Chromosoma

, Volume 114, Issue 2, pp 92–102

Mutant meiotic chromosome core components in mice can cause apparent sexual dimorphic endpoints at prophase or X–Y defective male-specific sterility

Authors

  • Nadine K. Kolas
    • Department of Molecular GeneticsAlbert Einstein College of Medicine
  • Edyta Marcon
    • Department of BiologyYork University
  • Michael A. Crackower
    • Department of Biochemistry and Molecular BiologyMerck Frosst Canada and Co.
  • Christer Höög
    • Department of Cell and Molecular Biology, The Medical Nobel InstituteKarolinska Institute
  • Josef M. Penninger
    • Institute of Molecular Biotechnology of the Austrian Academy of Sciences
  • Barbara Spyropoulos
    • Department of BiologyYork University
    • Department of BiologyYork University
Research Article

DOI: 10.1007/s00412-005-0334-8

Cite this article as:
Kolas, N.K., Marcon, E., Crackower, M.A. et al. Chromosoma (2005) 114: 92. doi:10.1007/s00412-005-0334-8

Abstract

Genetic modifications causing germ cell death during meiotic prophase in the mouse frequently have sexually dimorphic phenotypes where oocytes reach more advanced stages than spermatocytes. To determine to what extent these dimorphisms are due to differences in male versus female meiotic prophase development, we compared meiotic chromosome events in the two sexes in both wild-type and mutant mice. We report the abundance and time course of appearance of structural and recombination-related proteins of fetal oocyte nuclei. Oocytes at successive days post coitus show rapid, synchronous meiotic prophase development compared with the continuous spermatocyte development in adult testis. Consequently, a genetic defect requiring 2–3 days from the onset of prophase to reach arrest registers pachytene as the developmental endpoint in oocytes. Pachytene spermatocytes, on the other hand, which normally accumulate during days 4–10 after the onset of prophase, will be rare, giving the appearance of an earlier endpoint than in oocytes. We conclude that these different logistics create apparent sexually dimorphic endpoints. For more pronounced sexual dimorphisms, we examined meiotic prophase of mice with genetic modifications of meiotic chromosome core components that cause male but not female sterility. The correlations between male sterility and alterations in the organization of the sex chromosome cores and X–Y chromatin may indicate that impaired signals from the XY domain (XY chromosome cores, chromatin, dense body and sex body) may interfere with the progression of the spermatocyte through prophase. Oocytes, in the absence of the X–Y pair, do not suffer such defects.

Introduction

In mice, genetic modifications of meiotic chromosome core components and their associated proteins tend to have sexually dimorphic effects at meiotic prophase. Where the result is cell death at meiotic prophase, the developmental endpoint tends to be at a more advanced prophase stage in oocytes than in spermatocytes (Hunt and Hassold 2002). Disruption of SPO11, which normally generates DNA cuts in early prophase, leads to synaptic defects in early prophase spermatocytes and oocyte arrest at the later diplotene/dictyate stage (Baudat et al. 2000; Di Giacomo et al. 2005). Similarly, deletion of DMC1, which is involved in homology recognition and strand exchange, results in synaptic defects early in male prophase and later in oocytes (Pittman et al. 1998; Di Giacomo et al. 2005). The deletion of the Escherichia coli MutS mismatch repair homologs MSH4 and MSH5 also induces synaptic failure, more severely in spermatocytes than in oocytes (Kneitz et al. 2000; Di Giacomo et al. 2005).

In other cases, there is cell death of the spermatocytes at meiotic prophase but oocytes complete meiotic development. Deletion of the 30–33 kDa SYCP3 structural chromosome core protein results in defective early prophase, infertility in male mice and reduced fertility in females (Yuan et al. 2000, 2002). Mutations in the human SYCP3 protein also result in male sterility (Miyamoto et al. 2003). Deletion of the chromosome core/synaptonemal complex (SC)-associated protein, FKBP6, causes late prophase spermatocyte degeneration and infertility in males but females are mostly fertile (Crackower et al. 2003). Deletion of H2AX and a mutant isoform of the BRCA1 protein affect the X–Y cores/chromatin and result in pachytene arrest in spermatocytes, causing male, but not female sterility (Celeste et al. 2002; Xu et al. 2003). Deletion of Type IA DNA topoisomerase IIIβ, an autosomal and sex chromosome core-associated protein, does not appear to affect prophase but leads to aneuploid meiotic products and a generation-dependent reduction of fertility, more so in males than in females (Kwan et al. 2003).

Although female and male meiotic prophases have similar functions in terms of genetic recombination and chromosome segregation, there are prominent differences in the time course of meiotic prophase events. In the female mouse, prophase is mostly initiated at fetal age 13.5 days post coitus (dpc) and arrests prior to metaphase I 6 days later at birth. Conversely, in males, effective spermatogenesis does not start until puberty and it takes about 10–12 days for spermatocytes to progress through prophase (Clermont and Trott 1969; Roosen-Runge 1962). The level of reciprocal recombination in oocytes is somewhat higher (about 30 per nucleus, Table 1) than in the male (about 25 per spermatocyte, Moens et al. 2002). Levels of recombination-related proteins that are associated with meiotic chromosome cores are higher in oocytes than reported for spermatocytes (Moens et al. 2002). Germ cell proliferation is mostly a one-time fetal event in females (Johnson et al. 2004), whereas it is continuous in males. We develop the hypothesis that the differences in the developmental time course of prophase in males and females provide an adequate explanation for the apparent dimorphic arrests for early prophase defects.

To account for the differential fertility effects in mutant SYCP3 and FKBP6 mice (Yuan et al. 2000; Crackower et al. 2003), we compared meiotic prophase development in wild-type and mutant mice and report altered developmental patterns in the mutants. The results suggest that these mutations affect X–Y cores/chromatin and the sex body and that these impairments may be the cause of early spermatocyte death, whereas oocytes, which lack an X–Y sex body, are exempt from those effects. Reported male sterility in mutated H2AX and BRCA1 with X–Y cores/chromatin involvement (Celeste et al. 2002; Xu et al. 2003) may lend support to this presumptive relationship.

Materials and methods

Meiotic cell preparations

Spermatocytes for immunocytology were obtained from adult wild-type CD-1 or mutant mice. Single decapsulated testes were macerated in MEM and tubule fragments were allowed to settle in 1 ml Dulbecco’s Modified Eagles Medium (DMEM). The suspended testicular cells were transferred and pelleted at low speed and resuspended. One microliter of dilute cell suspension (1.5×106/ml) was added to a 20 μl drop of hypotonic salt solution (100 mM NaCl, pH 7.5) on each of 12 wells of 1% albumin-coated multiwell slides. Nuclei were left to settle for 20 min and were then fixed for 3 min in 2% paraformaldehyde with 0.03% SDS and 3 min without SDS. After three 1 min washes in 0.4% Kodak Photo-Flo 200, pH 8, the nuclei were air-dried. For electron microscopy, samples of suspended cells were added to 50 μl drops of hypotonic salt solution on plastic-coated slides. To obtain greater spreading of nuclear contents, cells were surface spread on a water bath of hypotonic salt solution and picked up on plastic-coated multiwell slides and fixed. Multiwell slides were used to test up to six different primary antibody treatments with two secondary antibody treatments each on a single slide. For electron microscopy, plain plastic-coated slides were touched to the hypotonic salt-water surface, fixed and washed. Oocytes were obtained from fetal ovaries from 13.5 to 19.5 dpc and up to 2.5 days post partum. Single ovaries were macerated in 20 μl MEM and 2 μl samples of suspended cells distributed as in the spermatocyte protocol.

For the TUNEL assay we used the In Situ Cell Death Detection Kit (Roche) according to the manufacturer’s instructions. The reaction mixture was applied to the spread oocytes after the chromosome cores were visualized with meiosis-specific primary antibodies (anti-SYCP1 and anti-SYCP3) and secondary antibodies. As a positive control, the cells were treated with 0.1 μg/ml of DNase in DMEM for 10 min to induce DNA breaks. As a negative control, the fluorescein isothiocyanate (FITC)-dUTP was added without terminal deoxynucleotidyl transferase (TdT).

For electron microscopy (EM) sections, seminiferous tubules were fixed for 1 h in 2% glutaraldehyde in s-collidine buffer, pH 7.4, and post-fixed for 1 h in 4% osmium tetroxide in buffer. Following dehydration and propylene oxide infiltration, the tubules were embedded in Epon. Thin sections were stained with uranyl acetate and lead citrate.

Immunocytology

For antibody treatment, the slides were washed with constant agitation for 10 min each in PBS with 0.4% Photo-Flo 200, PBS with 0.05% Triton X-100, and PBS with 10% antibody dilution buffer (ADB) for blocking [ADB is 10% goat serum, 3% bovine serum albumin (BSA), 0.05% Triton X-100, in PBS]. In the case of anti-goat antibodies, either donkey serum or egg albumin was used instead of goat serum and BSA. After wiping dry the areas between the wells, individual wells were treated with 10–20 μl of the same or different primary or secondary antibodies. Incubation with primary antibody for either 2 h or overnight at room temperature in a moist chamber was followed by the same washes and fluorochrome-conjugated secondary antibody for 1 h at 37°C. After washes, the slides were dipped in water with 0.4% Photo-Flo to remove salts and after thorough air drying, ProLong Antifade mounting agent (Molecular Probes, Eugene, Ore.), with or without 4 ng/ml of 4,6-diamidino-2-phenylindole (DAPI) for fluorescent DNA staining, was added along with a coverslip.

The same protocols were used to obtain meiotic prophase cells from SYCP3-deleted mice, (Yuan et al. 2000) and FKBP6-deleted mice, (Crackower et al. 2003). Numbers of foci were counted in photographic prints of individual nuclei or on the computer screen of digitally recorded nuclei. The numbers are shown in Table 1 and a summary version of representative numbers, prepared using the CorelDraw 10 graphic utility, is show in graphic form in Fig. 1 . Similarly, the numbers of cells in given meiotic prophase stages are shown in graphic form in Fig. 4.
Table 1

Distribution of meiotic protein foci in oocyte nuclei through prophase I (N/A not applicable)

Numbers of RAD51/DMC1 foci in oocyte nuclei through prophase I

 Stage

Number of oocytes

Range

Mean

Leptotene

Early

5

10–120

53

Mid

7

200–300

261

Late

12

230–420

283

Zygotene

Early

5

280–250

273

Mid

5

270–250

261

Late

4

290–190

236

Pachytene

Early

6

240–110

160

Mid

9

190–40

100

Late

Several

Sporadic

0

Numbers of RPA foci in oocyte nuclei through prophase I

 Stage

Number of oocytes

Range

Mean

Leptotene

Early

1

N/A

0

Mid

1

N/A

40

Late

4

330–460

380

Zygotene

Early

2

420–370

394

Mid

2

380–350

366

Late

2

300–280

290

Pachytene

Early

6

270–200

238

Mid

5

170–110

145

Late

6

100–40

72

Numbers of BLM foci in oocyte nuclei through prophase I

 Stage

Number of oocytes

Range

Mean

Zygotene

Mid

6

50–145

117

Late

7

150–240

175

Pachytene

Early

8

245–280–220

260

Mid

7

210–160

182

Late

8

160–125

139

Diplotene

Early

9

125–85

109

Mid

12

75–50

69

Numbers of MLH1 foci in oocyte nuclei through prophase I

 Stage

Number of oocytes

Range

Mean

Pachytene

Mid

16

22–30

25

Late

9

33–23

28

Diplotene

Early

7

34–17

26

Mid

13

28–16

22

Late

9

21–13

17

Fig. 1

Representative nuclei from Table 1 are shown in graphic form to illustrate the numbers of synaptonemal complex (SC)-associated foci, RAD51/DMC1, RPA, BLM, MLH1 at sequential stages of mouse oocyte meiotic prophase. The structural protein, SYCP3, of the lateral element and SYCP1 of the medial transverse filaments are recorded as percent completion per nucleus at the first part of prophase and percent loss at the last part of prophase. From Fig. 4, it is evident that leptotene and zygotene last about 1 day but pachytene lasts 2 days

Antibodies

Antibodies against meiotic chromosome cores and SCs were generated in rabbits and mice against whole hamster SCs, against hamster 111 kDa synaptic SYN1 protein, or the 30 kDa chromosome core protein, COR1 (Dobson et al. 1994). The current terminology for these proteins from mice is SYCP1 and SYCP3, respectively. The corresponding antibodies had previously been prepared in mice against rat SCs and proteins (Heyting et al. 1987). The mouse DMC1 and RAD51 full-length proteins were over-expressed in E. coli using the pET3 expression system (Habu et al. 1996). The (His)6×-tagged proteins were purified on a Ni-NTA agarose column (Qiagen) and injected into mice and rabbits (Tarsounas et al. 1999). Rabbit antibody to trimeric RPA (replication protein A) was a gift from C.J. Ingles (He et al. 1995). We also generated the antibody in our laboratory from the same construct in rabbits and mice. Rabbit anti-BLM (Bloom’s mutated) serum was provided by R. Freire (Unidad de Investigacion, Tenerife, Spain) (Moens et al. 2000) and prepared by us in mice against proteins from the same construct. Anti-MLH1 (MUTL homolog; Edelmann et al. 1996) was obtained commercially (Pharmingen). Production of antibody to TOPBP1 has been reported by Makiniemi et al. (2001) and the characterization at meiosis by Reini et al. (2004) and by Perera et al. (2004). We generated antibody against amino acids 369–575 of TOPBP1 in rabbit “Mini”. The antibody was verified by immunoblot (Fig. 3, insert), by immunofluorescence of leptotene/zygotene foci, and by intense fluorescence of X–Y chromosomes, the sex body and the staining of laggards. Production and characterization of antibody to FKBP6 has been reported by Crackower et al. (2003). Centromeres were labeled with human CREST serum, a gift from S. Varmuza, University of Toronto (Dobson et al. 1994). C-20 antibody against BRCA1 (Santa Cruz Biotechnology) was a gift from T.F. Lane (UCLA). Antibodies against cohesin SMC3 have been characterized previously (James et al. 2002; Eijpe et al. 2000) and were gifts from R. Jessberger (Technical University Dresden) and K. Yokomori (UCI). Fluorochrome-conjugated secondary antibodies were purchased from Cedarlane Laboratories.

Staging

Meiotic prophase stages of the mouse oocytes were based on the traditional parameters of age of the fetus, combined with the nuclear prophase characteristics of centromere distribution and the degree of chromosome core synapsis (Fig. 1, black lines). Once a reliable sequence of stages was established, secondarily the derived parameters of immunofluorescent protein foci were used for more precise definition of stages. The earliest meiotic prophases have 40 unpaired centromeres that occur in two to four clusters and short chromosome core segments identified with antibodies to chromosome core component SYCP3 or SMC3. Subsequently, the ongoing synapsis of pairs of cores was detected by the presence of the SYCP1 protein. The SYCP1 segments extend until the homologous autosomal cores are synapsed along their entire length. At the late stages of synapsis, the centromeres also synapse so that the numbers decline from 40 single to 20 double centromeres. Following the fully synapsed pachytene stage, the cores initiate repulsion, which is recognized by the separation of the cores and the shortening of SYCP1 segments. Eventually, there are 20 bivalents each with one or two chiasmata. The cores lose their continuity and the centromeres form two or three clusters in the female-specific arrested dictyotene stage of meiotic prophase.

High levels of RAD51, DMC1 and RPA are characteristic of early prophase stages while MLH1 foci mark mid to late pachytene and early diplotene. These additional parameters assist in classifying stages with morphological similarities such as zygotene and diplotene. They can also be used to differentiate unpaired cores at zygotene from the laggards at pachytene. In addition, TOPBP1-staining of laggards can prevent misclassification of pachytene stages with unpaired core segments as zygotene. This methodology is the same as that used for staging of spermatocytes in Moens et al. (2002).

Results

Immunofluorescence protein patterns of oocyte meiotic prophase

The patterns of chromosome core/SC proteins, SYCP1 and SYCP3, and core/SC-associated recombinases, RAD51/DMC1, replication protein A, RPA, Rec Q helicase, BLM, and MUTL homolog, MLH1, were determined by immunofluorescence analysis of some 200 wild-type oocyte nuclei in sequential developmental stages of meiotic prophase (Table 1). The data are illustrated in a simplified graph in Fig. 1. Selected points from Table 1 indicate the first appearance of protein species, the highest numbers of foci per nucleus observed, and the declining numbers during prophase development. The points are connected by curves to illustrate the continuity across meiotic prophase stages. Along the X-axis, the duration of the leptotene, zygotene and diplotene stages is approximately 1 day each while pachytene lasts about 2 days (see prophase progression in Fig. 4).

During oocyte leptotene, RAD51/DMC1 foci become associated with the newly formed core segments, increasing in number from 0 to about 420 foci, well in excess of the numbers recorded for spermatocytes (Moens et al. 2002). Subsequently RPA and BLM protein become associated with the cores/SCs (Fig. 2a). The numbers of foci decline during zygotene, pachytene and diplotene. One or two MLH1 foci per SC are evident from early pachytene into diplotene. Different from male prophase is the appearance of oocyte MLH1 foci when RPA foci are still numerous (Fig. 2b) while in the spermatocyte, there are only a few RPA foci left when the MLH1 foci appear (see summary in Fig. 9). In oocytes as well as spermatocytes, TOPBP1 foci are present on leptotene and zygotene cores but only few of these persist into pachytene. At pachytene, TOPBP1 is a pronounced component of the unpaired X–Y cores and the sex chromatin in the male (Fig. 2c,d) and is associated with cores that have failed to synapse at pachytene in oocytes (Fig. 3) and in spermatocytes (Reini et al. 2004; Perera et al. 2004). We use this characteristic to compare the frequency of spermatocyte nuclei with incomplete synapsis at pachytene [2 in 231 nuclei (<1%)] versus oocyte nuclei with “laggards” [12 in 137 nuclei, (9%), 17.5 dpc fetuses]. Spermatocytes were scored as early pachytene as defined by the presence of TOPBP1-positive X–Y cores (Fig. 2c) prior to the appearance of TOPBP1-positive sex bodies (Fig. 2d). Similarly, the oocytes are at early pachytene on 17.5 dpc (Fig. 4). These asynaptic segments apparently do not cause cell death during prophase because we did not observe degenerating meiotic prophase nuclei with SCs by the TUNEL assay or by the presence of prophase nuclei with condensed chromatin.
Fig. 2a–d

Organization of foci. a RPA becomes a component of the early RAD51/DMC1 foci, the putative sites of double-strand breaks. Subsequently, BLM protein also becomes a component of the foci as demonstrated by the consistent co-location of RPA (yellow-green fluorescence) and BLM (slightly displaced red fluorescence). b In oocyte pachytene nuclei, the MLH1 foci (green) appear while there is still an abundance of RPA foci (red) associated with the synaptonemal complex (SC). c TOPBP1 is a major component of the unpaired regions of X–Y chromosome cores at the pachytene stage of meiotic prophase (rhodamine red). SCs green; centromeres (cen) red. Bar represents 5 μm. d A more advanced pachytene stage, the X and Y cores as well as the sex body, SB, are intensely stained by anti-TOPBP1 (FITC). SCs yellow (rhodamine)

Fig. 3

TOPBP1 (red) is present in association with chromosome cores that have failed to complete synapsis at the time that all other chromosomes have completed synaptonemal complex formation (SC green). Pachytene oocyte nuclei (one shown here) were observed to have a higher incidence of such “laggards” relative to spermatocyte nuclei. Insert Immunoblot of rabbit antibody against the 180 kDa TOPBP1

Fig. 4

The meiotic prophase stages of fetal oocytes at successive days were determined by using the unique profiles of foci and structural proteins at sequential meiotic prophase stages (Table 1; Fig. 1). Meiotic development takes place from 14.5 days post coitus (dpc) to birth, which occurs at about 19.5 dpc, altogether 5–6 days. There appears to be an arrest at early pachytene on days 16.5 and 17.5

Each meiotic prophase stage has a unique profile of SCs and SC-associated proteins (Fig. 1), which enables us to construct a time course of oocyte meiotic prophase development. The classification of oocytes from fetuses at consecutive days post coitus is summarized in Fig. 4. At 14.5 dpc, most oocytes are in early prophase and some have advanced to the zygotene stage. 15.5 dpc fetuses have a rather wide range of stages from zygotene to early pachytene. On days 16.5 and 17.5, the oocytes accumulate in the early pachytene stage. On day 18.5, the oocytes progress into mid-late pachytene and at day 19.5 and at birth, there is a rapid exit into diplotene and dictyotene.

Chromosome core modifications in SYCP3−/− mice

SYCP3−/− mice lack the 30/33 kDa protein and, coincidentally, the 190 kDa chromosome-core component, SYCP2. However, the cohesins and core/SC-associated proteins are still present (Yuan et al. 2000; Pelttari et al. 2001). The cores, which can be visualized by antibodies to cohesin SMC3 (Figs. 5 and 6), are greatly lengthened and are aligned homologously but with limited synapsis as defined by short stretches of SYCP1 protein (Fig. 5) (Kolas et al. 2004). Antibody to TOPBP1 fails to reveal the unpaired X–Y cores or the X–Y chromatin (Fig. 5c) that anti-TOPBP1 recognizes in wild-type spermatocytes (Fig. 2c,d). Anti-γH2AX recognizes X–Y chromatin in the few spermatocytes that reach the pachytene stage (Fig. 5e) but the stained chromatin lacks the compaction seen in the wild-type sex body (Fig. 5d). The ultrastructurally defined sex bodies, SB, in wild-type spermatocytes (Fig. 5d) are not well developed in SYCP3−/− spermatocytes (Fig. 5e). Instead, the X–Y chromatin is located more towards the interior of the nucleus and it lacks the peripheral patches of dense chromatin seen in the sex body of wild-type spermatocytes (Fig. 5d). In the serial sections of the SYCP3−/− nucleus in Fig. 5e, there is an SC of the pseudo-autosomal X–Y region, PAR, that is attached to the nuclear envelope. There are short stretches of normal looking SCs. The unpaired autosomal cores at pachytene (Fig. 5b,g) that would be recognized by TOPBP1 in wild-type mice are not so recognized in the mutant mice (Fig. 5c,h) (scored in 50 spermatocyte and 50 oocyte pachytene nuclei). In the SYCP3−/− spermatocytes, the short SC-like segments range from 1 to 18 μm with averages of between 4 and 6 μm per nucleus (n=5) (Fig. 5a,b). In the SYCP3−/− oocytes, the synapsed segments are much longer, ranging from 3 to 30 μm with means of 4 to 22 μm per nucleus (n=7) (Fig. 5f,g). Even the longer SC-like segments in the oocyte do not extend the entire length of the chromosomes and unpaired cores are present in all pachytene oocytes (Fig. 5g,h, core). The SC segments in oocytes are capable of supporting MLH1 foci (Fig. 5g) even though the synapsis is incomplete. No MLH1 foci are observed in association with the short SCs in spermatocytes, possibly because these foci normally appear in late prophase (Fig. 9) past the time of SYCP3−/− spermatocyte arrest. In wild-type mice, spermatocyte SC lengths range from 3 to 20 μm (average=9 μm) and oocyte SC lengths from 10 to 30 μm (average around 20 μm).
Fig. 5a–h

Meiotic prophase effects of SYCP3−/− deletion. a Spermatocyte early prophase has many elongated cores (green) that are homologously aligned with short synaptic segments indicated by SYCP1 staining (yellow). b The synaptic segments are of variable but short length and will not acquire MLH1 foci. Most of the cores (core) remain unpaired. c The same nucleus as in b to demonstrate that the unpaired cores do not accumulate TOPBP1 in the mutant. The arrow marks a contaminant that is visible in b and c and serves as a point of reference. d Electron micrograph of a wild-type seminiferous tubule cross-section with three nuclei to demonstrate the morphology of sex bodies (SB). The SBs are appressed to the nuclear envelope and they have fine-grained chromatin with internal and peripheral dense chromatin segments. The insert demonstrates the compact γH2AX-positive X–Y chromatin. e Higher EM magnification of an SYCP3−/− spermatocyte to demonstrate that the fine-grained X–Y chromatin is not associated with the nuclear envelope and lacks dense chromatin segments. In the serial EM sections of the nucleus, the SC of the pseudo-autosomal region (PAR) is attached to the nuclear envelope. The insert also demonstrates the more dispersed γH2AX-positive X–Y chromatin. In the interior of the nucleus there are a few short SC segments (SC). f An SYCP3−/− early prophase oocyte nucleus in zygotene judging by the few synapsed cores and the clustering of most unpaired centromeres (cen, FITC). The numbers and localization of the 500 RPA foci (rhodamine, red) to unpaired cores and the SCs appear to be unaffected by the SYCP3−/− mutation. g At the pachytene stage, the synapsed segments (SC, red) are longer than in the spermatocytes and these contain MLH1 foci (MLH1, green). However, synapsis is incomplete and there are abundant unpaired cores (core). h Immunostaining with anti-TOPBP1 (FITC) does not produce the bright immunofluorescence of laggards that is seen in wild-type oocytes as in Fig. 3

Fig. 6a,b

Electron microscope images of chromosome cores in normal and SYCP3−/− mice visualized with antibody to the cohesin, SMC3, and to the central element filaments with anti-SYCP1. a SC of the wild type with parallel aligned lateral elements, LE (SMC3, 10 nm gold grains), and central element, CE (SYCP1, 5 nm) mostly in the central region of the SC. b SYCP3−/− oocytes have some stretches of SC but the lateral elements (10 nm grains) intermittently lack proper alignment and some of the SYCP1 (5 nm) is anomalously located at the periphery of the SC. Not shown, spermatocyte SC-like structure with yet more disorganized immunogold organization

Male-sterile phenotype of a BRCA1 mutation

Males homozygous for exon 11-deleted BRCA1(11/11), p53+/−, fail to complete meiosis but females are fertile (Xu et al. 2003). In wild-type males and females, BRCA1 protein is detected as foci in association with the unpaired cores of zygotene chromosomes (Fig. 7a). Later, at pachytene, these foci are no longer present. Instead, antibody to BRCA1 protein has a high affinity for the unsynapsed segments of the X and Y chromosome cores (Fig. 7b) (Fernandez-Capetillo et al. 2003; Turner et al. 2004) but not the synapsed X–X pair in oocytes. Apparently BRCA1 protein is a major component of the X and Y cores at pachytene and spermatocyte death may be a consequence of sex body impairment by the BRCA1 mutation.
Fig. 7a,b

SC/core-associated BRCA1 protein of wild-type male. a At zygotene of early prophase, the unpaired cores of partially synapsed chromosomes (green) are recognized by anti-BRCA1 antibody (red). b At pachytene, the BRCA1 protein appears to be a component of the unpaired X and Y cores (red). There is no pronounced staining of the sex body. Autosomal SCs are green

Male sterility in FKBP6−/−

The FKBP6 antigen is a component of autosome and sex chromosome cores (Fig. 8a,b). It is the only antigen currently known that is a component of the X chromosome-associated dense body, DB, (Fig. 8a). The DB is normally detected in mammalian spermatocytes by electron microscopy as a dense-staining round body that at first is attached to the X chromosome and later is either associated with the sex body or lies elsewhere in the nucleus. Deletion of the FKBP6 protein results in a defective male meiotic prophase phenotype and sterility, whereas the oocytes are less affected and females are fertile (Crackower et al. 2003). Notably, the pachytene spermatocytes fail to develop a sex body (Fig. 8c) of the type found in wild-type spermatocytes (Fig. 5d). Anti-γH2AX stains the X–Y cores and chromatin but it is disperse, lacking the compact appearance of a wild-type sex body (Fig. 5d). Anti-TOPBP1 recognizes the X–Y chromatin and TOPBP1-defined incomplete synapsis is more extensive (Fig. 8d, asterisks) and more frequent than in wild-type spermatocytes. In about 30% of the pachytene spermatocytes (124 nuclei) the X-chromosome core forms auto-synaptic loops that contain the synaptic protein SYCP1 (Fig. 8e–g). Even though the loop results from non-homologous synapsis, the X-core of the loop is no longer recognized by anti-TOPBP1. In 120 wild-type FKBP6+/+ spermatocytes, there were three nuclei that had a small, possibly autosynaptic, X loop. It appears that FKBP6 is required for normal X–Y chromosome behavior and sex body development.
Fig. 8a–g

FKBP6−/− meiotic prophase effects. a In the wild-type male, FKBP6 is a component of SC/cores (green) as well as the X chromosome-associated dense body, DB, which is known from electron microscopy of mammalian spermatocytes. FKBP6 is the first identified component of the DB. b Anti-SYCP3 and anti-centromere antibodies (red) do not stain the DB. c An electron micrograph of five FKBP6−/− pachytene spermatocytes in a cross-sectioned seminiferous tubule demonstrates that the wild-type sex body as in Fig. 5d is lacking in FKBP6−/− but fine-grained X–Y chromatin is aligned through the nucleus. The insert shows that the γH2AX-positive X–Y chromatin is similarly extended rather than compacted. d FKBP6−/− spermatocytes frequently have incompletely synapsed autosomal cores (laggards) that are recognized by anti-TOPBP1 antibody (green, asterisks). e, f, g FKBP6−/− spermatocytes have impaired X chromosome behavior with one or more auto-synapsed loops (SYCP3 green, SYCP1 red, PAR pseudo-autosomal region, cen centromere)

Discussion

Sex-specific phenotypes have been reported for mutants of chromosome core components that affect fertility. Hunt and Hassold (2002) summarize 12 mutations with sexually dimorphic meiotic phenotypes. A number of these, SPO11, DMC1, MSH4 and MSH5, cause spermatocyte death at early meiotic prophase (zygotene) while oocytes develop further into pachytene and diplotene. Other mutations, SYCP3, and more recently FKBP6, BRCA1 and H2AX, result in spermatocyte death but females are fertile. Hunt and Hassold conclude that “faced with adversity, oogenesis is more robust than spermatogenesis”. Differences in meiotic prophase timing are considered to be a possible minor source of the observed dimorphism but they note that the male–female differences in mouse mutants are real.

Our observations suggest that male sterility versus female fertility is indeed the result of real differences due to the effects of the mutations on X–Y chromosome organization in the male but not in the female. The dimorphic prophase cell arrest/deaths, on the other hand, is only apparent rather than real as a result of differential timing of meiotic prophase in the two sexes. Hunt and Hassold’s Fig. 1 gives a single time frame of meiotic prophase events in males and females that results in a pronounced disparity in the time of cell death between the two sexes. Figure 9 of this report juxtaposes the prophase time course of the two sexes, demonstrating that for a given meiotic prophase defect, oocytes will appear to reach a more advanced developmental stage than the spermatocytes.
Fig. 9

Summary of time course of core-associated proteins in days from the start of meiotic prophase (X-axis) in female (solid lines, this report) versus male prophase (dotted lines, from Moens et al. 2002, the maximum number of RAD51/DMC1 and RPA foci is about 280–300) in terms of numbers of foci per nucleus (Y-axis). The progression in terms of prophase stages is noted along the top of the graph. The oocytes reach the pachytene stage synchronously in 3–4 days after the start of prophase while spermatocytes accumulate pachytene stages from 3 to 11 days after the start of prophase. If the effects of a mutation causes germ cell breakdown some 4 days into prophase, then the oocytes will all have reached the pachytene stage while pachytene spermatocytes fail to accumulate. As a result, it will appear that the spermatocytes have an earlier checkpoint than the oocytes

Mutant ATM, SPO11, DMC1, MSH4 and other proteins lead to defects in meiotic prophase chromosome synapsis and eventually cell degeneration and death. There do not seem to be precise checkpoints. Some 4% of FKBP6−/− spermatocytes reach mid pachytene as defined by the acquisition of testis-specific histone H1 while in the wild type, 16% of the pachytene cells are H1t positive (Crackower et al. 2003). Similarly, a survey of some 500 ATM −/− spermatocytes (P.B.M., personal observations) demonstrates mostly leptotene and zygotene stages but also 3.5% of cell in pachytene (compared with 80% in wild type). There is a shift in the distribution of stages but not a single point of arrest. Mutant cells accumulate pleiotropic disorders over time and eventually arrest/die, in a process somewhat akin to the effects of a genetic disease or aging. If detrimental defects of a mutation set in on day 1 of meiotic prophase and require 2 to 3 days to develop a specific phenotype, the synchronous development of oocytes (Figs. 4 and 9) will cause most oocytes to arrest and later die at the pachytene or diplotene/dictyate stage and there will be few cells in zygotene stages. Oocytes that reach the dictyate stage decline in numbers after birth in SPO11−/− while ATM−/−, DMC1−/− and MSH5−/− oocytes degenerate somewhat earlier. Since SPO11 is epistatic to these other genes, Di Giacomo et al. (2005) attributed the earlier losses to the effect of unrepaired DNA cuts in a wild-type SPO11 background.

In the same time frame, the non-synchronous spermatocytes will fail to accumulate pachytene stages of days 4–10, giving an apparent zygotene checkpoint (Figs. 4 and 9). Thus the observed apparent sexual dimorphism in these cases may not require an explanation in terms of differential regulatory mechanisms.

On the other hand, sexually dimorphic phenotypes of mutations that result in spermatocyte death but have a less severe effect on oocyte survival and female fertility probably result from differential regulatory mechanisms. In several such cases, the X and Y sex chromosomes appear to be involved in male sterility of humans and laboratory mice. Crosses between 57Bl/6 and Mus spretus yield sterile males with asynapsed X and Y chromosomes while female offspring are fertile (Matsuda et al. 1992). Mice and human males with aneuploid sex chromosomes, XYY, XSxraO and others, have levels of cell losses at metaphase I and metaphase II that correlate with the synaptic failure of the sex chromosomes (Rodriguez and Burgoyne 2000, 2001; Solari and Rey Valzacchi 1997; Yogev et al. 2002).

In wild-type mice, SYCP3 and FKBP6 are chromosome core components of both oocytes and spermatocytes. Mutations cause male but not female sterility. The sex body fails to develop normally in SYCP3−/− (Fig. 5e) and in FKBP6−/− mice (Fig. 8c). In addition, the SYCP3−/− sex chromosomes do not accumulate TOPBP1 on the cores (Fig. 5a) or in the sex body (Fig. 5c) as is the rule in wild-type males (Fig. 2c,d). The lack of TOPBP1 recognition of laggards is common to spermatocytes and oocytes and it is therefore not likely to be the deciding factor for sexual dimorphism. The FKBP6−/− X-associated DB lacks its FKBP6 component (Fig. 8a,b) and 30% of the spermatocyte pachytene nuclei have an X chromosome core that undergoes autosynapsis (Fig. 8e–g) while wild-type males rarely have X-loops. To what extent the DB is a necessary requirement for normal sex chromosome development is not known at present but its regular occurrence in mammalian spermatocytes (Dresser and Moses 1980) suggests a functional involvement. It is unlikely that the several defects, the lack of a sex body, the incomplete DB and the X chromosome loops, would not affect male meiotic development.

The BRCA1 protein appears as foci associated with unpaired chromosome cores at zygotene in both sexes (Fig. 7a). Most of the foci do not persist into pachytene. Instead, the unpaired segments of the X and Y chromosomes acquire a BRCA1 component (Fig. 7b). In the wild-type males, the sex body accumulates histone H2AX, which does not occur in BRCA(11/11)p53+/− (Xu et al. 2003). The absence of γH2AX in H2AX−/− male mice results in male sterility (Celeste et al. 2002). In both these mutant mice, male sterility correlates with the failure of the X–Y chromatin to develop its normal association with γH2AX.

These investigations reveal two possible sources of sexually dimorphic effects of mutations of meiotic chromosome core components. As a result of the mutations, where oocytes and spermatocytes do not complete meiotic prophase as in mutated SPO11, DMC1, MSH4 and MSH5, the oocytes appear to reach a more advanced developmental stage. We attribute the apparent dimorphism to the rapid and synchronous prophase development of the oocytes relative to the asynchronous and slower development of the spermatocytes (Fig. 9). In the cases of chromosome core mutations that result in spermatocyte death but oocyte survival, SYCP3, FKBP6, BRCA1, H2AX, we see a correlation with failure of the X–Y cores/chromatin/sex body development that affects spermatocytes but not oocytes.

Acknowledgements

We thank Karen Rethoret for assistance with electron microscopy, Rolf Jessberger and K. Yokomori for antibodies to cohesins. The research was supported by a Discovery Grant from NSERC to P.B.M. and a studentship to E.M.

Copyright information

© Springer-Verlag 2005