Gene Expression Patterns 4 (2004) 495–504 www.elsevier.com/locate/modgep
Changes in the expression and localization of cohesin subunits during meiosis in a non-mammalian vertebrate, the medaka fish Toshiharu Iwai, Jibak Lee1, Atsushi Yoshii, Takehiro Yokota, Koichi Mita, Masakane Yamashita* Laboratory of Molecular and Cellular Interactions, Division of Biological Sciences, Graduate School of Science, Hokkaido University, Sapporo 060-0810, Japan Received 10 September 2003; received in revised form 1 March 2004; accepted 8 March 2004 Available online 17 April 2004
Abstract Until the onset of anaphase, sister chromatids are bound to each other by a multi-subunit protein complex called cohesin. Since chromosomes in meiosis behave differently from those in mitosis, the cohesion and separation of homologous chromosomes and sister chromatids in meiosis are thought to be regulated by meiosis-specific cohesin subunits. Actually, several meiosis-specific cohesin subunits, including Rec8, STAG3 and SMC1b, are known to exist in mammals; however, there are no reports of meiosis-specific cohesin subunits in other vertebrates. To investigate the protein expression and localization of cohesin subunits during meiosis in non-mammalian species, we isolated cDNA clones encoding SMC1a, SMC1b, SMC3 and Rad21 in the medaka and produced antibodies against recombinant proteins. Medaka SMC1b was expressed solely in gonads, while SMC1a, SMC3 and Rad21 were also expressed in other organs and in cultured cells. SMC1b forms a complex with SMC3 but not with Rad21, in contrast to SMC1a, which forms complexes with both SMC3 and Rad21. SMC1a and Rad21 were mainly expressed in mitotically dividing cells in the testis (somatic cells and spermatogonia), although their weak expression was detected in pre-leptotene spermatocytes. SMC1b was expressed in spermatogonia and spermatocytes. SMC1b was localized along the chromosomal arms as well as on the centromeres in meiotic prophase I, and its existence on the chromosomes persisted up to metaphase II, a situation different from that reported in the mouse, in which SMC1b is lost from the chromosome arms in late pachytene despite its universal presence in vertebrates. q 2004 Elsevier B.V. All rights reserved. Keywords: Chromosome; Cohesin; Medaka; Testis; Meiosis; Spermatogenesis; SMC1; SMC3; Rad21; cDNA; Coimmunoprecipitation; Immunohistochemistry
1. Results and discussion 1.1. cDNA cloning of medaka cohesins A multi-subunit protein complex called cohesin is responsible for the cohesion and separation of sister chromatids, and most of the components are well conserved from yeast to humans (Uhlmann, 2001). Meiosis is a special type of cell division to produce gametes. In both budding and fission yeast, Rec8, a meiotic version of the Scc1/Rad21-type subunit, is required for reductional chromosome segregation in meiosis (Watanabe and Nurse, 1999). * Corresponding author. Tel.: þ 81-11-706-4454; fax: þ81-11-706-4456. E-mail address:
[email protected] (M. Yamashita). 1 Present address: Graduate School of Science and Technology, Kobe University, Nada-Ku, Kobe 657-8501, Japan. 1567-133X/$ - see front matter q 2004 Elsevier B.V. All rights reserved. doi:10.1016/j.modgep.2004.03.004
In meiosis I, Rec11-associated Rec8 is destroyed at the arms of chromosomes (Kitajima et al., 2003), while Rec8 in the centromere regions remains intact by the protective function of Shugoshin (Kitajima et al., 2004). Homologous genes have been found in various species, and the protein localization of Rec8 in mammalian meiosis has been shown to be similar to that in yeast meiosis (Eijpe et al., 2003; Lee et al., 2003). In mammals, meiosis-specific cohesin subunits other than Rec8 have been found; SMC1b and STAG3 (also called SA3) have been identified as meiotic isoforms of SMC1 and SA1/SA2, respectively. A meiotic isoform of SMC1 has only been reported for mammals, and no homologous genes have been found in the genomes of yeast, Drosophila, C. elegans, and Arabidopsis. Therefore, the subunits of cohesin are different among species as well as in mitosis and meiosis, even though cohesin plays a central and common role in chromosome cohesion in all eukaryotes.
496
T. Iwai et al. / Gene Expression Patterns 4 (2004) 495–504
More information on cohesin subunits in various animals must be obtained to gain a better understanding of the mechanisms of sister chromatid cohesion and separation during mitosis and meiosis. Meiosis-specific cohesin subunits should be characterized in detail, since they have not been examined in vertebrates other than mammals. The present study is the first investigation of cohesin subunits during meiosis in non-mammalian vertebrates. Using the medaka fish as an experimental model, we isolated cDNA clones encoding four cohesin subunits, SMC1a, SMC1b, SMC3 and Rad21 (named MeSMC1a, MeSMC1b, MeSMC3 and MeRad21, respectively). Amino acid similarity of MeSMC1a to mouse SMC1a is 91%, whereas that to mouse SMC1b is only 52%. MeSMC1b shows relatively low sequence similarity to mouse SMC1a and SMC1b (50 and 53%, respectively), indicating that there is greater species variation in SMC1b than in SMC1a. Both MeSMC1a and MeSMC1b have nucleotide-binding Walker A and Walker B motifs, which are well conserved to the SMC family (Hirano, 2002). MeSMC3 has a Walker B box in the C-terminus, with sequence similarity of 90 and 96% to the equivalent regions of Xenopus and mouse SMC3, respectively. MeRad21 has 74 and 75% similarity to Xenopus and mouse counterparts, respectively. 1.2. Protein expression of medaka cohesin subunits To investigate protein expression of cohesin subunits in medaka, we produced antibodies against recombinant proteins of MeSMC1a, MeSMC1b, MeSMC3 and MeRad21. Medaka spermatogenic cells can be staged according to their nuclear morphology, but to stage the germ cells more precisely, we produced antibodies against medaka synaptonemal complex protein 3 (MeSCP3), a major component of the axial/lateral element of synaptonemal complexes, since SCP3 is used as a general marker for meiotic cells (Lammars et al., 1994). Using the antibodies, we examined the expression of cohesin subunits in medaka organs and cultured cells (OL32 cells) (Fig. 1). The reactions with the antibodies were diminished when the antibodies were preincubated with respective antigenic recombinant proteins, indicating specificity of the signals (data not shown). An antibody raised against recombinant MeSMC1a recognized a 155-kDa band in extracts from the ovary, liver and OL32 cells. In addition to the 155-kDa band, the antibody also recognized a 150-kDa band in the testis extract (Fig. 1A, SMC1). The 150-kDa band was detected with an antibody produced against MeSMC1b (Fig. 1A, SMC1b), indicating that the antibody raised against MeSMC1a can recognize both MeSMC1a and MeSMC1b and that the anti-MeSMC1b antibody is not cross-reactive with MeSMC1a. The antibody raised against MeSMC1a was therefore designated anti-MeSMC1 and used for detecting MeSMC1a and MeSMC1b simultaneously in immunoblots. When the anti-MeSMC1 antibody was absorbed with
recombinant MeSMC1b, the resulting antibody recognized the 155-kDa band but not the anti-MeSMC1b-reactive 150-kDa band (Fig. 1A, SMC1a). This antibody was therefore designated anti-MeSMC1a and used for specific detection of MeSMC1a in immunohistochemical analysis. Anti-MeSMC3 antibody and anti-MeRad21 antibody specifically recognized a 145-kDa band and a 120-kDa band, respectively, on immunoblots of medaka tissues and cultured cells (Fig. 1A, SMC3 and Rad21). MeSCP3 was detected as 27- and 30-kDa bands in testis extracts (Fig. 1A, SCP3). In the mouse, phosphorylation is known to be the cause of two forms of SCP3 with different electrophoretic mobility (Heyting et al., 1987; Lammars et al., 1994). To investigate the difference between protein expressions of MeSMC1a and MeSMC1b, we fractionated testicular cells to obtain a fraction enriched with primary spermatocytes (Shimizu et al., 1997). In the spermatocyte-rich fraction, MeSMC1b but no, or very little, MeSMC1a was detected, while both proteins were detected in the whole extract from the testis (Fig. 1B). Since the ovary also contains meiotic cells (oocytes), MeSMC1b is thought to be expressed in this organ. However, we were not able to detect MeSMC1b in the ovary by ordinary immunoblotting analysis (Fig. 1A, SMC1b). A possible reason for the failure in detection of SMC1b in the ovary is that the ratio of germ cells to somatic cells is very low in the ovary compared to that in the testis. To concentrate the protein contents, therefore, MeSMC1b (and MeSMC1a) in the ovarian extracts were immunoprecipitated with anti-MeSMC1 antibody and then analyzed by immunoblotting with the same antibody. As in the case of the testis, the anti-MeSMC1 immunoprecipitates from the ovary contained both MeSMC1a and MeSMC1b, whereas those from the liver contained only MeSMC1a (Fig. 1C). Thus, we conclude that MeSMC1b is a germ cell-specific cohesin subunit expressed exclusively in male and female gonads. 1.3. Cohesin complexes in medaka The molecular compositions of medaka cohesin complexes were examined by coimmunoprecipitation analyses using extracts from testes and liver. Anti-MeSMC1 immunoprecipitates from medaka liver extracts contained MeSMC1a, MeSMC3 and MeRad21 but not MeSMC1b (Fig. 1D, right panel). Accordingly, anti-MeSMC3 antibody and anti-MeRad21 antibody, but not anti-MeSMC1b antibody, precipitated the remaining subunits except for MeSMC1b from the extracts. The cohesin subunits found in the immunoprecipitates were absent in control precipitates without the antibodies (data not shown), thereby confirming the specific precipitation by the antibodies. These results indicate that medaka somatic cells contain a cohesin complex that is composed of, at least, MeSMC1a, MeSMC3 and MeRad21. On the other hand, different results were obtained by similar experiments using extracts from medaka testes
T. Iwai et al. / Gene Expression Patterns 4 (2004) 495–504
497
Fig. 1. Immunoblotting and coimmunoprecipitation analyses with anti-medaka cohesin antibodies. (A) Detection of medaka SMC1, SMC1a, SMC1b, SMC3, Rad21 and SCP3 by immunoblotting. Protein extracts from the ovary (O), testis (T), liver (L) and OL32 cells (OL) were blotted with antibodies indicated in each figure. Closed triangles indicate SMC1a, SMC3, Rad21 and SCP3, and an open triangle indicates SMC1b. (B) Presence of medaka SMC1b in meiotic cells. Extracts from the spermatocyte-rich fraction (SC) and whole testis (T) were blotted with anti-MeSMC1 antibody. The spermatocyte-rich fraction contains only SMC1b (open triangle), whereas the whole testis contains SMC1a (closed triangle) as well as SMC1b. (C) Detection of SMC1b in the medaka ovary. Anti-MeSMC1 immunoprecipitates from the ovary (O), testis (T) and liver (L) were blotted with the same antibody. In contrast to the presence of SMC1a (closed triangle) in all of the tissues, SMC1b (open triangle) is absent in the liver without meiotic cells. (D) Molecular compositions of cohesin subunits in the medaka testis and liver. Using the antibodies indicated, cohesin subunits were immunoprecipitated from testis or liver extracts and then immunoblotted.
(Fig. 1D, left panel). Anti-MeSMC1 antibody precipitated MeSMC1a, MeSMC1b, MeSMC3 and MeRad21 from the extracts. Accordingly, anti-MeSMC3 antibody precipitated the remaining subunits in the extracts. However, anti-MeRad21 antibody failed to precipitate MeSMC1b from the extracts, although it did precipitate MeSMC1a, MeSMC3 and MeRad21. Furthermore, anti-MeSMC1b antibody precipitated MeSMC1b and MeSMC3 but neither MeSMC1a nor MeRad21. These results demonstrate that the medaka testis comprises, at least, two types of cohesin complex: one consisting of MeSMC1a, MeSMC3 and MeRad21 and another consisting of MeSMC1b and MeSMC3 but devoid of MeRad21. 1.4. Changes in the expression patterns of cohesin subunits during meiosis Expression patterns of MeSMC1a, MeSMC1b, and MeRad21 in the testis were investigated in detail by
immunohistochemical analyses of frozen sections with specific antibodies for each protein and anti-MeSCP3 antibody for visualizing the synaptonemal complex as a meiosis-specific marker. Anti-MeSMC3 antibody provided no clear signal in this analysis. The meiotic cell cycle of medaka germ cells proceeds synchronously in each cyst, which is bordered by Sertoli cells (Gresik et al., 1973; Shibata and Hamaguchi, 1986). This characteristic together with the immunological detection of MeSCP3 enables the precise stages of spermatogenic cells in each cyst to be easily identified. MeSMC1a was intensely expressed in somatic cells that bordered the cysts (Fig. 2A,B). Spermatogonia also expressed MeSMC1a, although the signals were weaker than those in somatic cells (Fig. 2A,B). The MeSMC1a signals were generally absent in primary and secondary spermatocytes (Fig. 2A,B) or spermatids and spermatozoa (data not shown). In a few cases, however, a faint hazy signal of MeSMC1a was detected in the cells that were estimated to be primary spermatocytes judging from their nuclear size.
498
T. Iwai et al. / Gene Expression Patterns 4 (2004) 495–504
Fig. 2. Immunocytochemical analyses of SMC1a in the medaka testis. Frozen sections were incubated with anti-MeSMC1a alone (A,B) or in combination with anti-SCP3 antibody (C,D). The primary antibody was detected with Alexa 488 (green)-conjugated anti-mouse IgG (for SMC1a) and Alexa 546 (red)conjugated anti-guinea pig IgG (for SCP3). In (A) and (B), DNA was stained with propidium iodide (red). The regions indicated in (A) and (C) are magnified in (B) and (D), respectively. sg, spermatogonium; ps, primary spermatocyte; ss, secondary spermatocyte; p-l, pre-leptotene spermatocyte; l, leptotene spermatocyte; p, pachytene spermatocyte. Bar, 30 mm (A,C), 10 mm (B,D).
Double-staining of MeSMC1a and MeSCP3 (Fig. 2C,D) revealed that the MeSMC1a-expressing spermatocytes did not express MeSCP3, the expression of which begins at the leptotene stage (Eijpe et al., 2003). It is thus likely that
the expression of MeSMC1a in the primary spermatocytes is confined to the pre-leptotene stage. In mammals, SMC1a is expressed throughout spermatogenesis (Revenkova et al., 2001; Eijpe et al., 2003). The finding of limited expression of
T. Iwai et al. / Gene Expression Patterns 4 (2004) 495–504
MeSMC1a in spermatocytes in the early phase suggests that this protein is not involved in meiosis-specific chromosome behavior in the medaka. In mammals, SMC1b has been shown to be a meiosis-specific cohesin subunit (Revenkova et al., 2001). However, MeSMC1b was expressed in spermatogonia
499
as well as in spermatocytes (Fig. 3A,B), indicating that this protein is not a meiosis-specific but a germ cell-specific cohesin subunit. String-like signals, which represent the synaptonemal complex as stained by anti-MeSCP3 antibody, were detected in primary spermatocytes (Fig. 3B – D). The intensity of string-like signals diminished
Fig. 3. Immunocytochemical analyses of SMC1b in the medaka testis. The sections were stained in the same manner as that for analyses for which the results are shown in Fig. 2. The regions indicated in (A) and (C) are magnified in (B) and (D), respectively. The asterisk ( p ) in (B) shows a somatic cell, which does not express MeSMC1b. sg, spermatogonium; ps, primary spermatocyte; ss, secondary spermatocyte; p-l, pre-leptotene spermatocyte; l, leptotene spermatocyte; p, pachytene spermatocyte. Bar, 30 mm (A,C), 10 mm (B,D).
500
T. Iwai et al. / Gene Expression Patterns 4 (2004) 495–504
as meiosis proceeded, and they were not detected in secondary spermatocytes (Fig. 3A). MeSMC1b was not expressed in spermatids and spermatozoa (data not shown) or somatic cells (Fig. 3A,B). In addition to its expression in somatic cells, Rad21 has been reported to be expressed in meiotic cells in the mouse (Prieto et al., 2002; Parra et al., 2004) and yeast (Yokobayashi et al., 2003). MeRad21 was expressed most extensively in somatic cells that bordered the cysts (Fig. 4A,B). MeRad21 was also expressed weakly in spermatogonia, but was hardly detected in primary and secondary spermatocytes
(Fig. 4A,B) or spermatids (data not shown). Like MeSMC1a, however, a hazy signal of MeRad21 was detected in pre-leptotene spermatocytes as judged from their absence of MeSCP3 expression (Fig. 4C,D). Since Rad21 colocalizes to SCP3 and accumulates in the centromere regions from late prophase I to metaphase I in the mouse, Rad21 is thought to be involved in monopolar attachment in meiosis I (Parra et al., 2004). In striking contrast to this notion, the lack of colocalization of MeRad21 and MeSCP3 in medaka spermatocytes (Fig. 4C,D) suggests that MeRad21 is not involved in meiosis-specific chromosome behavior.
Fig. 4. Immunocytochemical analyses of Rad21 in the medaka testis. The sections were stained in the same manner as that for analyses for which the results are shown in Fig. 2. The regions indicated in (A) and (C) are magnified in (B) and (D), respectively. sg, spermatogonium; ps, primary spermatocyte; ss, secondary spermatocyte; p-l, pre-leptotene spermatocyte; l, leptotene spermatocyte; p, pachytene spermatocyte. Bar, 30 mm (A,C), 10 mm (B,D).
T. Iwai et al. / Gene Expression Patterns 4 (2004) 495–504
1.5. Changes in the localization of MeSMC1b on meiotic chromosomes We further investigated in detail the localization of MeSMC1b on meiotic chromosomes using chromosome spreads of spermatocytes, which enable close observations
501
of chromosomal morphology during meiosis. Primary spermatocytes already expressed MeSMC1b at the pre-leptotene stage (Fig. 5A, see also Fig. 3D). The signals of MeSMC1b were visible as fine dots spread over the entire nucleus. At the leptotene stage (Fig. 5B), the signals of MeSMC1b had become stronger and denser than those in
Fig. 5. Localization of MeSMC1b on meiotic chromosomes in the medaka. Chromosome spreads were stained with anti-MeSMC1b antibody and Alexa 488 (green)-conjugated anti-mouse IgG antibody except for (E). In (E), chromosome spread was stained with anti-MeSCP3 antibody and Alexa 546 (red)conjugated anti-guinea pig antibody. (D) and (E) represent the same chromosome spreads. The samples were simultaneously stained with DAPI (blue) for DNA. (A) pre-leptotene, (B) leptotene, (C) zygotene, (D) and (E) pachytene, (F) diplotene, (G) diakinesis, (H) metaphase I, (I) anaphase I, (J) metaphase II, (K) anaphase II, (L) higher magnification of the region indicated in H. Bar, 10 mm (A –K), 2 mm (L).
502
T. Iwai et al. / Gene Expression Patterns 4 (2004) 495–504
Fig. 6. Summary of the protein expression patterns of cohesin subunits (SMC1a, SMC1b and Rad21) and a component of the synaptonemal complex (SCP3) in the medaka. The expression of each protein is shown by lines (distinct expression is shown by solid lines, and hazy expression is shown by broken lines). sg, spermatogonium; p-l, pre-leptotene; l, leptotene; z, zygotene; p, pachytene; d, diplotene; MI, metaphase I; ss, secondary spermatocyte; st, spermatid; sp, spermatozoon.
the previous stage, with a change in their appearance from dots to short fine threads. At the zygotene stage, the threadlike signals of MeSMC1b on the whole nucleus gathered to formlong and distinct lines that marked the axes of chromosomes undergoing synapsis (Fig. 5C). At the pachytene stage, the MeSMC1b signals had become thicker and shorter than those found at the zygotene stage, appearing as 24 lines associated with each pair of homologous chromosomes (Fig. 5D). At this stage, MeSCP3 colocalized with MeSMC1b (Fig. 5E). At the diplotene stage (Fig. 5F), the homologous chromosomes underwent desynapsis, remaining bound to each other at the chiasmata and at the arms distal to chiasmata (Miyazaki and Orr-Weaver, 1994). Chromosomes were further condensed at the diakinesis stage (Fig. 5G), at which MeSMC1b was detected in both synaptic and desynaptic chromosomal regions. Unlike the line-like signals observed at the previous stages, the MeSMC1b signals found on the desynaptic chromosomes at the diakinesis stage were discontinuous and appeared as dots (Fig. 5G). At metaphase I (Fig. 5H), dotty MeSMC1b signals were found on all of the 24 fully condensed bivalent chromosomes, irrespective of the regions of arms and centromeres. MeSMC1b signals were also detected on cross-shaped bivalent chromosomes, which had probably been pulled in opposite directions by spindle microtubules (Fig. 5L). Homologous chromosomes were completely separated and pulled toward opposite poles at anaphase I, a stage at which MeSMC1b was still found on the whole regions of all chromosomes, irrespective of arms and centromeres (Fig. 5I). MeSMC1b remained detectable on all of the chromosomes even at metaphase II (Fig. 5J). The sister chromatids were separated at anaphase II, when MeSMC1b was no longer detectable on the chromosomes (Fig. 5K). These results clearly indicate that MeSMC1b remained localized on chromosomes until metaphase II, despite the fact that cohesin is expected to be lost along most of the chromosome length at this stage. This finding is in marked contrast to that in the mouse, in which SMC1b is largely lost from chromosome arms in late pachytene (Revenkova et al., 2001; Eijpe et al., 2003).
In summary, we have reported the protein expression of cohesin subunits during medaka spermatogenesis (Fig. 6). The following differences were notable found between medaka and mammals: (1) in contrast to the continuous expression of SMC1a during spermatogenesis in mammals, the expression of MeSMC1a is mainly in spermatogonia, (2) SMC1b is a meiosis-specific cohesin in mammals, whereas its expression is apparently initiated in spermatogonia (thus being a germ cell-specific cohesin in medaka), (3) in mammals, Rad21 is expressed in meiotic cells up to metaphase I, while the expression of MeRad21 ceases much earlier (before leptotene) in medaka, and (4) SMC1b is largely lost from chromosome arms in late pachytene in mammals, whereas it remains localized on chromosomes until metaphase II in medaka. These findings suggest that although the basic mechanisms are ubiquitous in all eukaryotes, the roles of cohesin subunits in assuring meiosis-specific chromosome behavior exhibit considerable species variations. To obtain a better understanding of the molecular mechanisms that regulate the behavior of chromosomes in meiosis, we need to focus on Rec8 and STAG3, the most prominent meiosis-specific cohesin subunits. However, we have not been able to isolate Rec8 and STAG3 from medaka despite the use of RT-PCR with various sets of degenerate primers including those used for cloning Rec8 in other bony fish such as blowfish (fugu) and zebrafish. Since the medaka genome project is proceeding, sequence data obtained from the project should enable us to obtain medaka Rec8 and STAG3 in the near future.
2. Experimental procedures 2.1. Cloning of cDNAs encoding medaka cohesins Using a first-strand cDNA synthesis kit (Gibco BRL LIFE Technology, Tokyo, Japan), cDNAs were produced from total RNA isolated from medaka testes with ISOGEN (Nippon Gene, Tokyo, Japan), and cDNA fragments of medaka cohesins were obtained by RT-PCR using the following
T. Iwai et al. / Gene Expression Patterns 4 (2004) 495–504
degenerate oligonucleotide primers: for rad21, 50 -TGGYTIGCICAYTGGGA-30 and 50 -GTCAAARTCRTGGAACTCCT30 ; for smc3, 50 -GARGARTGYATGAARAARAT-30 and 50 -GCYTGRTCIATYTCRTCRAA-30 ; for smc1a, 50 -AARGCIMRIMGITGGGAYGARAA-30 and 50 -GRTTYTTYTCRWARTCIARYTG-30 ; for smc1b, 50 -AARTCWGGAGTIATYTCTGGW-30 and 50 -ACYTTICCDATRTTIGTRTTRTC30 [Y (C or T); R (A or G); M (A or C); W (A or T); D (A or T or G); I (inosine)]; for scp3, 50 -ACTAAGAACCACATGACAGG-30 and 50 -GCCATTTCTCGTTTCAGTTC-30 (according to partial cDNA sequence of scp3 provided from Dr A. Kanamori, Nagoya University, Japan). After confirming the sequences, the amplified PCR fragments were labeled with digoxigenin (DIG) using a DIG DNA Labeling and Detection Kit (Roche Diagnostics, Tokyo, Japan) and used as hybridization probes to screen a cDNA library constructed from medaka testis (generously provided by Dr T. Kobayashi, NIBB, Japan). The isolated clones were sequenced to confirm that they are medaka homologs of cohesin subunits. The sequence data appear in the DDBJ/EMBL/GenBank databases with the accession numbers AB115697 (rad21), AB115698 (smc3), AB097255 (smc1a), AB115699 (smc1b) and AB162905 (scp3).
503
2.3. Extraction of proteins Medaka, Oryzias latipes, were purchased from a local fish farm (Yatomi, Aichi Prefecture, Japan) and kept in fresh water at 27 8C under artificial light conditions (14-h light and 10-h dark conditions). OL32 cells, which are an in vitrocultured cell line originally isolated from the fin of medaka (a generous gift from Dr H. Mitani, Tokyo University, Japan), were cultured in Leibovitz’s L-15 medium supplemented with 15% fetal calf serum and 10 mM HEPES (pH 7.5) at 33 8C. Cell fractions enriched with the primary spermatocytes were collected by sedimentation under normal gravity through continuous gradients of 1– 4% bovine serum albumin, according the method described previously (Shimizu et al., 1997). Proteins were extracted from tissues and cells by sonication in RIPA buffer (20 mM Tris – HCl (pH 7.4), 150 mM NaCl, 2 mM EDTA, 1% Nonidet P-40, 1% sodium deoxycholate, 0.1% SDS, 50 mM NaF, 1 mM Na3VO4, 5 mM 2-mercaptoethanol, 10 mg/ml aprotinin, 10 mg/ml leupeptin, 1 mM phenylmethylsulfonyl fluoride). After centrifugation at 16,000 £ g for 20 min, the supernatants were collected and stored at 2 80 8C until use. 2.4. Coimmunoprecipitation and immunohistochemistry
2.2. Production of antibodies DNA fragments encoding full lengths of MeRad21 and MeSCP3, C-termini of MeSMC1a, MeSMC1b and MeSMC3 were amplified by PCR using the following primer sets (50 and 30 primers introducing the Eco RI and Xho I (or Sal I) sites, respectively, as indicated by underlines): for MeRad21, CGGAATTCCAATGTTTTACGCCCACTTTG and CTACTCGAGAATAAGATTGATTGAATCTTGG TCC; for MeSMC1a, CGGAATTCGCAAAAAACTTGGAGGCGCTA and TTCGTCGACTTCATTTGGATTGGGATTCGC; for MeSMC1b, CGAATTCGACTGGAGCA GAAGCGC and GTGCTCGAGTGCAGTCTCTTTCAGGCT; for MeSMC3, TGGAATTCAAGAGAAGCTCATCAAGCGGC and TTTCTCGAGGCCGTGGGTGG TGTCGTCTTC, for MeSCP3, GGAATTCAAATGGAGTCTGTGAGAAAAGTG and CCGCTCGAGGAAGAACATGGTCTGCAG. The amplified fragments were ligated into the expression vector pET-21c (Novagen, Madison, WI), except for the fragment of MeSMC3, which was ligated into the expression vector pGEX-KG (Guan and Dixon, 1991). Recombinant proteins were produced and purified for use as antigens according to the method described previously (Hirai et al., 1992). Mouse antisera were raised against MeRad21, MeSMC1a and MeSMC1b, rabbit antisera were raised against MeSMC3 and guinea pig antisera were raised against MeSCP3. The antisera were affinity-purified with the antigenic proteins electroblotted onto an Immobilon membrane (Millipore, Tokyo, Japan).
Extracts from tissues and cells were incubated with either anti-MeSMC1, anti-MeSMC3, anti-MeSMC1b or anti-Rad21 antibody in the presence of Protein G Sepharose (Amersham Biosciences, Piscataway, NJ) at 4 8C overnight with rotor agitation. After washing six times in RIPA buffer, the immunoprecipitates were separated by SDS-PAGE with 7.5% gel, blotted onto an Immobilon membrane, and probed with primary and secondary antibodies, as described previously (Lee et al., 2003). Immunohistochemical observations were carried out with frozen sections of the testes, according to the method described previously (Lee et al., 2003). 2.5. Chromosome spreads of medaka spermatocytes Testes were minced with scissors and dissociated into single cells in Ca2þ/Mg2þ-free medaka Ringer’s solution (Iwamatsu, 1973) containing 0.2% collagenase (236 U/mg, Wako Pure Chemicals, Osaka, Japan), 0.6 U dispase (Gibco BRL) and 0.002% DNase I (Sigma-Aldrich, St Louise, MO) with gentle agitation. The cell suspension was filtered through a mesh with a pore size of 42 mm and loaded on a 25% Percoll (Amersham Biosciences) in medaka Ringer’s solution. Spermatocyte-rich fractions were obtained by centrifugation at 3000 rpm for 30 min. Meiotic nuclear spreads were prepared according to the method described by Lee et al. (2003) for mouse spermatocytes except that medaka spermatocytes put on poly-L -lysine-coated coverslips were treated with 75 mM KCl for 10 min instead of 85 mM NaCl for 3 min for mouse spermatocytes and that DNA
504
T. Iwai et al. / Gene Expression Patterns 4 (2004) 495–504
was counter-stained with 5 mg/ml DAPI (40 ,6-diamidino2-phenylindole).
Acknowledgements We thank Drs Hiroshi Mitani (Tokyo University, Tokyo), Tohru Kobayashi (National Institute for Basic Biology, Okazaki) and Akira Kanamori (Nagoya University, Aichi) for providing OL32 cells, medaka testis cDNA library and partial cDNA fragment of medaka scp3, respectively. We also thank Dr Yasushi Shibata (National Institute for Basic Biology, Okazaki) for technical advice on the production of anti-MeSCP3 antibody in guinea pig. This work was supported in part by Grants-in-Aid for Scientific Research from the Japan Society for the Promotion of Science to J.L. and from the Ministry of Education, Science, Sports and Culture of Japan to M.Y. (11236201) and by 21st Century COE Program ‘Center of Excellence for Advanced Life Science on the Base of Bioscience and Nanotechnology’ from the Ministry of Education, Science, Sports and Culture of Japan to M.Y.
References Eijpe, M., Offenberg, H., Jessberger, R., Revenkova, E., Heyting, C., 2003. Meiotic cohesin Rec8 marks the axial elements of rat synaptonemal complexes before cohesins SMC1b and SMC3. J. Cell Biol. 160, 657– 670. Gresik, E.W., Quirk, G.B., Hamilton, G.B., 1973. Fine structure of the Sertoli cell of the testis of the teleost Oryzias latipes. Gen. Comp. Endocrinol. 21, 341–352. Guan, K.L., Dixson, J.E., 1991. Eukaryotic proteins expressed in Escherichia coli: an improved thrombin cleavage and purification procedure of fusion proteins with glutathione S-transferase. Anal. Biochem. 192, 262–267. Heyting, C., Moens, P.B., van Raamsdonk, W., Dietrich, A.J.J., Vink, A.C.G., Redeker, E.J.W., 1987. Identification of two major components of the lateral elements of synaptonemal complexes. Eur. J. Cell Biol. 43, 148– 154. Hirai, T., Yamashita, M., Yoshikuni, M., Lou, Y.H., Nagahama, Y., 1992. Cyclin B in fish oocytes: its cDNA and amino acid sequences,
appearance during maturation, and induction of p34cdc2 activation. Mol. Reprod. Dev. 33, 131– 140. Hirano, T., 2002. The ABCs of SMC proteins: two-armed ATPases for chromosome condensation, cohesion, and repair. Genes Dev. 9, 587–599. Iwamatsu, T., 1973. Studies on oocyte maturation in the medaka, Oryzias latipes. Improvement of culture medium for oocytes in vitro. Jpn. J. Ichthyol. 20, 218–224. Kitajima, T.S., Yokobayashi, S., Yamamoto, M., Watanabe, T., 2003. Distinct cohesin complexes organize meiotic chromosome domains. Science 300, 1152–1155. Kitajima, T.S., Kashiwabara, S.A., Watanabe, Y., 2004. The conserved kinetochore protein shugoshin protects centromeric cohesion during meiosis. Nature 427, 510– 517. Lammers, J.H.M., Offenberg, H.H., van Aalderen, M., Vink, A.C.G., Dietrich, A.J.J., Heyting, C., 1994. The gene encoding a major component of the lateral elements of synaptonemal complexes of the rat is related to X-linked lymphocyte-regulated genes. Mol. Cell. Biol. 14, 1137–1146. Lee, J., Iwai, T., Yokota, T., Yamashita, M., 2003. Temporally and spatially selective loss of Rec8 protein from meiotic chromosomes during mammalian meiosis. J. Cell Sci. 116, 2781–2790. Miyazaki, W.Y., Orr-Weaver, T.L., 1994. Sister-chromatid cohesion in mitosis and meiosis. Annu. Rev. Genet. 28, 167 –187. Parra, M.T., Viera, A., Go´mez, R., Page, J., Benavente, R., Santos, J.L., et al., 2004. Involvement of the cohesin Rad21 and SCP3 in monopolar attachment of sister kinetochores during mouse meiosis I. J. Cell Sci. 117, 1221–1234. Prieto, I., Pezzi, N., Buesa, J.M., Kremer, L., Barthelemy, I., Carreiro, C., et al., 2002. STAG2 and Rad21 mammalian mitotic cohesins are implicated in meiosis. Eur. Mol. Biol. Org. Rep. 3, 543 –550. Revenkova, E., Eijpe, M., Heyting, C., Gross, B., Jessberger, R., 2001. Novel meiosis-specific isoform of mammalian SMC1. Mol. Cell. Biol. 21, 6984–6998. Shibata, N., Hamaguchi, S., 1986. Electron microscopic study of the blood-testis barrier in the teleost, Oryzias latipes. Zool. Sci. 3, 331 –338. Shimizu, Y., Shibata, N., Yamashita, M., 1997. Spermiogenesis without proceeding meiosis in the hybrid medaka between Oryzias latipes and O. curvinotus. J. Exp. Zool. 279, 102–112. Uhlmann, F., 2001. Chromosome cohesion and segregation in mitosis and meiosis. Curr. Opin. Cell Biol. 13, 754–761. Watanabe, Y., Nurse, P., 1999. Cohesin Rec8 is required for reductional chromosome segregation at meiosis. Nature 400, 461 –464. Yokobayashi, S., Yamamoto, M., Watanabe, Y., 2003. Cohesins determine the attachment manner of kinetochores to spindle microtubules at meiosis I in fission yeast. Mol. Cell. Biol. 23, 3965–3973.