germline x chromosomes exhibit contrasting patterns of histone

Developmental Biology 245, 71–82 (2002)doi:10.1006/dbio.2002.0634, available online at http://www.idealibrary.com on

Germline X Chromosomes Exhibit ContrastingPatterns of Histone H3 Methylationin Caenorhabditis elegans

Melanie Reuben and Rueyling Lin1

Department of Molecular Biology, University of Texas Southwestern Medical Center,Dallas, Texas 75390-9148

In mammals, one of the two somatic X chromosomes in the female is inactivated, thereby equalizing X chromosome-derived transcription in the two sexes, a process known as dosage compensation. In the germline, however, the situation isquite different. Both X chromosomes are transcriptionally active during female oogenesis, whereas the X and Ychromosomes are transcriptionally silent during male spermatogenesis. Previous studies suggest that Caenorhabditiselegans germline X chromosomes might have different transcriptional activity in the two sexes in a manner similar to thatin mammals. Using antibodies specific to H3 methylated at either lysine 4 or lysine 9, we show that the pattern ofsite-specific H3 methylation is different between X chromosomes and autosomes as well as between germline Xchromosomes from the two sexes in C. elegans. We show that the pachytene germline X chromosomes in both sexes lackMe(K4)H3 when compared with autosomes, consistent with their being transcriptionally inactive. This transcriptionalinactivity of germline X chromosomes is apparently transient in hermaphrodites because both X chromosomes stainbrightly for Me(K4)H3 after germ nuclei exit pachytene. The male single X chromosome, on the other hand, remains devoidof Me(K4)H3 staining throughout the germline. Instead, the male germline X chromosome exhibits a high level of Me(K9)H3that is not detected on any other chromosomes in either sex, consistent with stable silencing of this chromosome. Usingmutants defective in the sex determination pathway, we show that X-chromosomal Me(K9)H3 staining is determined by thesexual phenotype, and not karyotype, of the animal. We detect a similar high level of Me(K9)H3 in male mouse XY bodies,suggesting an evolutionarily conserved mechanism for silencing the X chromosome specifically in the malegermline. © 2002 Elsevier Science (USA)

Key Words: histone; methylation; germline; C. elegans; X chromosome; XY body; telomere.

INTRODUCTION

In species where females and males have different num-bers of X chromosomes, various dosage compensationmechanisms ensure an approximately equal amount ofX-encoded gene products in both sexes. In mammals, forexample, one of the two X chromosomes in the femalesoma inactivates to form the Barr body, whereas the singleX in the male soma is constitutively active (Avner andHeard, 2001; Willard, 1996). In the round worm, Caeno-rhabditis elegans, transcription from each of the two Xchromosomes in hermaphrodites is reduced to approxi-mately half of that from the single X chromosome in the

1 To whom correspondence should be addressed. Fax: (214) 648-

All rights reserved.

male soma (Chuang et al., 1994; Meyer and Casson, 1986).Although the transcription of X-linked genes is compen-sated in the somatic tissues between the two sexes, thisregulation appears to be reversed in the two germlines inseveral animal species (Henderson, 1964; McKee and Han-del, 1993; Monesi, 1965). In mammals, both X chromo-somes are transcriptionally active in the female germline asmeasured by the incorporation of [3H]-labeled uridine. Onthe contrary, the X and Y chromosomes in the malegermline are transcriptionally silent in similar assays.These silenced chromosomes heterochromatize to form thecytologically distinct XY body (Solari, 1974). The Xist gene,required for female somatic X inactivation, does not appearto play a role in the inactivation of male germline X and Ychromosomes (Marahrens et al., 1997). In addition, under-
acetylation of histone H4, a hallmark characteristic of the1196. E-mail: [email protected].
0012-1606/02 $35.00© 2002 Elsevier Science (USA)

71

inactive X in human and mouse female soma, is not foundassociated with the XY body in the mouse male germline(Armstrong et al., 1997). Therefore, the silencing mecha-nism for the male germline X and Y chromosomes appearsto be distinct from that for female somatic X chromosomes.However, despite our understanding of X-inactivation inmammalian female soma, the mechanism of and reason forsilencing of the male germline X and Y chromosomesremains unclear (Avner and Heard, 2001; Willard, 1996).

Heterochromatization of the male germline X chromo-some is not a mammal-specific phenomenon, having also

been observed in organisms such as the round worm, C.elegans (Goldstein, 1982). C. elegans has no Y chromosome,with two X chromosomes for hermaphrodites and one X formales (Madl and Herman, 1979). Heterochromatization ofC. elegans male germline X chromosomes was described byultrastructural analyses and occurs only in the pachytenestage of meiotic prophase I (Goldstein, 1982). Recent mi-croarray analysis showed that almost no sperm-enrichedgenes are X-linked, whereas oocyte-enriched genes areevenly distributed among autosomes and X chromosomes(Reinke et al., 2000). A similar under-representation of

FIG. 1. Silenced extra-chromosomal arrays contain a high level of Me(K9)H3 but no Me(K4)H3 staining. Representative sections of DAPIand antibody staining of the BS3128 hermaphrodite germline are shown. (A–G) Me(K4)H3 antibody. (H–K) Me(K9)H3 antibody. (A–D, H–K)Pachytene nuclei. (E–G) Diakineses oocytes. Arrows point to extrachromosomal DNA arrays. Scale bar, 1 �m.

72 Reuben and Lin

© 2002 Elsevier Science (USA). All rights reserved.

X-encoded genes amongst adult testis-derived transcriptshas also been observed in humans (Jones et al., 1997).Although an assay with [3H]-labeled uridine has not beenperformed, C. elegans male X chromosomes are likely to betranscriptionally silent based on their heterochromatic na-ture and the microarray result.

In eukaryotes, control of gene expression and the demar-cation of functional domains within chromosomes dependon the higher-order organization of chromatin. A number ofposttranslational modifications, including acetylation andphosphorylation of core histones at their N-terminal tails,have long been linked to changes in chromatin structureand transcriptional activity (see review by Jenuwein andAllis, 2001). Hyperacetylated histones have been most often

associated with open chromatin and transcriptional activa-tion. Phosphorylation of histone H3 at serine 10 has beenlinked to mitotic chromosome condensation. Recent evi-dence suggests that methylation of histones at specific sitesalso plays an important role in chromatin organization andgene regulation. Methylation of histone H3 at lysine 4[hereafter Me(K4)H3] appears to be associated with tran-scriptionally active chromatin (Noma et al., 2001; Strahl etal., 1999). On the contrary, histone H3 methylated at lysine9 [hereafter Me(K9)H3] has been shown to be associatedwith transcriptionally inactive heterochromatin (Bannisteret al., 2001; Lachner et al., 2001; Nakayama et al., 2001).However, unlike acetylation and phosphorylation, histonemethylation does not appear to be a dynamic event, as an

FIG. 2. Me(K4)H3 staining is excluded from the X chromosomes in the pachytene zone of male and hermaphrodite germlines. Single slicesof DAPI (red) and Me(K4)H3 (green) staining from hermaphrodite (A–G) and male (H–N) germlines are shown as low (A, H) or high (B–G,I–N) magnification images. (B–D, I–K) Pachytene nuclei. (E–G) Diakinetic oocyte nuclei. (L–N) Metaphase I nuclei. Arrows indicate Xchromosomes. Scale bar, 1 �m.

73Histone Methylation in C. elegans Germline


enzymatic activity able to specifically remove these methy-lations has not yet been identified. It is likely then thatchanges in chromatin organization via histone methyla-tions will correlate with a more stable change in transcrip-tional activity.

Using antibodies specific to H3 methylated at eitherlysine 4 or lysine 9, we show that the pattern of site-specificH3 methylation is different between X chromosomes andautosomes as well as between germline X chromosomesfrom the two sexes in C. elegans. While autosomes have ahigh level of H3 methylated at lysine 4, the germline Xchromosomes in both sexes are devoid of this modificationin pachytene nuclei. Absence of Me(K4)H3 staining on theX chromosome is transient in the hermaphrodite germline,as staining becomes apparent with further development ofthe germ nuclei into oocytes. On the other hand, theabsence of Me(K4)H3 staining on the male germline Xpersists throughout the germline. Even more dramaticdifferences between the germline X chromosomes in thetwo sexes are observed with the Me(K9)H3 antibody, whichstains the X chromosome only in the male, but not her-maphrodite, germlines. We show, using mutants that aredefective in sex determination, that the association ofMe(K9)H3 with X chromosomes is dependent on the sexualphenotype, but not the karyotype of the animals. We alsoshow that the XY body in mouse spermatocytes contains avery high level of Me(K9)H3, suggesting an evolutionarilyconserved mechanism for silencing X chromosomes in themale germline.

MATERIALS AND METHODS

Strains and Alleles

The N2 Bristol strain was used as the wild-type strain. Thegenetic markers used in this paper are: LGII, tra-2(e1095sd), mnCI;LGV, her-1(e1518); LGX, lon-2(e678). BS3128 contains the let-858-gfp transgene as an extrachromosomal array. All strains werecultured as described by Brenner (1974). Mouse testes were dis-sected from 3- to 4-month-old ICR male mice.

Immunofluorescence

C. elegans gonads were dissected and processed for immunoflu-orescence as described (Detwiler et al., 2001). Mouse testis spreadwas performed as described (Page et al., 1998) with few modifica-tions. After freezing on dry ice, the coverslip was removed and slidewas immersed into 0.3% Triton/PBS twice, followed by PBS threetimes, 5 min each. The slide was then blocked in 3.5% goatserum/0.5% BSA/PBS at room temperature for 30 min. Primaryantibody was done at 4°C overnight and the secondary antibodywas at room temperature for 1 h. The slide was washed with PBSthree times for 5 min each in between incubations and stained withDAPI for 10 min prior to mounting. Mouse testis sectioning andstaging of seminiferous tubules were performed in the pathologycore of UT Southwestern Medical Center. Antibody dilutions usedwere: Me(K4)H3, 1:5000; Me(K9)H3, 1:1500; Alexa488 secondaryantibody (Molecular Probes), 1:500.

Microscopy

Imaging of immunofluorescence was performed by using anOlympus Bmax 60F microscope and a MicroMax-512EBFT CCDcamera from Princeton Instruments as described previously (Hsu etal., 2000). Each photograph was collected as sequential imagesalong the Z-axis with 0.13 microns between adjacent images for atotal of approximately 80 slices per specimen. All images werecollected in gray scale and pseudocolored with a custom program,EditView4D ([email protected]), and projections (maxi-mum intensity) were done with the free program, ImageJ (http://rsb.info.nih.gov/ij).

RESULTS

Silenced, Extra-Chromosomal Transgenes ContainHigh Levels of Me(K9)H3 and Are Devoid ofMe(K4)H3 in the Germline

The C. elegans germline is a tube-like syncitium consist-ing of thousands of germ nuclei. Germ nuclei proliferatemitotically at the distal end of the gonad before enteringmeiosis and progressing through various stages of meioticprophase I. After exiting pachytene, germ nuclei in her-maphrodites continue to develop and cellularize into oo-cytes that are arrested at diakinesis of prophase I (Schedl,1997).

In C. elegans, transgenes introduced via microinjectionconcatemerize to form repetitive extrachromosomal arraysthat are often silenced specifically in the germline (Kelly etal., 1997). We first examined the ability of the Me(K4)H3and Me(K9)H3 antibodies to distinguish transcriptionallyactive or silent chromatin by using a strain, BS3128, carry-ing extrachromosomal transgenes of let-858-gfp. LET-858 isa ubiquitously expressed nuclear protein (Kelly et al., 1997).However, the extrachromosomal let-858-gfp in BS3128 isexpressed only in somatic nuclei and is silenced in thegermline (Kelly et al., 1997). The extrachromosomal trans-genes can be identified as small, isolated DAPI-stainingparticles in the germline nucleus (Figs. 1B, 1E, and 1I). In allgerm nuclei where the extrachromosomal transgenes couldbe unequivocally identified, Me(K4)H3 staining was ex-cluded from the transgenes (Figs. 1A–1G). This was particu-larly clear in nuclei of diakinetic oocytes, a stage whenchromosomes are well separated and transgenes are easilyidentified (Figs. 1E–1G). In dramatic contrast, the Me(K9)H3antibody stains the silenced extrachromosomal arraysstrongly when compared to the rest of the genomic chro-mosomes (Figs. 1H–1K). These results are consistent withrecent findings suggesting that the Me(K4)H3 andMe(K9)H3 antibodies specifically recognize transcription-ally competent and silenced chromatin, respectively16–20.

X Chromosomes of Germline Pachytene Nuclei DoNot Stain with the Me(K4)H3 Antibody

We then examined in detail the staining pattern of theMe(K4)H3 antibody in the adult hermaphrodite germline. In

74 Reuben and Lin


diakinetic oocytes where all chromosomes are well sepa-rated, Me(K4)H3 staining is readily detected on all chromo-somes (Figs. 2E–2G). However, in pachytene nuclei, wherehomologous chromosomes are synapsed and arrayed at theperiphery, we consistently observed that one of the sixchromosome pairs does not display Me(K4)H3 staining(Figs. 2A–2D). We tentatively identified the chromosomepair devoid of Me(K4)H3 staining as the X chromosomepair. X chromosomes are the shortest homolog pair inpachytene (Goldstein, 1982), and they are readily distin-guishable from autosomes when stained with DAPI. Thechromosome pair lacking Me(K4)H3 staining is consis-tently the shortest pair in nuclei, where relative size couldbe unequivocally determined. This observation suggeststhat, in the pachytene region of the hermaphrodite germ-line, both X chromosomes are transcriptionally inactivewhen compared with the autosomes. This difference ap-pears to be transient in the hermaphrodite germline becauseall chromosomes have an equal and high level of Me(K4)H3staining as germ nuclei develop further toward oocytes.

The single X chromosomes in the pachytene region ofmale germlines were also found to lack Me(K4)H3 staining.Male pachytene X chromosomes are heterochromatizedinto DAPI-staining particles that can be readily distin-guished from the well-synapsed autosomes (Figs. 2H and 2I;Goldstein, 1982). In every nucleus where the single Xchromosome could be unequivocally identified, there islittle, if any, Me(K4)H3 staining on these X chromosomes(Figs. 2H–2K). In contrast to the hermaphrodite X chromo-somes, where Me(K4)H3 staining reappears postpachytene,the male X chromosome fails to stain with Me(K4)H3 at anystage during germline development. Figures 2L–2N show acharacteristic metaphase I spermatocyte in which thesingle X chromosome, surrounded by five pairs of auto-somes, is devoid of Me(K4)H3 staining. This observationsuggests that, while inactivation of X chromosomes in thehermaphrodite germline appears transient, the X chromo-some in males remains inactive throughout the germline.

The X Chromosome in Male, but NotHermaphrodite, Germlines Stains Brightly withthe Me(K9)H3 Antibody

Our above observation suggests that germline X chromo-somes in C. elegans are subjected to distinct patterns oftranscriptional regulation in the two sexes. We furtherinvestigated whether this difference could also be reflected

by different levels of methylated histone H3 at lysine 9.Using the Me(K9)H3 antibody, we fail to observe strongstaining of any particular chromosome in the hermaphro-dite germline (Figs. 3A–3D). In dramatic contrast, the singleX chromosome in the male germline stained with theMe(K9)H3 antibody very intensely (Figs. 3E–3H). This stain-ing peaked in the pachytene zone and was absent inspermatids. Under lower magnification, the only stainingdetected was that on the X chromosome (Fig. 3E). Thesedata strongly suggest that X chromosomes in male andhermaphrodite germlines are subjected to different mecha-nisms of transcriptional regulation involving methylationof histone H3 at lysine 9.

In addition to striking staining of male germline Xchromosomes, we detected very sparse and patchy staining,likely heterochromatic regions, throughout the germ nucleiof both sexes. We could ascribe some of this patchy stainingto the telomeric regions of the chromosomes (arrowheads inFigs. 3B–3D). This telomeric staining with the Me(K9)H3antibody became pronounced in prometaphase and meta-phase stages of proliferating early blastomeres, when thechromosomes became highly condensed (Fig. 4). To ourknowledge, enrichment of telomeric Me(K9)H3 staining hasnot been described in other organisms and its function in C.elegans is currently unclear. Nonetheless, HP1, a proteinbelieved to bind to methylated lysine 9 in H3 (Bannister etal., 2001; Jacobs et al., 2001; Lachner et al., 2001), has beenshown to localize to telomeric regions and this telomericlocalization prevents inappropriate chromosome fusion inDrosophila (Fanti et al., 1998).

The X Chromosomal Staining of Me(K9)H3Correlates with the Sexual Phenotype but Not theKaryotype of an Animal

In addition to secondary phenotypic differences, adultmales and hermaphrodites differ in two basic ways: (1)hermaphrodites have two X chromosomes, whereas maleshave only one; and (2) adult hermaphrodites produce oo-cytes, whereas adult males produce sperm. We were able totest which of these differences accounts for the differentialX chromosomal Me(K9)H3 staining in the two sexes of C.elegans by using well-characterized mutants in the sexdetermination pathway.

C. elegans hermaphrodites first produce sperm in the L4stage before switching to producing exclusively oocytes asadults (Schedl, 1997). We examined the Me(K9)H3 staining

FIG. 3. X chromosomes in the male germline stain strongly with the Me(K9)H3 antibody. Single slices of DAPI (red) and Me(K9)H3 (green)staining from the pachytene region of hermaphrodite (A–D) and male (E–H) germlines are shown as low (A, E) or high (B–D, F–H)magnification images. (I) Summary of Me(K9)H3 and Me(K4)H3 staining in male and hermaphrodite germlines. Arrows, X chromosome.Arrowheads, Patchy heterochromatic staining on autosomes. Scale bar, 3 �m in (A) and (E), and 1 �m in the rest.FIG. 4. Me(K9)H3 staining of telomeric regions in developing embryos. Low (A) or high (B–D) magnifications of selected nuclei from a12-cell-stage embryo stained with DAPI and Me(K9)H3 antibody. Images are projected from a stack. Arrowheads point to representative,intense telomeric staining. Scale bar, 5 �m.



76 Reuben and Lin

pattern in the L4 hermaphrodite germline during spermato-genesis and observed no association of staining with anyparticular chromosome (data not shown). This observationsuggested that the Me(K9)H3 staining of male X chromo-somes is not simply a result of or a prerequisite for sper-matogenesis. We then examined Me(K9)H3 staining inmutants affecting the sex-determination pathway whichuncouples the sexual phenotype from karyotype. Loss-of-function mutations in tra-2 result in sex transformation ofXX animals to males that have a fully masculinized somaticgonad and germline (Hodgkin and Brenner, 1977). In 80%(n � 90) of tra-2(e1095sd) XX male animals examined, weobserved many germ nuclei containing a single chromo-some pair, presumably the X chromosomes, that stainedintensely with the Me(K9)H3 antibody (Figs. 5A–5C). Incontrast to mutations in tra-2, mutations in the gene her-1result in XO animals developing into relatively normal,fertile hermaphrodites (Hodgkin and Brenner, 1977). Thesingle X chromosome in her-1(e1518) XO animals appearsas a nonpaired univalent and can be easily recognized. Weobserved that the vast majority of her-1 XO hermaphroditesdo not contain any chromosome pair with intenseMe(K9)H3 staining (Figs. 5E–5G). Approximately 90% ofgonads examined showed no observable staining, whereas10% showed detectable staining (n � 65). These observa-tions suggest that the Me(K9)H3 staining of germline Xchromosomes is not determined by the number of X chro-mosomes, but instead is determined by the sexual pheno-type of the animal.

Interestingly, we observed that the XX chromosomes intra-2(e1095sd) mutant males align at the periphery ofpachytene nuclei, indistinguishable from the XX chromo-some pair in wild-type hermaphrodites pachytene nuclei.On the contrary, the single X chromosome in her-1(e1518)XO mutant hermaphrodite pachytene nuclei formed a tightDAPI staining focus similar to that in wild-type malepachytene nuclei. These observations suggest that the con-densation state of the X chromosome can be dissociatedfrom the Me(K9)H3 staining (see Discussion).

Mouse Male Germline X Chromosome alsoDisplays High Levels of Me(K9)H3

Because the X chromosome in the mammalian malegermline is known to be silenced (Henderson, 1964; McKeeand Handel, 1993; Monesi, 1965), we examined whether themale germline X chromosome in mouse also contains a

high level of Me(K9)H3. The heterochromatized XY body isspecific to late pachytene and diplotene spermatocytes andcan be identified cytologically with DAPI, often as anintense staining region delineated from the rest of thechromatin (arrow in Fig. 6A; Solari, 1974). Other brightDAPI staining regions, which distribute throughout thenucleus, are likely to be the constitutive pericentric hetero-chromatin on the autosomes (arrowheads in Fig. 6A). It hasrecently been shown, using a different antibody, that theconstitutive pericentric heterochromatin as well as the XYbody contain H3 methylated at lysine 9 (Peters et al.,2001b). With our Me(K9)H3 antibody, we also observedMe(K9)H3 staining on the XY body (Figs. 6A–6C). However,the XY body staining level was very high compared with thepericentric heterochromatin in late pachytene spermato-cytes. The temporal pattern of Me(K9)H3 staining in devel-oping spermatocytes was analyzed using sections of tubulesfrom adult male mouse testes (Figs. 6D–6G). We detected ahigh level of XY body staining with the Me(K9)H3 antibodyonly in the late pachytene and diplotene spermatocytes intubules at stages IX–XI. Figure 6 shows a representativesection of a stage X–XI tubule in which many pachytenenuclei contain a single bright spot when stained with theMe(K9)H3 antibody. This result suggests an evolutionarilyconserved mechanism, involving methylated histone H3 atlysine 9, in the transcriptional silencing of male germline Xchromosomes.

DISCUSSION

In this report, we document an interesting developmentalregulation of histone H3 methylation on germline X chro-mosomes. X chromosomes of both the hermaphrodite andmale germlines in C. elegans do not stain for Me(K4)H3 inpachytene nuclei. As hermaphrodite germline X chromo-somes develop past pachytene, they display Me(K4)H3staining, whereas the single male X chromosome remainsnegative for Me(K4)H3 staining throughout the germline. Instriking contrast, the male germline X chromosome con-tains a high level of Me(K9)H3. This association ofMe(K9)H3 with the male germline X chromosome wasfound to be determined by the sexual phenotype instead ofthe karyotype of the animal. The inactive XY body in themouse male germline also contains a high level ofMe(K9)H3, suggesting an evolutionarily conserved mecha-

FIG. 5. The X chromosomal staining of Me(K9)H3 correlates with the sexual phenotype but not the karyotype of an animal.Representative nuclei from the germline pachytene region of wild-type male (D), tra-2 male (A–C), wild-type hermaphrodite (H), and her-1hermaphrodites (E–G) are stained with DAPI (red) and Me(K9)H3 antibody. Arrows point to X chromosomes. Scale bar, 1 �m.FIG. 6. Mouse male germline X chromosomes also display high levels of Me(K9)H3. (A–C) Spread late pachytene mouse spermatocytenuclei stained with DAPI (red) and Me(K9)H3 (green) antibody. (D–G) A testis section (estimated to be stage X–XI) stained with DAPI (D,F) and Me(K9)H3 (E, G). (F, G) High magnification images of boxed areas in (D) and (E). Arrow, XY body. Arrowheads, Pericentricheterochromatin. Bar, 10 �m in (A–C), and 100 �m in (D–E).

78 Reuben and Lin


nism for the silencing of the maternally derived X chromo-some in the male germline.

X Chromosome vs Autosomes in the Germline

Our findings with the Me(K4)H3 antibody suggest that Xchromosomes are transcriptionally inactive compared toautosomes in the pachytene region of the C. elegans germ-line in both sexes. The significance of this difference intranscriptional activity between autosomes and X chromo-some as well as the region of the germline exhibiting thisdifference remains unclear. Based on the genetic character-ization of four mes genes (mes-2, mes-3, mes-4, and mes-6),it has been proposed previously that the X chromosome andautosomes might be subjected to different transcriptionalregulation in the C. elegans germline (Garvin et al., 1998).Mutations in any of these mes genes result in desilencing ofnormally silenced extrachromosomal transgenes in thegermline (Kelly and Fire, 1998). In addition, all mes muta-tions cause maternal-effect sterility (Garvin et al., 1998).This sterility is very likely due to inappropriate desilencingof certain X-linked gene(s) in the germline, because theseverity of the mes mutant phenotype is directly propor-tional to the number of X chromosomes. Consistent withthe notion that X chromosomes are desilenced in mesmutants, we observed Me(K4)H3 staining on all chromo-somes in the germline of animals derived from homozygousmes mutant mothers (R.L., unpublished observation). How-ever, in homozygous mes mutants derived from heterozy-gous mutant mothers, whose germlines are normal andfertile, the Me(K4)H3 staining appears normal (R.L., unpub-lished observation). These observations make it unlikelythat Me(K4)H3 plays a causal role in the Mes phenotype.

X Chromosomes in Males and Hermaphrodites

X chromosomes in both sexes do not stain for Me(K4)H3during pachytene, although the hermaphrodite X chromo-somes do exhibit Me(K4)H3 staining as germ nuclei exitpachytene. In contrast, the X chromosome remains un-stained for Me(K4)H3 throughout the male germline. Theseresults suggest that the hermaphrodite X chromosomesbecome transcriptionally active in diakinesis, whereas themale X remains inactive throughout the germline. Thesedata are consistent with microarray results where an equalrepresentation of oocyte-enriched genes on the X and theautosomes was found, whereas almost no sperm-enrichedgenes were found to be encoded by the X chromosome(Reinke et al., 2000).

We propose that transcriptional activity of X chromo-somes in male and hermaphrodites is subjected to differ-ent mechanisms of regulation. X chromosomes in thepachytene region of the hermaphrodite germline are ini-tially transcriptionally inactive and are marked with un-methylated histone H3. Upon exiting pachytene, X chro-mosomes are activated either by methylation of histone H3at lysine 4 or by a yet unknown mechanism, with concomi-

tant methylation of histone H3 at lysine 4 and presumablyother modifications associated with transcriptionally activechromatin (e.g., acetylation of H3 at lysines 9 and 14; Litt etal., 2001). X chromosomes in male, on the other hand, aremarked by a high level of Me(K9)H3 throughout the germ-line. It has been recently shown that methylation of histoneH3 lysine 4 is inhibited by a preexisting methyl group onlysine 9 and vice versa (Wang et al., 2001). This couldexplain why male germline X chromosomes do not stainwith Me(K4)H3 antibody after exiting pachytene. We pro-pose that the high level of Me(K9)H3 on the male germlineX chromosomes prevents the addition of Me(K4)H3 andconsequently the activation of transcription.

It has been shown that one of the two histone methyl-transferases (HMTases) identified in mice (Suv39h2) isspecifically expressed in the mouse male germline(O’Carroll et al., 2000). It is possible that the C. elegansHMTase responsible for X chromosomal Me(K9)H3 stainingis also expressed exclusively in the male germline. How-ever, Suv39h2 does not appear to be responsible for theMe(K9)H3 staining of the mouse XY body (Peters et al.,2001b), nor is the gene(s) responsible for the germline Xchromosomal staining in C. elegans known. Depletion of aC. elegans gene transcript, C15H11.5, that shares the mostsimilarity in sequence to known HMTases (Aagaard et al.,1999; Rea et al., 2000) with RNA interference, results in aloss of Me(K9)H3 staining in approximately 30–60% of theprogeny from injected animals (R.L., unpublished data).This RNAi result is consistent with C15H11.5 functioningas an HMTase for lysine 9 in vivo. However, to date wehave not been able to detect any in vitro HMTase activityassociated with C15H11.5 (B. Strahl, C. D. Allis, and R.L.,unpublished data). This is not surprising as it has recentlybeen shown that several methyltransferases exhibit noHMTase activity in vitro when expressed in bacteria asrecombinant proteins, but do exhibit HMTase activitywhen assayed in the context of their native complex (Rea etal., 2000; Roguev et al., 2001). Current experiments areaimed to address whether C15H11.5 exhibits HMTaseactivity in worm extracts.

Inactivation of Male Germline X Chromosomes

Because an enzyme activity that removes the methylgroup has not yet been identified, histone methylation hasbeen considered to be a more stable modification. This ideareceives strong support from the recent demonstration thatthe stably inactive X chromosomes in female mammalsoma are marked by Me(K9)H3 (Boggs et al., 2001; Peters etal., 2001a). Silencing of apparently the entire X chromo-some in the male germline with Me(K9)H3 may seemdrastic, however, as active transcription from this chromo-some is required later in developing embryos. Permanentinactivation of the male-derived X chromosome may beavoided because in most animals, the majority of histonesare not transferred from sperm to zygote. Sperm genome inmost animals, in fact, undergoes histone displacement that



removes most histones and replaces them with protaminesduring late stages of spermatogenesis (Loir and Lanneau,1978; and see review by Fuentes-Mascorro et al., 2000).Histone displacement would likely reduce the amount ofmethylated histones carried over from sperm to zygote.Therefore, Me(K9)H3 could well function as an excellentmechanism for male germline-specific silencing. This isconsistent with our observation that the germline in L4hermaphrodites contains no intense Me(K9)H3 staining ofX chromosomes despite undergoing spermatogenesis, be-cause these X chromosomes must remain competent fortranscription as hermaphrodites develop into adults andproduce exclusively oocytes. Although a sperm-specifichistone H1 is expressed during spermatogenesis (Sanicola etal., 1990), none of the antibodies to core histones that wetested (n � 5; unpublished observation) stain spermatids,suggesting that most histones are displaced during sper-matogenesis in C. elegans. However, in C. elegans, thehighly condensed spermatid DNA is not found associatedwith protamines (L’Hernault, 1997), and exactly how his-tone displacement operates is presently unclear.

Silencing of the X chromosome in the male germline hasbeen demonstrated in mammals, and we now show thatthis is likely the case for C. elegans. As this appears anevolutionarily conserved phenomenon, what pressuresmight maintain this conservation? It is unlikely that thisdifference in germline X chromosome transcriptional activ-ity is a form of dosage compensation, as it runs counterin-tuitive to a mechanism for compensating the difference inchromosome numbers. In addition, somatic dosage com-pensation does not take effect until approximately the30-cell stage in embryos in C. elegans (Chuang et al., 1994).

Several possibilities, including ones listed below, havebeen proposed to explain the heterochromatization andsilencing of the XY body in mammals. First, certain geneson the X or Y chromosomes might be detrimental to malemeiosis (Lifschytz and Lindsley, 1972). Second, heterochro-matization could prevent inappropriate pairing and recom-bination between nonhomologous regions of X and Ychromosomes (McKee and Handel, 1993). Third, hetero-chromatization of X and Y chromosomes could preventtriggering a checkpoint for asynapsed chromosomes (Jab-lonka and Lamb, 1988). Gene silencing in this scenariowould be merely a consequence of the heterochromatiza-tion. Interestingly, our characterization with C. eleganstra-2 and her-1 mutants has uncoupled the condensationstate of the chromatin and the level of Me(K9)H3 ongermline X chromosomes. While the level of Me(K9)H3 isdetermined by the sexual phenotype of an animal, thecondensation state of the X chromosome appears to corre-late with the number of X chromosomes. One intriguinginterpretation of our data is that silencing and condensationstate of X chromosomes serve different biological functions.The condensation state correlates well with karyotypes: ahighly condensed state was only observed when a single,unpaired X chromosome is present in the germline. This isin agreement with the second and third hypotheses noted

above that condensation might prevent inappropriate pair-ing between the X chromosome and autosomes or triggeringa checkpoint for asynapsed chromosomes. Silencing of Xchromosomes, on the other hand, is specific to malesregardless of their karyotypes, consistent with the firsthypothesis above that X-encoded genes might be harmful tomale meiosis. The X chromosome of male animals,whether it be in worms or in humans, is always inheritedfrom their mothers. It remains an intriguing possibilitytherefore, that a mechanism has evolved whereby thecontribution of the maternal genome to the biogenesis ofsperm is minimized. Although somewhat speculative, ourdata highlight dramatic developmental changes in lysine 4vs lysine 9 methylation of histone H3 and provide addi-tional supporting evidence for the histone code hypothesis(Strahl and Allis, 2000).

ACKNOWLEDGMENTS

We thank C. David Allis, Leon Avery, Scott Cameron, DennisMckearin, Eric Olson, Scott Robertson, and Jim Waddle for criticalreading of the manuscript. Special thanks to C. David Allis forsharing unpublished information and unpublished reagents beforethey were commercially available. We appreciate Brian Strahl forhis initial efforts in performing the in vitro HMTase assay. Wethank Jim Richardson and the histology core at UT SouthwesternMedical Center for mouse sectioning and staging of testis tubules.We also thank Tim Schedl for strain BS3128 and the C. elegansGenome Consortium for all other strains described in this study.This work is supported by an NIH ROI grant (HD37933).

REFERENCES

Aagaard, L., Laible, G., Selenko, P., Schmid, M., Dorn, R., Schotta,G., Kuhfittig, S., Wolf, A., Lebersorger, A., Singh, P. B., Reuter,G., and Jenuwein, T. (1999). Functional mammalian homologuesof the Drosophila PEV-modifier Su(var)3–9 encode centromere-associated proteins which complex with the heterochromatincomponent M31. EMBO J. 18, 1923–1938.

Armstrong, S. J., Hulten, M. A., Keohane, A. M., and Turner, B. M.(1997). Different strategies of X-inactivation in germinal andsomatic cells: Histone H4 underacetylation does not mark theinactive X chromosome in the mouse male germline. Exp. CellRes. 230, 399–402.

Avner, P., and Heard, E. (2001). X-chromosome inactivation:Counting, choice and initiation. Nat. Rev. Genet. 2, 59–67.

Bannister, A. J., Zegerman, P., Partridge, J. F., Miska, E. A., Thomas,J. O., Allshire, R. C., and Kouzarides, T. (2001). Selective recog-nition of methylated lysine 9 on histone H3 by the HP1 chromodomain. Nature 410, 120–124.

Boggs, B. A., Cheung, P., Heard, E., Spector, D. L., Chinault, A. C.,and Allis, C. D. (2001). Differentially methylated forms ofhistone H3 show unique association patterns with inactivehuman X chromosomes. Nat. Genet. 30, 73–76.

Brenner, S. (1974). The genetics of Caenorhabditis elegans. Genet-ics 77, 71–94.

Chuang, P. T., Albertson, D. G., and Meyer, B. J. (1994). DPY-27: Achromosome condensation protein homolog that regulates C.

80 Reuben and Lin


elegans dosage compensation through association with the Xchromosome. Cell 79, 459–474.

Detwiler, M. R., Reuben, M., Li, X., Rogers, E., and Lin, R. (2001).Two zinc finger proteins, OMA-1 and OMA-2, are redundantlyrequired for oocyte maturation in C. elegans. Dev. Cell 1,187–199.

Fanti, L., Giovinazzo, G., Berloco, M., and Pimpinelli, S. (1998).The heterochromatin protein 1 prevents telomere fusions inDrosophila. Mol. Cell 2, 527–538.

Fuentes-Mascorro, G., Serrano, H., and Rosado, A. (2000). Spermchromatin. Arch. Androl. 45, 215–225.

Garvin, C., Holdeman, R., and Strome, S. (1998). The phenotype ofmes-2, mes-3, mes-4 and mes-6, maternal-effect genes requiredfor survival of the germline in Caenorhabditis elegans, is sensi-tive to chromosome dosage. Genetics 148, 167–185.

Goldstein, P. (1982). The synaptonemal complexes of Caenorhab-ditis elegans: Pachytene karyotype analysis of male and her-maphrodite wild-type and him mutants. Chromosoma 86, 577–593.

Henderson, S. A. (1964). RNA synthesis during male meiosis andspermatogenesis. Chromosoma 15, 345–366.

Hodgkin, J. A., and Brenner, S. (1977). Mutations causing transfor-mation of sexual phenotype in the nematode Caenorhabditiselegans. Genetics 86, 275–287.

Hsu, J. Y., Sun, Z. W., Li, X., Reuben, M., Tatchell, K., Bishop,D. K., Grushcow, J. M., Brame, C. J., Caldwell, J. A., Hunt, D. F.,Lin, R., Smith, M. M., and Allis, C. D. (2000). Mitotic phosphor-ylation of histone H3 is governed by Ipl1/aurora kinase andGlc7/PP1 phosphatase in budding yeast and nematodes. Cell 102,279–291.

Jablonka, E., and Lamb, M. J. (1988). Meiotic pairing constraintsand the activity of sex chromosomes. J. Theor. Biol. 133, 23–36.

Jacobs, S. A., Taverna, S. D., Zhang, Y., Briggs, S. D., Li, J.,Eissenberg, J. C., Allis, C. D., and Khorasanizadeh, S. (2001).Specificity of the HP1 chromo domain for the methylatedN-terminus of histone H3. EMBO J. 20, 5232–5241.

Jenuwein, T., and Allis, C. D. (2001). Translating the histone code.Science 293, 1074–1080.

Jones, M. H., Zhang, Y., Tirosvoutis, K. N., Davey, P. M., Webster,A. R., Walsh, D., Spurr, N. K., and Affara, N. A. (1997). Chromo-somal assignment of 311 sequences transcribed in human adulttestis. Genomics 40, 155–167.

Kelly, W. G., and Fire, A. (1998). Chromatin silencing and themaintenance of a functional germline in Caenorhabditis elegans.Development 125, 2451–2456.

Kelly, W. G., Xu, S., Montgomery, M. K., and Fire, A. (1997).Distinct requirements for somatic and germline expression of agenerally expressed Caernorhabditis elegans gene. Genetics 146,227–238.

Lachner, M., O’Carroll, D., Rea, S., Mechtler, K., and Jenuwein, T.(2001). Methylation of histone H3 lysine 9 creates a binding sitefor HP1 proteins. Nature 410, 116–120.

L’Hernault, S. W. (1997). Spermatogenesis. In :C. elegans II “ (D. L.Riddle, T. Blumenthal, B. J. Meyer, and J. R. Priess, Eds.), pp.271–294. Cold Spring Harbor Laboratory Press, Plainview, NY.

Lifschytz, E., and Lindsley, D. L. (1972). The role of X-chromosomeinactivation during spermatogenesis (Drosophila-allocycly-chromosome evolution-male sterility-dosage compensation).Proc. Natl. Acad. Sci. USA 69, 182–186.

Litt, M. D., Simpson, M., Gaszner, M., Allis, C. D., and Felsenfeld,G. (2001). Correlation between histone lysine methylation and

developmental changes at the chicken beta-globin locus. Science293, 2453–2455.

Loir, M., and Lanneau, M. (1978). Transformation of ram spermatidchromatin. Exp. Cell Res. 115, 231–243.

Madl, J. E., and Herman, R. K. (1979). Polyploids and sex determi-nation in Caenorhabditis elegans. Genetics 93, 393–402.

Marahrens, Y., Panning, B., Dausman, J., Strauss, W., and Jaenisch,R. (1997). Xist-deficient mice are defective in dosage compensa-tion but not spermatogenesis. Genes Dev. 11, 156–166.

McKee, B. D., and Handel, M. A. (1993). Sex chromosomes,recombination, and chromatin conformation. Chromosoma 102,71–80.

Meyer, B. J., and Casson, L. P. (1986). Caenorhabditis eleganscompensates for the difference in X chromosome dosage betweenthe sexes by regulating transcript levels. Cell 47, 871–881.

Monesi, V. (1965). Differential rate of ribonucleic acid synthesis inthe autosomes and sex chromosomes during male meiosis in themouse. Chromosoma 17, 11–21.

Nakayama, J., Rice, J. C., Strahl, B. D., Allis, C. D., and Grewal, S. I.(2001). Role of histone H3 lysine 9 methylation in epigeneticcontrol of heterochromatin assembly. Science 292, 110–113.

Noma, K., Allis, C. D., and Grewal, S. I. (2001). Transitions indistinct histone H3 methylation patterns at the heterochromatindomain boundaries. Science 293, 1150–1155.

O’Carroll, D., Scherthan, H., Peters, A. H., Opravil, S., Haynes, A. R.,Laible, G., Rea, S., Schmid, M., Lebersorger, A., Jerratsch, M.,Sattler, L., Mattei, M. G., Denny, P., Brown, S. D., Schweizer, D.,and Jenuwein T. (2000). Isolation and characterization of Suv39h2,a second histone H3 methyltransferase gene that displays testis-specific expression. Mol. Cell. Biol. 20, 9423–9433.

Page, J., Suja, J. A., Santos, J. L., and Rufas, J. S. (1998). Squashprocedure for protein immunolocalization in meiotic cells. Chro-mosome Res. 6, 639–642.

Peters, A. H., Mermoud, J. E., O’Carroll, D., Pagani, M., Schweizer,D., Brockdorff, N., and Jenuwein, T. (2001a). Histone H3 lysine 9methylation is an epigenetic imprint of facultative heterochro-matin. Nat. Genet. 30, 77–80.

Peters, A. H., O’Carroll, D., Scherthan, H., Mechtler, K., Sauer, S.,Schofer, C., Weipoltshammer, K., Pagani, M., Lachner, M., Kohl-maier, A., Opravil, S., Doyle, M., Sibilia, M., and Jenuwein, T.(2001b). Loss of the suv39h histone methyltransferases impairsMammalian heterochromatin and genome stability. Cell 107,323–337.

Rea, S., Eisenhaber, F., O’Carroll, D., Strahl, B. D., Sun, Z. W., Schmid,M., Opravil, S., Mechtler, K., Ponting, C. P., Allis, C. D., andJenuwein, T. (2000). Regulation of chromatin structure by site-specific histone H3 methyltransferases. Nature 406, 593–599.

Reinke, V., Smith, H. E., Nance, J., Wang, J., Van Doren, C., Begley,R., Jones, S. J., Davis, E. B., Scherer, S., Ward, S., and Kim, S. K.(2000). A global profile of germline gene expression in C. elegans.Mol. Cell 6, 605–616.

Roguev, A., Schaft, D., Shevchenko, A., Pijnappel, W. W., Wilm,M., Aasland, R., and Stewart, A. F. (2001). The Saccharomycescerevisiae Set1 complex includes an Ash2 homologue andmethylates histone 3 lysine 4. EMBO J. 20, 7137–7148.

Sanicola, M., Ward, S., Childs, G., and Emmons, S. W. (1990).Identification of a Caenorhabditis elegans histone H1 genefamily. Characterization of a family member containing anintron and encoding a poly(A)� mRNA. J. Mol. Biol. 212,259–268.

Schedl, T. (1997). Developmental genetics of the germ line. In “C.elegans II ” (D. L. Riddle, T. Blumenthal, B. J. Meyer, and J. R.



Priess, Eds.), pp. 241–270. Cold Spring Harbor Laboratory Press,Plainview, NY.

Solari, A. J. (1974). The behavior of the XY pair in mammals. Int.Rev. Cytol. 38, 273–317.

Strahl, B. D., and Allis, C. D. (2000). The language of covalenthistone modifications. Nature 403, 41–45.

Strahl, B. D., Ohba, R., Cook, R. G., and Allis, C. D. (1999).Methylation of histone H3 at lysine 4 is highly conserved andcorrelates with transcriptionally active nuclei in Tetrahymena.Proc. Natl. Acad. Sci. USA 96, 14967–14972.

Wang, H., Cao, R., Xia, L., Erdjument-Bromage, H., Borchers, C.,Tempst, P., and Zhang, Y. (2001). Purification and functionalcharacterization of a histone H3-lysine 4-specific methyltrans-ferase. Mol. Cell 8, 1207–1217.

Willard, H. F. (1996). X chromosome inactivation, XIST, andpursuit of the X-inactivation center. Cell 86, 5–7.

Received for publication February 18, 2002Accepted February 20, 2002

Published online March 28, 2002

82 Reuben and Lin


germline x chromosomes exhibit contrasting patterns of histone

Documents