit, Fisher)were used as gelling agents before autoclaving for 12 min at120C.Whereas G0 contained no additional component, 120 mMglucose was added to G120G, 30 mM sucrose was added toG30S, 60 mM sucrose was added to G60S, 90 mM sucrosewas added to G90S, and 120 mM sucrose was added to G120S.Sixty mM mannitol was added to G60ML for a control as anunmetabolized osmotic agent. Sugars or mannitol were addedto the culture medium before sterilization: sucrose and glucose(Prolabo, RP NORMAPUR) and mannitol (Merck).Gametophyte cultures were placed in a growth chamber atconstant 241C and 24 3 mmol m 2 s 1 of photosyntheticallyactive radiation for 16 h d 1 provided by fluorescentlamps, a 1/1 mixture of Philips TLD36W82, 2700K, andTLD36W33, 4000K.For the primary sexual differentiation experiment, at t p10 d, spores were sown on G0 petri dishes and then placedin the growth chamber. At t p 0, gametophytes were trans-ferred to test-media dishes. Test media were G0, G30S, G60S,G90S, G120S, G120G, and G60ML. Sixteen randomly chosengametophytes were transferred to each test-medium dish, witheach medium replicated in three dishes (for P11 and V1) orfour dishes (for P2, V2, A1, A2, T1, T2). The gametophyteswere observed under a binocular dissecting microscope everyweek after transfer to the test media until the end of the experiment.For each gametophyte, viability and gender weredetermined according to the presence or absence of antheridiaand archegonia, with each gametophyte recorded as female,male, bisexual, undifferentiated, or dead. At the end of theexperiment, 5 wk for E. telmateia, 8 wk for E. arvense andE. variegatum,and9wkforE. palustre, pairs of gametophyteswere transferred into a flooded G30S tube to assess the fertilityof some of the males and females grown on G0, G30S, G90S,and G120S. In all cultures, a sporophyte developed after3–4 wk, indicating that the sex organs and gametes werefunctional.For hermaphroditization experiments, spores were sown onG0 dishes, and gametophytes were transferred to G0 for primarysex expression. When the gender of gametophytes couldbe determined, female gametophytes were transferred to testmedia and further grown as above. Nine randomly chosenfemale gametophytes for P12 and 12 female gametophytes forT2 and V1 were transferred to each test-medium dish, andeach medium was replicated over three dishes. The female gametophyteswere observed every week after transfer to the testmediumplates, and the appearance of antheridia was recorded.Statistical AnalysesThe statistical analyses included only plates in which lessthan three gametophytes failed to grow or to differentiate( undifferentiated + dead ! 3). In the early sex-differentiationexperiment, the sex ratio was defined as ( males + bisexuals)/( males + bisexuals + females), accounting for the observationof male sex expression both in initially male gametophytesand in rare bisexuals deriving from initially female gametophytes.The sex ratio on each plate was square root– andarcsine-transformed and then submitted to ANOVA, withtest medium and plant species as fixed effects and plantorigin (nested within species) as a random effect (generallinear models procedure; SAS Institute 1988). In the hermaphroditizationexperiment, where only one plant per specieswas used, the statistical treatment was the same asabove, except that the sex ratio on each plate was definedas bisexuals/(bisexuals + females) and that the only effect in-cluded in the model was test medium. Tukey’s studentizedrange test was performed for multiple comparisons. Therelation between sucrose concentration and the sex ratioobtained on each dish was assessed by using the SAS regressionprocedure.ResultsTo test the effect of sugar addition on sex expression in fourspecies of Equisetum, we performed two different experiments,examining the influence of growth medium on sex expression,starting with either 11-d-old undifferentiated Equisetum gametophytes(table 1; fig. 1) or older differentiated female Equisetumgametophytes (fig. 2).

828 INTERNATIONAL JOURNAL OF PLANT SCIENCESFig. 1 Sex ratios in cultures of Equisetum gametophytes. For each plant, bars represent the mean sex ratio ([ male + bisexual]/[ male + bisexual + female]) and associated standard error obtained on the corresponding test medium. The number of repeats (n) and the resultof multiple comparisons (Tukey’s studentized range test) are also indicated. Treatments with the same letter do not significantly differ (P 10.05). Treatments that differ from G0 ( P ! 0.05) are indicated with bars in black. The duration of growth on test media is specified with thename of the plant. A1 and A2, Equisetum arvense; P11 and P2, Equisetum palustre; T1 and T2, Equisetum telmateia; V1 and V2, Equisetumvariegatum.stress could account in part for the observed effect because theaddition of an osmotic agent, 60 mM mannitol, or a sugar,60 mM sucrose, produced similar sex ratios (fig. 1). Higherconcentrations of mannitol should be used to test thishypothesis.When reviewing the variations in sex ratio of cultured Equisetumgametophytes, investigators have noticed that underconditions favoring vigorous vegetative growth of young gametophytes,e.g., high light levels, most spores differentiateinto initially female individuals (Duckett 1970). However, less

GUILLON & RAQUIN—SEX DETERMINATION IN EQUISETUM 829Fig. 2 Timing of hermaphrodite appearance in initially femalegametophyte cultures of P12 (Equisetum palustre), T2 (Equisetum telmateia),and V1 (Equisetum variegatum). For each plant and test medium,the mean sex ratio with its associated standard error is shownas a function of time. Each treatment was replicated in three dishes.Circles, G0 (no sucrose added); squares, G60S (60 mM sucrose); triangles,G90S (90 mM sucrose).favorable regimes, e.g., crowding, drought, mineral deficiency,or high temperatures, produce a majority of males (Duckett1970; Mohan Ram and Chatterjee 1970; Hauke 1977). Observationsof wild populations of Equisetum gametophyteshave shown that high frequencies of males correspond withlow growth rates, possibly resulting from environmental stress(Duckett and Duckett 1980). Environmental factors are alsoknown to influence sex ratios in several seed plant species,with males most typically tending to be overrepresented inresource-limited or less favorable environments (Lloyd andBawa 1984; Korpelainen 1998). This general pattern of observationsis consistent with the hypothesis that exogenous sugarscould be a stress to cultured gametophytes of Equisetum.Evolutionary Significance of Sex Determinationin Equisetum GametophytesEquisetum species have a haplo-diplophasic life cycle, characterizedby the alternation of a perennial sporophytic generationand a short-lived gametophytic generation. Not all sporophytepopulations produce spores, and they can propagateefficiently through vegetative reproduction. New populationsmay be founded either by broken-off or dispersed vegetativeparts of the sporophytes or by dispersed spores and sexualreproduction between emerging gametophytes. Studies of geneticdiversity in E. arvense and Equisetum hyemale populations,as well as high fertilization rates found in wild gametophytepopulations, indicate that sexual reproduction occursand cannot be dismissed as marginal in Equisetum (Duckettand Duckett 1980; Soltis et al. 1988; Korpelainen and Kolkkala1996).In the wild, observed differences in sex ratios between populationsof the same species at the same time and betweenpopulations at the same locality in different years all indicatean important environmental component for sex determinationin Equisetum (Duckett and Duckett 1980). Evolutionary theoriespredict that environmental sex determination is favored(i) when early in life, an individual enters a patchy environmentthat has a large effect on its lifetime fitness (some patches beingrelatively more beneficial to females than to males and somebeing the opposite), (ii) when individuals have little controlover which patch they enter, and (iii) when individuals fromdifferent patches mate randomly (Charnov and Bull 1977; Bull1981). Equisetum gametophytes might fit this model if naturalpopulations inhabited patchy places made of microenvironmentsconferring more or less an advantage to one sex. It hasalso been shown that short-lived species are less prone toevolve environmental sex determination because environmentalfluctuations may cause detrimental sex-ratio variations insuch a system (Bull and Bulmer 1989). For this reason, strongsex-specific fitness effects of the environment would be expectedin order to explain the occurrence of environmental sexdetermination in the seasonal Equisetum gametophytes in thewild. Since the literature does not provide such fitness-relateddata, new work is clearly needed to examine whether the biologyof Equisetum gametophytes conforms to the predictionsof the Charnov and Bull (1977) model.As shown here, male sex expression can be induced in fourspecies of Equisetum by the addition of sugar to the growthmedium. However, the addition of sugar in our experimentaldesign does not mimic any specific environmental factor thatgametophytes would encounter in the wild. Hence, sex determinationin response to sugar concentration is unlikely to haveevolved per se, but exogenous sugars must somehow interferewith the physiology of the developing gametophyte in such away that sex determination is affected. Sex determination inhorsetails seems to integrate many physiological signals arisingfrom the interaction of the gametophyte with its environment.Indeed, many environmental factors, such as light quality andlevel, competition, and pH, are also able to modify sex expressionin Equisetum gametophytes (Mohan Ram and Chatterjee1970; Hauke 1971, 1977). In this context, it is quiteunderstandable that sugar concentration in the growth mediumis one among many factors that influence sex determinationof gametophytes cultured in vitro.Lability of sex expression occurs in many plant taxa, butonly in homosporous pteridophytes is labile sex the rule (Korpelainen1998). Indeed, sex determination in horsetails is reminiscentof that found in ferns, where undifferentiated game-

830 INTERNATIONAL JOURNAL OF PLANT SCIENCEStophytes respond to antheridogen hormones secreted by femalegametophytes. The evolutionary significance of this informationsystem in fern gametophytes is still a matter of debate(Willson 1981; Haig and Westoby 1988; Korpelainen 1998).Because ferns and horsetails are sister clades (Kenrick andCrane 1997; Pryer et al. 2001), sex determination in horsetailscould be evolutionarily related to that found in ferns, but antheridogenhormones have not been detected in horsetails(Hauke 1971). Alternatively, environmental sex determinationmight have evolved more than once in relation to the distinctivelife history of ferns and horsetails because they both are homosporoushaplo-diplophasic land plants with perennial sporophytesand autonomous short-lived gametophytes. In particular,it has been proposed that the extreme lability in thesex expression of homosporous pteridophytes may be primarilyrelated to their peculiar mating system (Korpelainen1998).AcknowledgmentsWe would like to thank Jacques Moret and the people atthe Conservatoire Botanique National du Bassin Parisien fortheir support of this work. We thank Claire Lavigne, MarcGirondot, Bryan Epperson, and Jacqui Shykoff for discussionsand useful comments on the manuscript. Odylle Cudelou, OliviaCalvez, Alexandra Deville, and Amélie Pinot are acknowledgedfor their help in gametophyte culture.Literature CitedBull JJ 1981 Evolution of environmental sex determination from genotypicsex determination. Heredity 47:173–184.Bull JJ, MG Bulmer 1989 Longevity enhances selection of environmentalsex determination. Heredity 63:315–320.Charnov EL, JJ Bull 1977 When is sex environmentally determined?Nature 266:828–830.Duckett JG 1970 Sexual behaviour of the genus Equisetum, subgenusEquisetum. Bot J Linn Soc 63:327–352.——— 1972 Sexual behaviour of the genus Equisetum, subgenusHippochaete. Bot J Linn Soc 65:87–108.——— 1977 Towards an understanding of sex determination in Equisetum:an analysis of regeneration in gametophytes of the subgenusEquisetum. Bot J Linn Soc 74:215–242.——— 1979 Comparative morphology of the gametophytes of Equisetum,subgenus Hippochaete and the sexual behaviour of E. ramosissimumsubsp. debile, (Roxb.), E. hyemale var. affine (Engelm.)A.A., and E. laevigatum A.Br. Bot J Linn Soc 79:179–203.Duckett JG, AR Duckett 1980 Reproductive biology and populationdynamics of wild gametophytes of Equisetum. Bot J Linn Soc 80:1–40.Estelle MA, C Sommerville 1987 Auxin resistant mutants of Arabidopsisthaliana with an altered morphology. Mol Gen Genet 207:200–206.Haig D, M Westoby 1988 Sex expression in homosporous ferns: anevolutionary perspective. Evol Trends Plants 2:111–119.Hauke RL 1971 The effect of light quality and intensity on sexualexpression in Equisetum gametophytes. Am J Bot 58:373–377.——— 1977 Experimental studies on growth and sexual determinationin Equisetum gametophytes. Am Fern J 67:18–31.Kenrick P, PR Crane 1997 The origin and early evolution of plantson land. Nature 389:33–39.Korpelainen H 1998 Labile sex expression in plants. Biol Rev 73:157–180.Korpelainen H, M Kolkkala 1996 Genetic diversity and populationstructure in the outcrossing populations of Equisetum arvense andE. hyemale (Equisetaceae). Am J Bot 83:58–62.Laroche J, C Guervin, C Le Coq 1978 Le génome et l’expressionsexuelle chez l’Equisetum arvense L. I. Comportement du gamétophytecultivé sur milieu témoin. Rev Gen Bot 85:221–254.Laroche J, Van Thi Dao 1972 Peut-on admettre un déterminisme génétiquedu sexe chez certaines espèces d’Equisetum? Bull Soc BotFr 119:363–372.Lloyd DG, KS Bawa 1984 Modification of the gender of seed plantsin varying conditions. Evol Biol 17:255–338.Mohan Ram HY, J Chatterjee 1970 Gametophytes of Equisetum ramosissimumsubsp. ramosissimum. II. Sexuality and its modification.Phytomorphology 20:151–172.Murashige T, F Skoog 1962 A revised medium for rapid growth andbioassays with tobacco tissue culture. Physiol Plant 15:473–497.Pryer KM, H Schneider, AR Smith, R Cranfill, PG Wolf, JS Hunt, SDSipes 2001 Horsetails and ferns are a monophyletic group and theclosest living relatives to seed plants. Nature 409:618–621.SAS Institute 1988 SAS/STAT user’s guide, release 6.03 ed. SAS Institute,Cary, N.C.Soltis DE, PS Soltis, RD Noyes 1988 An electrophoretic investigationof intragametophytic selfing in Equisetum arvense. Am J Bot 75:231–237.Willson MF 1981 Sex expression in fern gametophytes: some evolutionarypossibilities. J Theor Biol 93:403–409.