- Email: [email protected]
b
Biochimica et Biophysica Acta 1493 (2000) 33^40
www.elsevier.com/locate/bba
Testis- and developmental stage-speci¢c expression of hnRNP A2/B1 splicing isoforms, B0a/b Miwa Matsui a , Hisashi Horiguchi a; *, Hiroshi Kamma b , Masachika Fujiwara c , Rieko Ohtsubo a , Takesaburo Ogata a b c
a Center for Medical Sciences, Ibaraki Prefectural University of Health Sciences, Ibaraki 300-0394, Japan Department of Pathology, Institute of Basic Medical Sciences, University of Tsukuba, Ibaraki 305-8575, Japan Department of Pathology, Kensei General Hospital, 604 Sukita, Iwase, Nishi-Ibaraki, Ibaraki 309-1295, Japan
Received 29 February 2000; received in revised form 15 May 2000 ; accepted 26 May 2000
Abstract Heterogeneous nuclear ribonucleoproteins (hnRNPs) A2 and B1 are abundant nuclear proteins that bind to nascent RNAs synthesized by RNA polymerase II. Previously we had found that the splicing isoforms hnRNP B0a/b, from which the ninth exon of the A2/B1 gene is excluded, are abundantly expressed in testis. We postulated that B0a/b are testis-specific isoforms, and investigated the expression of A2/B1 and B0a/b in rat tissues and in postnatal development of rat testes using RNase protection assay, immunoblotting, and immunohistochemistry. We found that hnRNP B0a/b mRNAs are expressed in several tissues but that the testis alone expresses B0a/b proteins. A sequential study using neonatal rat testes demonstrated that B0a/b mRNAs are produced after 17 days of age, but not translated until 4 weeks of age when round spermatids appear in addition to spermatogonia and spermatocytes. Immunohistochemically, hnRNP A2/B1 isoforms are expressed during spermatogenesis from spermatogonia through round spermatids, whereas the expression of A1 is restricted to spermatogonia. This expression pattern in the rat testis is maintained from birth through adulthood. These results suggest that the expression of the hnRNP A2/B1 gene is partly regulated by a testis-specific post-transcriptional mechanism, and that the products of the A2/B1 gene, especially hnRNP B0a/b, are involved in spermatogenesis. ß 2000 Elsevier Science B.V. All rights reserved. Keywords : Heterogeneous nuclear ribonucleoprotein A2/B1; Heterogeneous nuclear ribonucleoprotein B0a/b ; Splicing isoform; Tissue-speci¢c expression; Testis; Spermatogenesis
1. Introduction Heterogeneous nuclear ribonucleoproteins (hnRNPs) bind to nascent RNAs synthesized by RNA polymerase II and are thought to function in a wide range of cellular activities, such as pre-mRNA processing, mRNA localization, translation, and mRNA turnover [1^3]. In recent years, post-transcriptional as well as transcriptional mechanisms have attracted considerable attention in the regulation of gene expression. RNA binding proteins, including hnRNPs, play central roles in the post-transcriptional regulation of gene expression [4,5].
* Corresponding author. Fax: +81-298-40-2313; E-mail : [email protected]
HnRNPs were ¢rst identi¢ed as a major group of chromatin-associated RNA binding proteins [6]. Subsequently, the major hnRNPs were cloned and characterized, and about 30 hnRNP proteins have been identi¢ed and are designated from A1 to U at present [1,5]. Of these, there is a group of basic hnRNP proteins collectively called the A/B group proteins [1,7]. HnRNP A1, A2, and B1 proteins are its main members. The structural feature of these proteins is that they consist of two conserved amino-terminal RNA binding domains (RBDs, or RNA recognition motifs) linked to a glycine-rich domain that contains an RGG box and a nuclear localization signal termed M9 [1,8^11]. A2/B1 and A1 proteins are quite similar in their gene and protein structures, and they are thought to be evolutionarily duplicated [12^14]. A1 protein is produced from multiple genes [15,16], and its splicing isoform A1B is produced by insertion of the 7 bis exon [17]. A2 and B1 proteins are produced by alternative splicing from a single-copy gene, and they di¡er from each other only by a
0167-4781 / 00 / $ ^ see front matter ß 2000 Elsevier Science B.V. All rights reserved. PII: S 0 1 6 7 - 4 7 8 1 ( 0 0 ) 0 0 1 5 4 - 8
BBAEXP 93428 25-8-00
Cyaan Magenta Geel Zwart
34
M. Matsui et al. / Biochimica et Biophysica Acta 1493 (2000) 33^40
12-amino-acid insertion in the N-terminal RBD of B1 [8,13,14]. Recently, we isolated the splicing isoforms of hnRNP A2/B1, named B0a/b [18]. The ninth exon of the A2/B1 gene, which is structurally equivalent to the 7 bis exon of A1 gene, is excluded in B0a/b [13,14,18]. Thus, the overall structure of the A2/B1 gene is still more analogous to that of the A1 gene. A2/B1 and A1 proteins also share some functional similarities, such as hnRNP complex formation [1] and in vitro UUAG-speci¢c RNA binding [19]. They in£uence in vitro alternative 5P splicing site selection as antagonists of SR proteins [7,20,21]. Due to these features, A/B group proteins are thought to be ubiquitously expressed in all cells and constitutively function through their interactions with other molecules. It is still unclear why the diversity of A/B group proteins exists or what di¡ers among them. Recent studies, however, suggest a cell type-speci¢c function of hnRNP A2 in spite of its ubiquitous expression. It has been demonstrated that hnRNP A2 speci¢cally binds myelin basic protein mRNA in cultured myelinating cells and plays a central role in its transport, localization, and translation [22]. In addition, we have previously demonstrated the expression patterns of A2/B1 and A1 in the seminiferous tubules, suggesting they have di¡erent roles in spermatogenesis [18,23]. We also found that the splicing isoforms B0a/b are abundantly expressed in testis [18]. It is known that some RNA binding proteins are essential for spermatogenesis, and that some genes produce testis-speci¢c isoforms by alternative splicing or polyadenylation [24^26]. These observations led us to postulate that B0a/b are testis-speci¢c isoforms that play an important role in testis. To address this supposition, we investigated the expression of A2/B1 gene products among rat tissues and in postnatal rat testes from birth to adulthood, using RNase protection assay (RPA), immunoblotting, and immunohistochemistry. In this study, we demonstrated that the expression of hnRNP B0a/b, splicing isoforms of hnRNP A2/B1, is regulated in a testis-speci¢c manner, and that B0a/b are expressed in postnatal development of the testis. 2. Materials and methods
and frozen with liquid nitrogen to extract mRNAs and proteins. At the same time, the testis tissues were also ¢xed in Bouin's solution (saturated picric acid:formaldehyde: acetic acid = 75:25:5) for histological and immunohistochemical analysis (see below). 2.2. Preparation of probes and RPA We used a plasmid DNA containing the rat hnRNP B1 coding sequence previously cloned [18] (GenBank/EMBL/ DDBJ accession number AB006816). A rat hnRNP B1 cDNA 5P fragment digested with EcoRI and HinfI (nt 1^ 306, probe 1 in Fig. 1), encompassing the hnRNP B1-speci¢c region, i.e., exon 2, was cloned in a pBluescript II KS+ vector (pBSII KS+; Stratagene, La Jolla, CA, USA) digested with EcoRI and EcoRV. PCR ampli¢cation was used to isolate a cDNA fragment of rat hnRNP B1 (nt 468^811, probe 2 in Fig. 1), containing part of the exon 9 region, and it was subcloned into a pBluescript SK+ vector (pBS SK+ ; Stratagene) digested with BamHI and SalI. The pTRI RNA 18S antisense control template (Ambion, Austin, TX, USA) was used as internal control and calibration. To synthesize the K-32 P-labeled antisense RNA probes, the plasmid DNAs of probe 1 and probe 2 were linearized with the EcoRI site in the pBSII KS+ polylinker and with the BamHI site in the pBS SK+ polylinker, and transcribed with T3 and T7 RNA polymerase in the presence of [K-32 P]UTP, respectively [27]. Total RNAs of rat tissues were extracted and puri¢ed by the acidic guanidinium isothiocyanate^phenol^chloroform method [28]. RPAs were performed using a Ribonuclease Protection Assay kit (RPAII ; Ambion). The probes were hybridized to 25 or 10 Wg of total RNA from the rat tissues at 42³C for 16^18 h. After hybridization, the mixtures were digested with RNase solution containing RNase A/RNase T1 mixture in quantities recommended by the manufacturer. Protected fragments were analyzed on 5% denaturing polyacrylamide gels in 1UTBE bu¡er and visualized by autoradiography. The image analysis of the autoradiograms was performed using a model GS-700 Imaging Densitometer (Bio-Rad Laboratories, Hercules, CA, USA). 2.3. Gel electrophoresis and immunoblotting
2.1. Source of tissue samples To investigate di¡erential expression among tissues, male and female Wistar rats aged 10 weeks were purchased from Japan SLC, Inc. (Hamamatsu, Japan). For sequential analysis of postnatal development of testes, male Wistar rats aged 3, 4, 5, and 10 weeks, and pregnant Wistar rats were also purchased. The pregnant rats were maintained in our laboratory, and their male neonates were used in experiments 1, 5, 10, 14, and 17 days after birth. They were anesthetized with pentobarbital sodium and killed, and various tissues were immediately removed
BBAEXP 93428 25-8-00
Total protein extracts from rat tissues were prepared and adjusted to contain approximately the same total amount of the protein. Protein samples (10 Wg) were separated on 10% SDS^polyacrylamide gels (SDS-PAGE) [29]. Immunoblotting was performed and probed with anti-hnRNP A2/B1 (DP3) mouse monoclonal antibody in 1:5000 dilution as previously described [18]. Bound antibodies were detected with peroxidase-conjugated anti-mouse immunoglobulin antibody (Amersham Pharmacia Biotech, Uppsala, Sweden) and the ECL system (Amersham Pharmacia Biotech).
Cyaan Magenta Geel Zwart
M. Matsui et al. / Biochimica et Biophysica Acta 1493 (2000) 33^40
2.4. Histological and immunohistochemical analysis The ¢xed testis tissues were embedded in para¤n, serially cut at 3 Wm thickness, and mounted on glass slides. Depara¤nized tissue sections were reacted with periodic acid^Schi¡ (PAS) reagent and counterstained with hematoxylin. For immunohistochemistry, the mouse monoclonal antibodies used were anti-hnRNP A2/B1 (DP3 [18]) in 1:500 dilution and anti-hnRNP A1 (9H10; kindly provided by Dr. G. Dreyfuss) in 1:500 dilution. We used a Catalyzed Signal Ampli¢cation system (CSA system; Dako Corporation, Carpinteria, CA, USA) with slight modi¢cation to the manufacturer's instruction. Brie£y, the depara¤nized tissue sections were pretreated in a microwave oven at 95³C for 1 min in a target retrieval solution (provided by the manufacturer, based on 10 mM citrate bu¡er, pH 6.0) and incubated with the above ¢rst antibodies for 1 h at room temperature. Bound antibodies were detected by the CSA system. Control sections were incubated with non-immunized mouse serum substituted for the primary antibodies. Aminoethyl-carbamazole (Aldrich, Milwaukee, WI, USA) was used as a coloring substrate for detection of peroxidase, and followed by counterstaining with hematoxylin. Stained specimens were examined using a microscope. 3. Results 3.1. Testis-speci¢c expression of hnRNP B0 Here we use the terms Ex2(+), Ex2(3), Ex9(+), and Ex9(3) to distinguish the splicing sites of the products
35
of the hnRNP A2/B1 gene. For example, Ex2(+) designates the product into which the second exon was spliced. We also use the term `B0' for the Ex9(3) isoform, since the B0a (Ex2(3), Ex9(3)) and B0b (Ex2(+), Ex9(3)) mRNAs are indistinguishable at the same time because of the limitation of the probe length in RPA (Fig. 1). We regarded the sum of Ex2(+) and Ex2(3) or Ex9(+) and Ex9(3) mRNAs as the total amount of hnRNP A2/ B1 gene-derived transcripts and applied the term `total A2/ B1 mRNA' to it. The quanti¢ed relative values of RPA among rat tissues are shown in Fig. 2. The protection probe 1 produced protected bands of 306 nt and 264 nt, corresponding to Ex2(+) and Ex2(3) mRNA respectively, in all rat tissues examined (Fig. 2A). Although the amounts of Ex2(+) and Ex2(3) mRNAs varied across di¡erent tissues, the relative amounts of Ex2(+) to total A2/B1 mRNAs were nearly constant, ranging from 0.05 to 0.18 (mean 0.13) (Fig. 2D). This indicates that the alternative splicing producing the Ex2(+) messages correlates with the amounts of total A2/B1 mRNAs. The protection probe 2 produced protected bands of 344 nt and 291 nt, corresponding to Ex9(+) and Ex9(3) mRNA respectively. The Ex9(+) mRNAs, meaning the sum of the original hnRNP A2 and B1 mRNAs, were identi¢ed in all rat tissues as expected, whereas the expression levels of the Ex9(3) mRNAs varied considerably among the di¡erent tissues, i.e., the Ex9(3) messages were abundant in testis and thymus, less abundant in spleen and intestine, and much rarer or undetectable in other tissues (Fig. 2B,E). Immunoblotting of proteins from rat tissues and HeLa cells detected proteins of the expected sizes for hnRNP A2 (36 kDa) and B1 (38 kDa).
Fig. 1. Structure of the hnRNP A2/B1 and B0a/b transcripts, coding regions. Regions that correspond to the structural or functional domains of the gene product are shown: RBD-I, RNA binding domain I; RBD-II, RNA binding domain II; RGG motifs, R-G-G repeat-containing sequence; M9, nuclear localization signal. The conserved sequence in hnRNP B1, corresponding to exon 2 (Ex2), is indicated by the small hatched box. The excluded sequence in the glycine-rich domain is indicated by the small black box, corresponding to the ninth exon of the human hnRNP A2/B1 gene (Ex9). Probe 1 and 2 used in RPA are designed to distinguish the alternatively spliced mRNAs of the exons 2 and 9 respectively. The numbers indicate the nucleotide sequence positions of hnRNP A2/B1 from the initiation codon, ATG.
BBAEXP 93428 25-8-00
Cyaan Magenta Geel Zwart
36
M. Matsui et al. / Biochimica et Biophysica Acta 1493 (2000) 33^40
Fig. 2. RPA among rat tissues. 25 Wg total RNA of indicated rat tissues was analyzed by RPA with probe 1 (A) and probe 2 (B); as a control, 0.5 Wg total RNA of those tissues was analyzed with an 18S ribosomal RNA probe (C) as described in Section 2. The no target controls, without and with RNase treatment, are seen on the left side of A^C as `probe' and `control' respectively. The probes show the expected results ; each primarily full-length probe is 368 bp (A), 415 bp (B), and 128 bp (C). The controls show no protected signals as expected. The size markers are also demonstrated in A^C: `100 bp ladder', [Q-32 P]ATP-labeled 100 bp DNA ladder. The protected messages in A and B were quanti¢ed as described in Section 2, using the 18S ribosomal RNA signals as calibration standard. (D, E) The ratio of each optical density to the sum of the lung Ex2(+) and Ex2(3) optical densities in A, or Ex9(+) and Ex9(3) in B, was calculated as the relative optical density. The estimated values of the Ex2(+) (306 bp, hatched bars) and Ex2(3) (264 bp, open bars) mRNAs are presented in D, and those of the Ex9(+) (344 bp, open bars) and Ex9(3) (291 bp, hatched bars) in E. The height of the bar, i.e., open bar plus hatched bar, in D and E indicates the total amount of A2/B1 messages, and the numbers above the bars denote the ratio of the Ex2(+) or Ex9(3) to the total amount of the messages.
It also showed that the testis alone contained an additional 33 kDa band, which is the expected size of splicing isoform B0, speci¢cally B0a excluding the second exon (Fig. 3). B0b protein was scarcely detected in rat testis, but it is di¤cult to visualize under the same immunoblot condition as Fig. 3 (data not shown). Thus, the exclusion of the ninth exon occurs in a tissue-speci¢c manner, and the expression of B0 proteins is testis-speci¢c. 3.2. Developmental stage-speci¢c expression of hnRNP B0 in rat testis We next investigated the mRNA expression levels of hnRNP A2/B1 isoforms during postnatal development of the rat testis from birth to adulthood (Fig. 4). The total A2/B1 mRNAs gradually increased after birth to 4 weeks of age, and then the expression level decreased to the adult one. During this period, the Ex2(+) mRNAs increased in amount almost proportionally to the total A2/B1 mRNAs (Fig. 4A,D), which indicates that the alternative splicing producing the Ex2(+) messages is constitutive in postnatal development of the testis.
BBAEXP 93428 25-8-00
Fig. 3. Immunoblot analysis of rat tissues with anti-hnRNP A2/B1 antibody. Samples of indicated rat tissues and HeLa cells were separated by electrophoresis in SDS^10% polyacrylamide gel and transferred onto a membrane as described in Section 2. Immunoblotting using anti-hnRNP A2/B1 monoclonal antibody DP3 was performed. The positions of the hnRNP A2/B1 isoforms are shown to the right of the ¢gure. An additional 39 kDa band (asterisk) is supposed to be another isoform of hnRNP A2/B1 or to be post-translationally modi¢ed hnRNP A2/B1 proteins [18].
Cyaan Magenta Geel Zwart
M. Matsui et al. / Biochimica et Biophysica Acta 1493 (2000) 33^40
37
Fig. 4. RPA of rat testis during postnatal developmental. 10 Wg total RNA of rat testis and liver tissues at indicated ages was analyzed by RPA with probe 1 (A), probe 2 (B), and 0.5 Wg total RNA of those tissues was analyzed with a 18S ribosomal RNA probe (C) as described in Section 2. The size markers and controls are also demonstrated in A^C in the same manner as in Fig. 2. The protected messages in A and B were quanti¢ed as described in Section 2, using the 18S ribosomal RNA signals as calibration standard. (D, E) The ratio of each optical density to the sum of the 10-week-old testis Ex2(+) and Ex2(3) optical densities in A, or Ex9(+) and Ex9(3) in B, was calculated as the relative optical density. The estimated values of the mRNAs are presented in D and E in the same manner as in Fig. 2.
However, the Ex9(3) mRNAs became evident by the 17th day after birth, and they gradually increased in amount from then until 4 weeks of age (Fig. 4B,E). The relative amounts of Ex9(3) mRNAs decreased after the 5th week, but Ex9(3) mRNAs still constituted about 40% of total A2/B1 mRNAs. It was shown that the proportion of Ex9(3) to total A2/B1 mRNAs reached the level of the adult rat by the fourth week after birth. In contrast, the expression levels of hnRNP A2/B1 mRNAs in the liver
Fig. 5. Immunoblot analysis of rat testis during postnatal development with anti-hnRNP A2/B1 antibody. Samples of rat testis samples of indicated ages were separated by 10% SDS^PAGE and transferred onto a membrane as described in Section 2. Immunoblotting using anti-hnRNP A2/B1 monoclonal antibody DP3 was performed. The positions of the hnRNP A2/B1 isoforms are shown to the right of the ¢gures. Note that 33 kDa bands, which is the expected size of hnRNP B0a, abruptly appear after 4 weeks of age. Asterisked bands are described in Fig. 3.
BBAEXP 93428 25-8-00
were relatively constant during the same period. Immunoblot analysis demonstrated that hnRNP B0 proteins suddenly appeared after the 4th week of age (Fig. 5). Thus, the expression of Ex9(3) products is regulated in a postnatal developmental stage-speci¢c manner, though there was a time lag between the expression of the messages and that of the proteins. 3.3. Histology and immunohistochemistry of rat testis during postnatal development We previously demonstrated the expression patterns of hnRNP A2/B1 and A1 proteins in adult rodent testis [18,23]. We investigated here their expression patterns during postnatal development. Fig. 6 shows the histology with PAS staining and the immunohistochemistry using antiA2/B1 (DP3) and anti-A1 (9H10) monoclonal antibodies. As seen in Fig. 6 (PAS; A, D, G) and described in the literature [30], spermatogonia and spermatocytes occupied the seminiferous tubules of the 3-week-old rat testis. Round spermatids, which have a PAS-positive acrosomal granule [31], appeared in the 4-week-old testis. The 10week-old testis showed a completely mature appearance, containing all di¡erentiating cells from spermatogonia through spermatozoa, though each seminiferous tubule
Cyaan Magenta Geel Zwart
38
M. Matsui et al. / Biochimica et Biophysica Acta 1493 (2000) 33^40
Fig. 6. Histology and immunohistochemistry of rat testis during postnatal development. Testis tissue sections were reacted with PAS reagent and counterstained with hematoxylin (PAS; A, D, G). The cell types are shown as the following abbreviations : Spg, spermatogonium ; Spc, spermatocyte; Rsd, round spermatid; SC, Sertoli cell. Arrows indicate PAS-positive acrosomal granules. The tissue sections were also stained for hnRNP A1 (9H10; B, E, H) and hnRNP A2/B1 (DP3; C, F, I, J) as described in Section 2. Immunohistochemically positive nuclei are stained reddish-brown. Small black triangles in J indicate the spermiogenesis border between steps 8 and 9 [31]. The number of weeks after birth is shown to the left of the ¢gures: 3W, 4W, and 10W. Bar = 10 Wm.
shows an independent stage of spermatogenesis in sexually mature rodents. Immunohistochemically, the expression of hnRNP A1 proteins, detected by monoclonal antibody 9H10, is restricted to the nuclei of spermatogonia and Sertoli cells from 3 to 10 weeks of age (Fig. 6B,E,H). In the 3-weekold testis, hnRNP A2/B1 isoforms, detected by monoclonal antibody DP3, were expressed in the nuclei of spermatogonia, spermatocytes, and Sertoli cells (Fig. 6C), and they were also expressed in round spermatids in the 4and 10-week-old rat testis (Fig. 6F,I). In addition, it should be noted that hnRNP A2/B1 proteins abruptly faded away during spermiogenesis between steps 8 and 9 (Fig. 6J). Thus, hnRNP A1 proteins are restricted to sper-
BBAEXP 93428 25-8-00
matogonia, whereas A2/B1 proteins are expressed from spermatogonia through round spermatids during postnatal development of rat testis. It is impossible to distinguish B0 proteins from A2/B1 ones immunohistochemically because they are produced by alternative splicing and the available antibody recognizes their common epitope. Considering the immunoblotting result (Fig. 5), however, the expression of B0 proteins coincides with the appearance of round spermatids, which were strongly stained with a monoclonal antibody DP3. The reproducibility of RPA, immunoblotting, and immunohistochemical data was con¢rmed by at least three independent experiments using samples from several other rats.
Cyaan Magenta Geel Zwart
M. Matsui et al. / Biochimica et Biophysica Acta 1493 (2000) 33^40
4. Discussion Our results show that the expression of Ex2(3) and Ex2(+) messages is ubiquitous and that alternative splicing, including the second exon of hnRNP A2/B1 gene, constitutively occurs. This is consistent with previously reported data for hnRNP A2/B1 [14,18,32] because the Ex2(3) message is nearly equal to the A2 mRNA and the Ex2(+) to the B1 in most tissues. On the other hand, alternative splicing excluding the ninth exon occurs in a tissue-speci¢c manner. If the alternative splicing pathways involving the second and ninth exons of the A2/B1 gene occur independently and constitutively, the four types of messages (A2, B1, B0a, and B0b) are eventually produced in all tissues. In the sequential study using neonatal rat testes, B0 proteins were abruptly expressed at the fourth week of age, and thereafter their high expression was maintained (Fig. 5), although B0 mRNAs had become evident earlier (Fig. 4). In other words, some quantities of B0 mRNAs are already produced in rat testis after 17 days of age, but not translated until the appearance of round spermatids. In addition, relatively abundant Ex9(3) messages were produced not only in testis but also in thymus and spleen (Fig. 2B,E), yet the expression of Ex9(3), B0, proteins was restricted to the testis (Fig. 3). Therefore, our results suggest the existence of a testis-speci¢c post-transcriptional mechanism that allows the translation of Ex9(3) mRNA, and also suggest that the mechanism does not function in the testis until the appearance of round spermatids. It is known that some mRNAs are transcribed in the earlier stages of spermatogenesis but that their translation occurs in the later stages [25,26,33]. The time lag of the expression between B0 messages and proteins in testis may be related to such a mechanism. In previous studies [18,23] and the present one, we demonstrated that the isoforms of the A2/B1 gene are immunohistochemically expressed through spermatogenesis except for elongated spermatids and spermatozoa, whereas hnRNP A1 is restricted to spermatogonia. The present study also shows that these expression patterns of A2/B1 and A1 proteins in the seminiferous tubules are kept from birth to adulthood. This indicates that their roles di¡er in the testis, although hnRNP A2/B1 and A1 share a quite similar structure and their genes are thought to be evolutionarily duplicated [13,14]. The expression of A2/B1 mRNAs transiently increased after birth through 4 weeks of age, and then declined to the adult level. The A2/B1 gene products, not only B0 but also A2/B1, seem to be involved in postnatal development of the testis, as hnRNP A2/B1 are involved in development of mammalian lung [34]. What role does B0 play in testis ? Our results suggest that B0 participates in spermatogenesis, because B0 proteins appear in a developmental stage of the testis when spermatogenesis becomes mature. The A2/B1 and B0 pro-
BBAEXP 93428 25-8-00
39
teins possess glycine-rich domains that di¡er from each other as a result of the alternative splicing of the ninth exon, which is located in the mid portion of the C-terminal glycine-rich domain (see Fig. 1) [13,14]. It has been demonstrated that the glycine-rich regions are important in protein^protein and protein^nucleic acid interactions [7,35,36]. Therefore, A2/B1 and B0 can interact with different molecules or interact with the same molecule with di¡erent a¤nities and/or speci¢cities in testis. Since the ninth exon of the A2/B1 gene, which is excluded from B0, is structurally equivalent to the 7 bis exon of the A1 gene [13,14,18], the overall structure of B0 corresponds to A1. It is interesting that B0 proteins are probably expressed in later spermatogenesis, when A1 proteins are repressed. In summary, hnRNP B0 proteins are testis-speci¢c splicing isoforms of the hnRNP A2/B1 gene, and the expression of B0 is regulated by a post-transcriptional mechanism. B0 proteins are involved in spermatogenesis and development of the testis, though the molecules interacting with B0 remain to be identi¢ed further. Acknowledgements This work was supported by a grant from Ibaraki Prefectural University of Health Sciences (to H.H.) and by a Japanese Grant-in-Aid for Scienti¢c Research (to H.K.). We thank Dr. Akiyoshi Fukamizu and Dr. Norio Ishida for their interest and useful suggestions.
References [1] G. Dreyfuss, M.J. Matunis, S. Pin¬ol-Roma, C.G. Burd, HnRNP proteins and the biogenesis of mRNA, Annu. Rev. Biochem. 62 (1993) 289^321. [2] S. Nakielny, G. Dreyfuss, Nuclear export of proteins and RNAs, Curr. Opin. Cell Biol. 9 (1997) 420^429. [3] A.R. Krainer (Ed.), Eukaryotic mRNA Processing, IRL Press, Oxford, 1997. [4] H. Siomi, G. Dreyfuss, RNA-binding proteins as regulators of gene expression, Curr. Opin. Genet. Dev. 7 (1997) 345^353. [5] A.M. Krecic, M.S. Swanson, HnRNP complexes: composition, structure, and function, Curr. Opin. Cell Biol. 11 (1999) 363^371. [6] M.S. Swanson, G. Dreyfuss, Classi¢cation and puri¢cation of proteins of heterogeneous nuclear ribonucleoprotein particles by RNAbinding speci¢cities, Mol. Cell. Biol. 8 (1988) 2237^2241. [7] A. Mayeda, S.H. Munroe, J.F. Ca¨ceres, A.R. Krainer, Function of conserved domains of hnRNP A1 and other hnRNP A/B proteins, EMBO J. 13 (1994) 5483^5495. [8] C.G. Burd, M.S. Swanson, M. Go«rlach, G. Dreyfuss, Primary structures of the heterogeneous nuclear ribonucleoprotein A2, B1, and C2 proteins: a diversity of RNA binding proteins is generated by small peptide inserts, Proc. Natl. Acad. Sci. USA 86 (1989) 9788^9792. [9] H. Siomi, G. Dreyfuss, A nuclear localization domain in the hnRNP A1 protein, J. Cell Biol. 129 (1995) 551^560. [10] V.W. Pollard, W.M. Michael, S. Nakielny, M.C. Siomi, F. Wang, G. Dreyfuss, A novel receptor-mediated nuclear protein import pathway, Cell 86 (1996) 985^994.
Cyaan Magenta Geel Zwart
40
M. Matsui et al. / Biochimica et Biophysica Acta 1493 (2000) 33^40
[11] E. Izaurralde, A. Jarmolowski, C. Beisel, I.W. Mattaj, G. Dreyfuss, U. Fischer, A role for the M9 transport signal of hnRNP A1 in mRNA nuclear export, J. Cell Biol. 137 (1997) 27^35. [12] Y.J. Kim, B.S. Baker, Isolation of RRM-type RNA-binding protein genes and the analysis of their relatedness by using a numerical approach, Mol. Cell. Biol. 13 (1993) 174^183. [13] G. Biamonti, M. Ruggiu, S. Saccone, G. Della Valle, S. Riva, Two homologous genes, originated by duplication, encode the human hnRNP proteins A2 and A1, Nucleic Acids Res. 22 (1994) 1996^ 2002. [14] T. Kozu, B. Henrich, K.P. Scha«fer, Structure and expression of the gene (HNRPA2B1) encoding the human hnRNP protein A2/B1, Genomics 25 (1995) 365^371. [15] M. Buvoli, G. Biamonti, P. Tsoulfas, M.T. Bassi, A. Ghetti, S. Riva, C. Morandi, cDNA cloning of human hnRNP protein A1 reveals the existence of multiple mRNA isoforms, Nucleic Acids Res. 16 (1988) 3751^3770. [16] G. Biamonti, M. Buvoli, M.T. Bassi, C. Morandi, F. Cobianchi, S. Riva, Isolation of an active gene encoding human hnRNP protein A1. Evidence for alternative splicing, J. Mol. Biol. 207 (1989) 491^ 503. [17] M. Buvoli, F. Cobianchi, M.G. Bestagno, A. Mangiarotti, M.T. Bassi, G. Biamonti, S. Riva, Alternative splicing in the human gene for the core protein A1 generates another hnRNP protein, EMBO J. 9 (1990) 1229^1235. [18] H. Kamma, H. Horiguchi, L. Wan, M. Matsui, M. Fujiwara, M. Fujimoto, T. Yazawa, G. Dreyfuss, Molecular characterization of the hnRNP A2/B1 proteins: Tissue-speci¢c expression and novel isoforms, Exp. Cell Res. 246 (1999) 399^411. [19] Y. Kajita, J. Nakayama, M. Aizawa, F. Ishikawa, The UUAG-speci¢c RNA binding protein, heterogeneous nuclear ribonucleoprotein D0. Common modular structure and binding properties of the 2xRBD-Gly family, J. Biol. Chem. 270 (1995) 22167^22175. [20] A. Hanamura, J.F. Ca¨ceres, A. Mayeda, B.R. Franza Jr., A.R. Krainer, Regulated tissue-speci¢c expression of antagonistic premRNA splicing factors, RNA 4 (1998) 430^444. [21] A. Mayeda, S.H. Munroe, R.-M. Xu, A.R. Krainer, Distinct functions of the closely related tandem RNA-recognition motifs of hnRNP A1, RNA 4 (1998) 1111^1123. [22] J.H. Carson, S. Kwon, E. Barbarese, RNA tra¤cking in myelinating cells, Curr. Opin. Neurobiol. 8 (1998) 607^612. [23] H. Kamma, D.S. Portman, G. Dreyfuss, Cell type-speci¢c expression of hnRNP proteins, Exp. Cell Res. 221 (1995) 187^196.
BBAEXP 93428 25-8-00
[24] E.M. Eddy, D.A. O'Brien, Gene expression during mammalian meiosis, Curr. Top. Dev. Biol. 37 (1998) 141^200. [25] N.B. Hecht, Molecular mechanisms of male germ cell di¡erentiation, BioEssays 20 (1998) 555^561. [26] J.P. Venables, I.C. Eperon, The roles of RNA-binding proteins in spermatogenesis and male infertility, Curr. Opin. Genet. Dev. 9 (1999) 346^354. [27] D.A. Melton, P.A. Krieg, M.R. Rebagliati, T. Maniatis, K. Zinn, M.R. Green, E¤cient in vitro synthesis of biologically active RNA and RNA hybridization probes from plasmids containing a bacteriophage SP6 promoter, Nucleic Acids Res. 12 (1984) 7035^7056. [28] P. Chomczynski, N. Sacchi, Single-step method of RNA isolation by acid guanidinium thiocyanate-phenol-chloroform extraction, Anal. Biochem. 162 (1987) 156^159. [29] U.K. Laemmli, Cleavage of structural proteins during the assembly of the head of bacteriophage T4, Nature 227 (1970) 680^685. [30] Y. Clermont, B. Perey, Quantitative study of the cell population of the seminiferous tubules in immature rats, Am. J. Anat. 100 (1957) 241^268. [31] L.D. Russell, R.A. Ettlin, A.P. Sinha Hikimm, E.D. Clegg, Histological and Histopathlogical Evaluation of the Testis, Cache River Press, Bolesta, FL, 1990. [32] M. Faura, J. Renau Piqueras, O. Bachs, R. Bosser, Di¡erential distribution of heterogeneous nuclear ribonucleoproteins in rat tissues, Biochem. Biophys. Res. Commun. 217 (1995) 554^560. [33] R.E. Braun, Post-transcriptional control of gene expression during spermatogenesis, Semin. Cell Dev. Biol. 9 (1998) 483^489. [34] L.M. Montuenga, J. Zhou, I. Avis, M. Vos, A. Martinez, F. Cuttitta, A.M. Treston, M. Sunday, J.L. Mulshine, Expression of heterogeneous nuclear ribonucleoprotein A2/B1 changes with critical stages of mammalian lung development, Am. J. Respir. Cell Mol. Biol. 19 (1998) 554^562. [35] J.R. Casas-Finet, J.D. Smith Jr., A. Kumar, J.G. Kim, S.H. Wilson, R.L. Karpel, Mammalian heterogeneous ribonucleoprotein A1 and its constituent domains. Nucleic acid interaction, structural stability and self-association, J. Mol. Biol. 229 (1993) 873^889. [36] L. Cartegni, M. Maconi, E. Morandi, F. Cobianchi, S. Riva, G. Biamonti, hnRNP A1 selectively interacts through its Gly-rich domain with di¡erent RNA-binding proteins, J. Mol. Biol. 259 (1996) 337^348.
Cyaan Magenta Geel Zwart
www.elsevier.com/locate/bba
Testis- and developmental stage-speci¢c expression of hnRNP A2/B1 splicing isoforms, B0a/b Miwa Matsui a , Hisashi Horiguchi a; *, Hiroshi Kamma b , Masachika Fujiwara c , Rieko Ohtsubo a , Takesaburo Ogata a b c
a Center for Medical Sciences, Ibaraki Prefectural University of Health Sciences, Ibaraki 300-0394, Japan Department of Pathology, Institute of Basic Medical Sciences, University of Tsukuba, Ibaraki 305-8575, Japan Department of Pathology, Kensei General Hospital, 604 Sukita, Iwase, Nishi-Ibaraki, Ibaraki 309-1295, Japan
Received 29 February 2000; received in revised form 15 May 2000 ; accepted 26 May 2000
Abstract Heterogeneous nuclear ribonucleoproteins (hnRNPs) A2 and B1 are abundant nuclear proteins that bind to nascent RNAs synthesized by RNA polymerase II. Previously we had found that the splicing isoforms hnRNP B0a/b, from which the ninth exon of the A2/B1 gene is excluded, are abundantly expressed in testis. We postulated that B0a/b are testis-specific isoforms, and investigated the expression of A2/B1 and B0a/b in rat tissues and in postnatal development of rat testes using RNase protection assay, immunoblotting, and immunohistochemistry. We found that hnRNP B0a/b mRNAs are expressed in several tissues but that the testis alone expresses B0a/b proteins. A sequential study using neonatal rat testes demonstrated that B0a/b mRNAs are produced after 17 days of age, but not translated until 4 weeks of age when round spermatids appear in addition to spermatogonia and spermatocytes. Immunohistochemically, hnRNP A2/B1 isoforms are expressed during spermatogenesis from spermatogonia through round spermatids, whereas the expression of A1 is restricted to spermatogonia. This expression pattern in the rat testis is maintained from birth through adulthood. These results suggest that the expression of the hnRNP A2/B1 gene is partly regulated by a testis-specific post-transcriptional mechanism, and that the products of the A2/B1 gene, especially hnRNP B0a/b, are involved in spermatogenesis. ß 2000 Elsevier Science B.V. All rights reserved. Keywords : Heterogeneous nuclear ribonucleoprotein A2/B1; Heterogeneous nuclear ribonucleoprotein B0a/b ; Splicing isoform; Tissue-speci¢c expression; Testis; Spermatogenesis
1. Introduction Heterogeneous nuclear ribonucleoproteins (hnRNPs) bind to nascent RNAs synthesized by RNA polymerase II and are thought to function in a wide range of cellular activities, such as pre-mRNA processing, mRNA localization, translation, and mRNA turnover [1^3]. In recent years, post-transcriptional as well as transcriptional mechanisms have attracted considerable attention in the regulation of gene expression. RNA binding proteins, including hnRNPs, play central roles in the post-transcriptional regulation of gene expression [4,5].
* Corresponding author. Fax: +81-298-40-2313; E-mail : [email protected]
HnRNPs were ¢rst identi¢ed as a major group of chromatin-associated RNA binding proteins [6]. Subsequently, the major hnRNPs were cloned and characterized, and about 30 hnRNP proteins have been identi¢ed and are designated from A1 to U at present [1,5]. Of these, there is a group of basic hnRNP proteins collectively called the A/B group proteins [1,7]. HnRNP A1, A2, and B1 proteins are its main members. The structural feature of these proteins is that they consist of two conserved amino-terminal RNA binding domains (RBDs, or RNA recognition motifs) linked to a glycine-rich domain that contains an RGG box and a nuclear localization signal termed M9 [1,8^11]. A2/B1 and A1 proteins are quite similar in their gene and protein structures, and they are thought to be evolutionarily duplicated [12^14]. A1 protein is produced from multiple genes [15,16], and its splicing isoform A1B is produced by insertion of the 7 bis exon [17]. A2 and B1 proteins are produced by alternative splicing from a single-copy gene, and they di¡er from each other only by a
0167-4781 / 00 / $ ^ see front matter ß 2000 Elsevier Science B.V. All rights reserved. PII: S 0 1 6 7 - 4 7 8 1 ( 0 0 ) 0 0 1 5 4 - 8
BBAEXP 93428 25-8-00
Cyaan Magenta Geel Zwart
34
M. Matsui et al. / Biochimica et Biophysica Acta 1493 (2000) 33^40
12-amino-acid insertion in the N-terminal RBD of B1 [8,13,14]. Recently, we isolated the splicing isoforms of hnRNP A2/B1, named B0a/b [18]. The ninth exon of the A2/B1 gene, which is structurally equivalent to the 7 bis exon of A1 gene, is excluded in B0a/b [13,14,18]. Thus, the overall structure of the A2/B1 gene is still more analogous to that of the A1 gene. A2/B1 and A1 proteins also share some functional similarities, such as hnRNP complex formation [1] and in vitro UUAG-speci¢c RNA binding [19]. They in£uence in vitro alternative 5P splicing site selection as antagonists of SR proteins [7,20,21]. Due to these features, A/B group proteins are thought to be ubiquitously expressed in all cells and constitutively function through their interactions with other molecules. It is still unclear why the diversity of A/B group proteins exists or what di¡ers among them. Recent studies, however, suggest a cell type-speci¢c function of hnRNP A2 in spite of its ubiquitous expression. It has been demonstrated that hnRNP A2 speci¢cally binds myelin basic protein mRNA in cultured myelinating cells and plays a central role in its transport, localization, and translation [22]. In addition, we have previously demonstrated the expression patterns of A2/B1 and A1 in the seminiferous tubules, suggesting they have di¡erent roles in spermatogenesis [18,23]. We also found that the splicing isoforms B0a/b are abundantly expressed in testis [18]. It is known that some RNA binding proteins are essential for spermatogenesis, and that some genes produce testis-speci¢c isoforms by alternative splicing or polyadenylation [24^26]. These observations led us to postulate that B0a/b are testis-speci¢c isoforms that play an important role in testis. To address this supposition, we investigated the expression of A2/B1 gene products among rat tissues and in postnatal rat testes from birth to adulthood, using RNase protection assay (RPA), immunoblotting, and immunohistochemistry. In this study, we demonstrated that the expression of hnRNP B0a/b, splicing isoforms of hnRNP A2/B1, is regulated in a testis-speci¢c manner, and that B0a/b are expressed in postnatal development of the testis. 2. Materials and methods
and frozen with liquid nitrogen to extract mRNAs and proteins. At the same time, the testis tissues were also ¢xed in Bouin's solution (saturated picric acid:formaldehyde: acetic acid = 75:25:5) for histological and immunohistochemical analysis (see below). 2.2. Preparation of probes and RPA We used a plasmid DNA containing the rat hnRNP B1 coding sequence previously cloned [18] (GenBank/EMBL/ DDBJ accession number AB006816). A rat hnRNP B1 cDNA 5P fragment digested with EcoRI and HinfI (nt 1^ 306, probe 1 in Fig. 1), encompassing the hnRNP B1-speci¢c region, i.e., exon 2, was cloned in a pBluescript II KS+ vector (pBSII KS+; Stratagene, La Jolla, CA, USA) digested with EcoRI and EcoRV. PCR ampli¢cation was used to isolate a cDNA fragment of rat hnRNP B1 (nt 468^811, probe 2 in Fig. 1), containing part of the exon 9 region, and it was subcloned into a pBluescript SK+ vector (pBS SK+ ; Stratagene) digested with BamHI and SalI. The pTRI RNA 18S antisense control template (Ambion, Austin, TX, USA) was used as internal control and calibration. To synthesize the K-32 P-labeled antisense RNA probes, the plasmid DNAs of probe 1 and probe 2 were linearized with the EcoRI site in the pBSII KS+ polylinker and with the BamHI site in the pBS SK+ polylinker, and transcribed with T3 and T7 RNA polymerase in the presence of [K-32 P]UTP, respectively [27]. Total RNAs of rat tissues were extracted and puri¢ed by the acidic guanidinium isothiocyanate^phenol^chloroform method [28]. RPAs were performed using a Ribonuclease Protection Assay kit (RPAII ; Ambion). The probes were hybridized to 25 or 10 Wg of total RNA from the rat tissues at 42³C for 16^18 h. After hybridization, the mixtures were digested with RNase solution containing RNase A/RNase T1 mixture in quantities recommended by the manufacturer. Protected fragments were analyzed on 5% denaturing polyacrylamide gels in 1UTBE bu¡er and visualized by autoradiography. The image analysis of the autoradiograms was performed using a model GS-700 Imaging Densitometer (Bio-Rad Laboratories, Hercules, CA, USA). 2.3. Gel electrophoresis and immunoblotting
2.1. Source of tissue samples To investigate di¡erential expression among tissues, male and female Wistar rats aged 10 weeks were purchased from Japan SLC, Inc. (Hamamatsu, Japan). For sequential analysis of postnatal development of testes, male Wistar rats aged 3, 4, 5, and 10 weeks, and pregnant Wistar rats were also purchased. The pregnant rats were maintained in our laboratory, and their male neonates were used in experiments 1, 5, 10, 14, and 17 days after birth. They were anesthetized with pentobarbital sodium and killed, and various tissues were immediately removed
BBAEXP 93428 25-8-00
Total protein extracts from rat tissues were prepared and adjusted to contain approximately the same total amount of the protein. Protein samples (10 Wg) were separated on 10% SDS^polyacrylamide gels (SDS-PAGE) [29]. Immunoblotting was performed and probed with anti-hnRNP A2/B1 (DP3) mouse monoclonal antibody in 1:5000 dilution as previously described [18]. Bound antibodies were detected with peroxidase-conjugated anti-mouse immunoglobulin antibody (Amersham Pharmacia Biotech, Uppsala, Sweden) and the ECL system (Amersham Pharmacia Biotech).
Cyaan Magenta Geel Zwart
M. Matsui et al. / Biochimica et Biophysica Acta 1493 (2000) 33^40
2.4. Histological and immunohistochemical analysis The ¢xed testis tissues were embedded in para¤n, serially cut at 3 Wm thickness, and mounted on glass slides. Depara¤nized tissue sections were reacted with periodic acid^Schi¡ (PAS) reagent and counterstained with hematoxylin. For immunohistochemistry, the mouse monoclonal antibodies used were anti-hnRNP A2/B1 (DP3 [18]) in 1:500 dilution and anti-hnRNP A1 (9H10; kindly provided by Dr. G. Dreyfuss) in 1:500 dilution. We used a Catalyzed Signal Ampli¢cation system (CSA system; Dako Corporation, Carpinteria, CA, USA) with slight modi¢cation to the manufacturer's instruction. Brie£y, the depara¤nized tissue sections were pretreated in a microwave oven at 95³C for 1 min in a target retrieval solution (provided by the manufacturer, based on 10 mM citrate bu¡er, pH 6.0) and incubated with the above ¢rst antibodies for 1 h at room temperature. Bound antibodies were detected by the CSA system. Control sections were incubated with non-immunized mouse serum substituted for the primary antibodies. Aminoethyl-carbamazole (Aldrich, Milwaukee, WI, USA) was used as a coloring substrate for detection of peroxidase, and followed by counterstaining with hematoxylin. Stained specimens were examined using a microscope. 3. Results 3.1. Testis-speci¢c expression of hnRNP B0 Here we use the terms Ex2(+), Ex2(3), Ex9(+), and Ex9(3) to distinguish the splicing sites of the products
35
of the hnRNP A2/B1 gene. For example, Ex2(+) designates the product into which the second exon was spliced. We also use the term `B0' for the Ex9(3) isoform, since the B0a (Ex2(3), Ex9(3)) and B0b (Ex2(+), Ex9(3)) mRNAs are indistinguishable at the same time because of the limitation of the probe length in RPA (Fig. 1). We regarded the sum of Ex2(+) and Ex2(3) or Ex9(+) and Ex9(3) mRNAs as the total amount of hnRNP A2/ B1 gene-derived transcripts and applied the term `total A2/ B1 mRNA' to it. The quanti¢ed relative values of RPA among rat tissues are shown in Fig. 2. The protection probe 1 produced protected bands of 306 nt and 264 nt, corresponding to Ex2(+) and Ex2(3) mRNA respectively, in all rat tissues examined (Fig. 2A). Although the amounts of Ex2(+) and Ex2(3) mRNAs varied across di¡erent tissues, the relative amounts of Ex2(+) to total A2/B1 mRNAs were nearly constant, ranging from 0.05 to 0.18 (mean 0.13) (Fig. 2D). This indicates that the alternative splicing producing the Ex2(+) messages correlates with the amounts of total A2/B1 mRNAs. The protection probe 2 produced protected bands of 344 nt and 291 nt, corresponding to Ex9(+) and Ex9(3) mRNA respectively. The Ex9(+) mRNAs, meaning the sum of the original hnRNP A2 and B1 mRNAs, were identi¢ed in all rat tissues as expected, whereas the expression levels of the Ex9(3) mRNAs varied considerably among the di¡erent tissues, i.e., the Ex9(3) messages were abundant in testis and thymus, less abundant in spleen and intestine, and much rarer or undetectable in other tissues (Fig. 2B,E). Immunoblotting of proteins from rat tissues and HeLa cells detected proteins of the expected sizes for hnRNP A2 (36 kDa) and B1 (38 kDa).
Fig. 1. Structure of the hnRNP A2/B1 and B0a/b transcripts, coding regions. Regions that correspond to the structural or functional domains of the gene product are shown: RBD-I, RNA binding domain I; RBD-II, RNA binding domain II; RGG motifs, R-G-G repeat-containing sequence; M9, nuclear localization signal. The conserved sequence in hnRNP B1, corresponding to exon 2 (Ex2), is indicated by the small hatched box. The excluded sequence in the glycine-rich domain is indicated by the small black box, corresponding to the ninth exon of the human hnRNP A2/B1 gene (Ex9). Probe 1 and 2 used in RPA are designed to distinguish the alternatively spliced mRNAs of the exons 2 and 9 respectively. The numbers indicate the nucleotide sequence positions of hnRNP A2/B1 from the initiation codon, ATG.
BBAEXP 93428 25-8-00
Cyaan Magenta Geel Zwart
36
M. Matsui et al. / Biochimica et Biophysica Acta 1493 (2000) 33^40
Fig. 2. RPA among rat tissues. 25 Wg total RNA of indicated rat tissues was analyzed by RPA with probe 1 (A) and probe 2 (B); as a control, 0.5 Wg total RNA of those tissues was analyzed with an 18S ribosomal RNA probe (C) as described in Section 2. The no target controls, without and with RNase treatment, are seen on the left side of A^C as `probe' and `control' respectively. The probes show the expected results ; each primarily full-length probe is 368 bp (A), 415 bp (B), and 128 bp (C). The controls show no protected signals as expected. The size markers are also demonstrated in A^C: `100 bp ladder', [Q-32 P]ATP-labeled 100 bp DNA ladder. The protected messages in A and B were quanti¢ed as described in Section 2, using the 18S ribosomal RNA signals as calibration standard. (D, E) The ratio of each optical density to the sum of the lung Ex2(+) and Ex2(3) optical densities in A, or Ex9(+) and Ex9(3) in B, was calculated as the relative optical density. The estimated values of the Ex2(+) (306 bp, hatched bars) and Ex2(3) (264 bp, open bars) mRNAs are presented in D, and those of the Ex9(+) (344 bp, open bars) and Ex9(3) (291 bp, hatched bars) in E. The height of the bar, i.e., open bar plus hatched bar, in D and E indicates the total amount of A2/B1 messages, and the numbers above the bars denote the ratio of the Ex2(+) or Ex9(3) to the total amount of the messages.
It also showed that the testis alone contained an additional 33 kDa band, which is the expected size of splicing isoform B0, speci¢cally B0a excluding the second exon (Fig. 3). B0b protein was scarcely detected in rat testis, but it is di¤cult to visualize under the same immunoblot condition as Fig. 3 (data not shown). Thus, the exclusion of the ninth exon occurs in a tissue-speci¢c manner, and the expression of B0 proteins is testis-speci¢c. 3.2. Developmental stage-speci¢c expression of hnRNP B0 in rat testis We next investigated the mRNA expression levels of hnRNP A2/B1 isoforms during postnatal development of the rat testis from birth to adulthood (Fig. 4). The total A2/B1 mRNAs gradually increased after birth to 4 weeks of age, and then the expression level decreased to the adult one. During this period, the Ex2(+) mRNAs increased in amount almost proportionally to the total A2/B1 mRNAs (Fig. 4A,D), which indicates that the alternative splicing producing the Ex2(+) messages is constitutive in postnatal development of the testis.
BBAEXP 93428 25-8-00
Fig. 3. Immunoblot analysis of rat tissues with anti-hnRNP A2/B1 antibody. Samples of indicated rat tissues and HeLa cells were separated by electrophoresis in SDS^10% polyacrylamide gel and transferred onto a membrane as described in Section 2. Immunoblotting using anti-hnRNP A2/B1 monoclonal antibody DP3 was performed. The positions of the hnRNP A2/B1 isoforms are shown to the right of the ¢gure. An additional 39 kDa band (asterisk) is supposed to be another isoform of hnRNP A2/B1 or to be post-translationally modi¢ed hnRNP A2/B1 proteins [18].
Cyaan Magenta Geel Zwart
M. Matsui et al. / Biochimica et Biophysica Acta 1493 (2000) 33^40
37
Fig. 4. RPA of rat testis during postnatal developmental. 10 Wg total RNA of rat testis and liver tissues at indicated ages was analyzed by RPA with probe 1 (A), probe 2 (B), and 0.5 Wg total RNA of those tissues was analyzed with a 18S ribosomal RNA probe (C) as described in Section 2. The size markers and controls are also demonstrated in A^C in the same manner as in Fig. 2. The protected messages in A and B were quanti¢ed as described in Section 2, using the 18S ribosomal RNA signals as calibration standard. (D, E) The ratio of each optical density to the sum of the 10-week-old testis Ex2(+) and Ex2(3) optical densities in A, or Ex9(+) and Ex9(3) in B, was calculated as the relative optical density. The estimated values of the mRNAs are presented in D and E in the same manner as in Fig. 2.
However, the Ex9(3) mRNAs became evident by the 17th day after birth, and they gradually increased in amount from then until 4 weeks of age (Fig. 4B,E). The relative amounts of Ex9(3) mRNAs decreased after the 5th week, but Ex9(3) mRNAs still constituted about 40% of total A2/B1 mRNAs. It was shown that the proportion of Ex9(3) to total A2/B1 mRNAs reached the level of the adult rat by the fourth week after birth. In contrast, the expression levels of hnRNP A2/B1 mRNAs in the liver
Fig. 5. Immunoblot analysis of rat testis during postnatal development with anti-hnRNP A2/B1 antibody. Samples of rat testis samples of indicated ages were separated by 10% SDS^PAGE and transferred onto a membrane as described in Section 2. Immunoblotting using anti-hnRNP A2/B1 monoclonal antibody DP3 was performed. The positions of the hnRNP A2/B1 isoforms are shown to the right of the ¢gures. Note that 33 kDa bands, which is the expected size of hnRNP B0a, abruptly appear after 4 weeks of age. Asterisked bands are described in Fig. 3.
BBAEXP 93428 25-8-00
were relatively constant during the same period. Immunoblot analysis demonstrated that hnRNP B0 proteins suddenly appeared after the 4th week of age (Fig. 5). Thus, the expression of Ex9(3) products is regulated in a postnatal developmental stage-speci¢c manner, though there was a time lag between the expression of the messages and that of the proteins. 3.3. Histology and immunohistochemistry of rat testis during postnatal development We previously demonstrated the expression patterns of hnRNP A2/B1 and A1 proteins in adult rodent testis [18,23]. We investigated here their expression patterns during postnatal development. Fig. 6 shows the histology with PAS staining and the immunohistochemistry using antiA2/B1 (DP3) and anti-A1 (9H10) monoclonal antibodies. As seen in Fig. 6 (PAS; A, D, G) and described in the literature [30], spermatogonia and spermatocytes occupied the seminiferous tubules of the 3-week-old rat testis. Round spermatids, which have a PAS-positive acrosomal granule [31], appeared in the 4-week-old testis. The 10week-old testis showed a completely mature appearance, containing all di¡erentiating cells from spermatogonia through spermatozoa, though each seminiferous tubule
Cyaan Magenta Geel Zwart
38
M. Matsui et al. / Biochimica et Biophysica Acta 1493 (2000) 33^40
Fig. 6. Histology and immunohistochemistry of rat testis during postnatal development. Testis tissue sections were reacted with PAS reagent and counterstained with hematoxylin (PAS; A, D, G). The cell types are shown as the following abbreviations : Spg, spermatogonium ; Spc, spermatocyte; Rsd, round spermatid; SC, Sertoli cell. Arrows indicate PAS-positive acrosomal granules. The tissue sections were also stained for hnRNP A1 (9H10; B, E, H) and hnRNP A2/B1 (DP3; C, F, I, J) as described in Section 2. Immunohistochemically positive nuclei are stained reddish-brown. Small black triangles in J indicate the spermiogenesis border between steps 8 and 9 [31]. The number of weeks after birth is shown to the left of the ¢gures: 3W, 4W, and 10W. Bar = 10 Wm.
shows an independent stage of spermatogenesis in sexually mature rodents. Immunohistochemically, the expression of hnRNP A1 proteins, detected by monoclonal antibody 9H10, is restricted to the nuclei of spermatogonia and Sertoli cells from 3 to 10 weeks of age (Fig. 6B,E,H). In the 3-weekold testis, hnRNP A2/B1 isoforms, detected by monoclonal antibody DP3, were expressed in the nuclei of spermatogonia, spermatocytes, and Sertoli cells (Fig. 6C), and they were also expressed in round spermatids in the 4and 10-week-old rat testis (Fig. 6F,I). In addition, it should be noted that hnRNP A2/B1 proteins abruptly faded away during spermiogenesis between steps 8 and 9 (Fig. 6J). Thus, hnRNP A1 proteins are restricted to sper-
BBAEXP 93428 25-8-00
matogonia, whereas A2/B1 proteins are expressed from spermatogonia through round spermatids during postnatal development of rat testis. It is impossible to distinguish B0 proteins from A2/B1 ones immunohistochemically because they are produced by alternative splicing and the available antibody recognizes their common epitope. Considering the immunoblotting result (Fig. 5), however, the expression of B0 proteins coincides with the appearance of round spermatids, which were strongly stained with a monoclonal antibody DP3. The reproducibility of RPA, immunoblotting, and immunohistochemical data was con¢rmed by at least three independent experiments using samples from several other rats.
Cyaan Magenta Geel Zwart
M. Matsui et al. / Biochimica et Biophysica Acta 1493 (2000) 33^40
4. Discussion Our results show that the expression of Ex2(3) and Ex2(+) messages is ubiquitous and that alternative splicing, including the second exon of hnRNP A2/B1 gene, constitutively occurs. This is consistent with previously reported data for hnRNP A2/B1 [14,18,32] because the Ex2(3) message is nearly equal to the A2 mRNA and the Ex2(+) to the B1 in most tissues. On the other hand, alternative splicing excluding the ninth exon occurs in a tissue-speci¢c manner. If the alternative splicing pathways involving the second and ninth exons of the A2/B1 gene occur independently and constitutively, the four types of messages (A2, B1, B0a, and B0b) are eventually produced in all tissues. In the sequential study using neonatal rat testes, B0 proteins were abruptly expressed at the fourth week of age, and thereafter their high expression was maintained (Fig. 5), although B0 mRNAs had become evident earlier (Fig. 4). In other words, some quantities of B0 mRNAs are already produced in rat testis after 17 days of age, but not translated until the appearance of round spermatids. In addition, relatively abundant Ex9(3) messages were produced not only in testis but also in thymus and spleen (Fig. 2B,E), yet the expression of Ex9(3), B0, proteins was restricted to the testis (Fig. 3). Therefore, our results suggest the existence of a testis-speci¢c post-transcriptional mechanism that allows the translation of Ex9(3) mRNA, and also suggest that the mechanism does not function in the testis until the appearance of round spermatids. It is known that some mRNAs are transcribed in the earlier stages of spermatogenesis but that their translation occurs in the later stages [25,26,33]. The time lag of the expression between B0 messages and proteins in testis may be related to such a mechanism. In previous studies [18,23] and the present one, we demonstrated that the isoforms of the A2/B1 gene are immunohistochemically expressed through spermatogenesis except for elongated spermatids and spermatozoa, whereas hnRNP A1 is restricted to spermatogonia. The present study also shows that these expression patterns of A2/B1 and A1 proteins in the seminiferous tubules are kept from birth to adulthood. This indicates that their roles di¡er in the testis, although hnRNP A2/B1 and A1 share a quite similar structure and their genes are thought to be evolutionarily duplicated [13,14]. The expression of A2/B1 mRNAs transiently increased after birth through 4 weeks of age, and then declined to the adult level. The A2/B1 gene products, not only B0 but also A2/B1, seem to be involved in postnatal development of the testis, as hnRNP A2/B1 are involved in development of mammalian lung [34]. What role does B0 play in testis ? Our results suggest that B0 participates in spermatogenesis, because B0 proteins appear in a developmental stage of the testis when spermatogenesis becomes mature. The A2/B1 and B0 pro-
BBAEXP 93428 25-8-00
39
teins possess glycine-rich domains that di¡er from each other as a result of the alternative splicing of the ninth exon, which is located in the mid portion of the C-terminal glycine-rich domain (see Fig. 1) [13,14]. It has been demonstrated that the glycine-rich regions are important in protein^protein and protein^nucleic acid interactions [7,35,36]. Therefore, A2/B1 and B0 can interact with different molecules or interact with the same molecule with di¡erent a¤nities and/or speci¢cities in testis. Since the ninth exon of the A2/B1 gene, which is excluded from B0, is structurally equivalent to the 7 bis exon of the A1 gene [13,14,18], the overall structure of B0 corresponds to A1. It is interesting that B0 proteins are probably expressed in later spermatogenesis, when A1 proteins are repressed. In summary, hnRNP B0 proteins are testis-speci¢c splicing isoforms of the hnRNP A2/B1 gene, and the expression of B0 is regulated by a post-transcriptional mechanism. B0 proteins are involved in spermatogenesis and development of the testis, though the molecules interacting with B0 remain to be identi¢ed further. Acknowledgements This work was supported by a grant from Ibaraki Prefectural University of Health Sciences (to H.H.) and by a Japanese Grant-in-Aid for Scienti¢c Research (to H.K.). We thank Dr. Akiyoshi Fukamizu and Dr. Norio Ishida for their interest and useful suggestions.
References [1] G. Dreyfuss, M.J. Matunis, S. Pin¬ol-Roma, C.G. Burd, HnRNP proteins and the biogenesis of mRNA, Annu. Rev. Biochem. 62 (1993) 289^321. [2] S. Nakielny, G. Dreyfuss, Nuclear export of proteins and RNAs, Curr. Opin. Cell Biol. 9 (1997) 420^429. [3] A.R. Krainer (Ed.), Eukaryotic mRNA Processing, IRL Press, Oxford, 1997. [4] H. Siomi, G. Dreyfuss, RNA-binding proteins as regulators of gene expression, Curr. Opin. Genet. Dev. 7 (1997) 345^353. [5] A.M. Krecic, M.S. Swanson, HnRNP complexes: composition, structure, and function, Curr. Opin. Cell Biol. 11 (1999) 363^371. [6] M.S. Swanson, G. Dreyfuss, Classi¢cation and puri¢cation of proteins of heterogeneous nuclear ribonucleoprotein particles by RNAbinding speci¢cities, Mol. Cell. Biol. 8 (1988) 2237^2241. [7] A. Mayeda, S.H. Munroe, J.F. Ca¨ceres, A.R. Krainer, Function of conserved domains of hnRNP A1 and other hnRNP A/B proteins, EMBO J. 13 (1994) 5483^5495. [8] C.G. Burd, M.S. Swanson, M. Go«rlach, G. Dreyfuss, Primary structures of the heterogeneous nuclear ribonucleoprotein A2, B1, and C2 proteins: a diversity of RNA binding proteins is generated by small peptide inserts, Proc. Natl. Acad. Sci. USA 86 (1989) 9788^9792. [9] H. Siomi, G. Dreyfuss, A nuclear localization domain in the hnRNP A1 protein, J. Cell Biol. 129 (1995) 551^560. [10] V.W. Pollard, W.M. Michael, S. Nakielny, M.C. Siomi, F. Wang, G. Dreyfuss, A novel receptor-mediated nuclear protein import pathway, Cell 86 (1996) 985^994.
Cyaan Magenta Geel Zwart
40
M. Matsui et al. / Biochimica et Biophysica Acta 1493 (2000) 33^40
[11] E. Izaurralde, A. Jarmolowski, C. Beisel, I.W. Mattaj, G. Dreyfuss, U. Fischer, A role for the M9 transport signal of hnRNP A1 in mRNA nuclear export, J. Cell Biol. 137 (1997) 27^35. [12] Y.J. Kim, B.S. Baker, Isolation of RRM-type RNA-binding protein genes and the analysis of their relatedness by using a numerical approach, Mol. Cell. Biol. 13 (1993) 174^183. [13] G. Biamonti, M. Ruggiu, S. Saccone, G. Della Valle, S. Riva, Two homologous genes, originated by duplication, encode the human hnRNP proteins A2 and A1, Nucleic Acids Res. 22 (1994) 1996^ 2002. [14] T. Kozu, B. Henrich, K.P. Scha«fer, Structure and expression of the gene (HNRPA2B1) encoding the human hnRNP protein A2/B1, Genomics 25 (1995) 365^371. [15] M. Buvoli, G. Biamonti, P. Tsoulfas, M.T. Bassi, A. Ghetti, S. Riva, C. Morandi, cDNA cloning of human hnRNP protein A1 reveals the existence of multiple mRNA isoforms, Nucleic Acids Res. 16 (1988) 3751^3770. [16] G. Biamonti, M. Buvoli, M.T. Bassi, C. Morandi, F. Cobianchi, S. Riva, Isolation of an active gene encoding human hnRNP protein A1. Evidence for alternative splicing, J. Mol. Biol. 207 (1989) 491^ 503. [17] M. Buvoli, F. Cobianchi, M.G. Bestagno, A. Mangiarotti, M.T. Bassi, G. Biamonti, S. Riva, Alternative splicing in the human gene for the core protein A1 generates another hnRNP protein, EMBO J. 9 (1990) 1229^1235. [18] H. Kamma, H. Horiguchi, L. Wan, M. Matsui, M. Fujiwara, M. Fujimoto, T. Yazawa, G. Dreyfuss, Molecular characterization of the hnRNP A2/B1 proteins: Tissue-speci¢c expression and novel isoforms, Exp. Cell Res. 246 (1999) 399^411. [19] Y. Kajita, J. Nakayama, M. Aizawa, F. Ishikawa, The UUAG-speci¢c RNA binding protein, heterogeneous nuclear ribonucleoprotein D0. Common modular structure and binding properties of the 2xRBD-Gly family, J. Biol. Chem. 270 (1995) 22167^22175. [20] A. Hanamura, J.F. Ca¨ceres, A. Mayeda, B.R. Franza Jr., A.R. Krainer, Regulated tissue-speci¢c expression of antagonistic premRNA splicing factors, RNA 4 (1998) 430^444. [21] A. Mayeda, S.H. Munroe, R.-M. Xu, A.R. Krainer, Distinct functions of the closely related tandem RNA-recognition motifs of hnRNP A1, RNA 4 (1998) 1111^1123. [22] J.H. Carson, S. Kwon, E. Barbarese, RNA tra¤cking in myelinating cells, Curr. Opin. Neurobiol. 8 (1998) 607^612. [23] H. Kamma, D.S. Portman, G. Dreyfuss, Cell type-speci¢c expression of hnRNP proteins, Exp. Cell Res. 221 (1995) 187^196.
BBAEXP 93428 25-8-00
[24] E.M. Eddy, D.A. O'Brien, Gene expression during mammalian meiosis, Curr. Top. Dev. Biol. 37 (1998) 141^200. [25] N.B. Hecht, Molecular mechanisms of male germ cell di¡erentiation, BioEssays 20 (1998) 555^561. [26] J.P. Venables, I.C. Eperon, The roles of RNA-binding proteins in spermatogenesis and male infertility, Curr. Opin. Genet. Dev. 9 (1999) 346^354. [27] D.A. Melton, P.A. Krieg, M.R. Rebagliati, T. Maniatis, K. Zinn, M.R. Green, E¤cient in vitro synthesis of biologically active RNA and RNA hybridization probes from plasmids containing a bacteriophage SP6 promoter, Nucleic Acids Res. 12 (1984) 7035^7056. [28] P. Chomczynski, N. Sacchi, Single-step method of RNA isolation by acid guanidinium thiocyanate-phenol-chloroform extraction, Anal. Biochem. 162 (1987) 156^159. [29] U.K. Laemmli, Cleavage of structural proteins during the assembly of the head of bacteriophage T4, Nature 227 (1970) 680^685. [30] Y. Clermont, B. Perey, Quantitative study of the cell population of the seminiferous tubules in immature rats, Am. J. Anat. 100 (1957) 241^268. [31] L.D. Russell, R.A. Ettlin, A.P. Sinha Hikimm, E.D. Clegg, Histological and Histopathlogical Evaluation of the Testis, Cache River Press, Bolesta, FL, 1990. [32] M. Faura, J. Renau Piqueras, O. Bachs, R. Bosser, Di¡erential distribution of heterogeneous nuclear ribonucleoproteins in rat tissues, Biochem. Biophys. Res. Commun. 217 (1995) 554^560. [33] R.E. Braun, Post-transcriptional control of gene expression during spermatogenesis, Semin. Cell Dev. Biol. 9 (1998) 483^489. [34] L.M. Montuenga, J. Zhou, I. Avis, M. Vos, A. Martinez, F. Cuttitta, A.M. Treston, M. Sunday, J.L. Mulshine, Expression of heterogeneous nuclear ribonucleoprotein A2/B1 changes with critical stages of mammalian lung development, Am. J. Respir. Cell Mol. Biol. 19 (1998) 554^562. [35] J.R. Casas-Finet, J.D. Smith Jr., A. Kumar, J.G. Kim, S.H. Wilson, R.L. Karpel, Mammalian heterogeneous ribonucleoprotein A1 and its constituent domains. Nucleic acid interaction, structural stability and self-association, J. Mol. Biol. 229 (1993) 873^889. [36] L. Cartegni, M. Maconi, E. Morandi, F. Cobianchi, S. Riva, G. Biamonti, hnRNP A1 selectively interacts through its Gly-rich domain with di¡erent RNA-binding proteins, J. Mol. Biol. 259 (1996) 337^348.
Cyaan Magenta Geel Zwart