- Email: [email protected]
Cytoplasmic Localization of Cyclin D3 in Seminiferous Tubules during Testicular Development
EXPERIMENTAL CELL RESEARCH ARTICLE NO.
234, 27–36 (1997)
EX973590
Cytoplasmic Localization of Cyclin D3 in Seminiferous Tubules during Testicular Development Min Jeong Kang,* Mi Kyung Kim,† Amy Terhune,* Jong Kwan Park,‡ Young Ho Kim,† and Gou Young Koh*,1 *Department of Physiology and Institute of Cardiovascular Research and ‡Department of Urology, Chonbuk National University Medical School, Chonju, 560-180, Republic of Korea; †Department of Microbiology, College of Natural Sciences, Kyungpook National University, Taegu 702-701, Republic of Korea
in the cytoplasm lead us to speculate that cyclin D3 may have functions in male germ cells other than mitosis. q 1997 Academic Press
Using a newly developed polyclonal antibody against murine cyclin D3, we have found that protein levels of cyclin D3 were highly detectable only in thymus and testis in rats. Since testis offer unique opportunities to examine the cell cycle in vivo, we examined the temporal and spatial expression of cyclin D3 and the DNA synthesis indicator, proliferating cell nuclear antigen (PCNA), in the rat testis during development. The protein levels of cyclin D3 protein in testis from 7 days to 3 months old were almost constant and then decreased gradually thereafter. The protein levels of cyclin D1 and PCNA were high in the testis of 7- and 14-day-old rats and decreased during testicular development. In the seminiferous tubules of 7-day-old newborns, cyclin D3 was surprisingly located in cytoplasm of stem cells that had bigger nuclei than the nuclei of surrounding cells. Interestingly, cyclin D3 immunopositive cells did not immunostain with PCNA in nuclei. In the adult testis, anti-cyclin D3 antibody strongly stained the cytoplasm of early stage primary spermatocytes, lightly stained pachytene spermatocytes, but did not stain elongated spermatids. There was no detectable cyclin D3 in Sertoli cells, interstitial cells, or fibroblasts within seminiferous tubules, or in blood vessels within the interstitial matrix. The known cyclin D3 partner, cyclin dependent kinase 4, was located mainly in nuclei of spermatogonia and in early stage primary spermatocytes. Strong PCNA immunopositive staining was located in the nuclei of spermatogonia in adult testis. These results indicate that cyclin D3 is detectable in meiotically active male germ cells (PCNA-negative cells), but is conspicuously absent from mitotically active spermatogonia (PCNA-positive cells). Moreover, in contrast to in vitro reports, cyclin D3 is not located in the nucleus, but rather in the cytoplasm of male germ cells in vivo. Taken together, the presence of cyclin D3 in spermatocytes and its location
INTRODUCTION
A number of cell-cycle-regulatory proteins have been identified and categorized as either cyclins, cyclin-dependent kinases (CDKs), or cyclin-dependent kinase inhibitors [1–3]. Cyclins are a family of proteins whose production oscillates during the cell cycle. Based on sequence similarities, they are subdivided into the classes A, B, D, E, F, and G [1–5]. Cyclins are also classified as G1 and S phase cyclins, or G2 and mitotic (M) phase cyclins, depending on the role and the protein abundance during the cell cycle. Each cyclin requires association with its specific CDK to regulate the progression of the cell cycle. Several CDKs, including cdc2, CDK2, CDK4, and CDK6 have been characterized with respect to their temporal activation and their cyclin partners [5, 6]. D-type cyclins are G1 phase cyclins that associate with CDK4 and CDK6, whereas, cyclin E associates with CDK2. Both types of cyclins allow the cell to proceed through the G1/S phase transition. Cyclin A associates with CDK2 or cdc2 and has regulatory effects in S and G2 phases. Cyclin B binds to cdc2 and controls entry into the mitotic phase. Cyclin/CDK function is regulated by cyclin availability and phosphorylation state of specific serine or threonine residues of CDKs. We have developed a polyclonal antibody using the murine form of cyclin D3 recombinant protein and have found that among brain, thymus, heart, liver, spleen, kidney, testis, ovary, and smooth muscle protein levels of cyclin D3 are the highest in thymus and are the second highest in testis of rats. The testis is composed of seminiferous epithelium where germ cells advance from the periphery to the lumen while transforming from spermatogonia to spermatozoa within 45 – 53 days in the rat [7]. This process, known as spermatogenesis, involves spermatogonia under-
1
To whom correspondence and reprint requests should be addressed at Department of Physiology and Institute of Cardiovascular Research, Chonbuk National University School of Medicine, San 220, Keum-Am-Dong, Chonju, 560-180, Republic of Korea. Fax: 82652-74-9892. E-mail: [email protected]. 27
AID
ECR 3590
/
6i23$$$181
06-09-97 12:30:36
0014-4827/97 $25.00 Copyright q 1997 by Academic Press All rights of reproduction in any form reserved.
eca
28
KANG ET AL.
FIG. 1. Tissue distribution of cyclin D3 mRNA (top) and protein (bottom) in adult rats. Total RNA (20 mg) and total lysate protein (50 mg) from various tissues: B (brain), Ty (Thymus), H (heart), L (liver), S (spleen), K (kidney), Te (testis), O (ovary), and Sm (skeletal muscle). Radiolabeled cDNA probe of cyclin D3 was used for hybridization (see Materials and Methods). The blot was rehybridized with a GAPDH probe to verify equal loading of RNA in each lane (middle). The dilution factor of cyclin D3 antibody and horseradish peroxidaseconjugated anti-rabbit IgG secondary antibody for Western blotting were described under Material and Methods. Results were similar from two experiments.
going mitosis, spermatocytes undergoing meiosis, and differentiating spermatids. Therefore, the cell cycle can be studied in this dynamic system. Wolgemuth’s group has recently reported that (1) cdc2 and cyclin B1-associated kinase activities are located in meiotically dividing pachytene spermatocytes, but are absent from postmeiotic spermatids [8], (2) cyclin A expression is predominant in spermatogonia, and highest levels are found in cells where premeiotic DNA synthesis occurs, preleptene spermatocytes [10], and (3) expression of cyclin D3 mRNA is present in germ cells and is highest in nondividing, haploid, round spermatids [9]. The presence and function of cyclins are currently being investigated, yet protein levels and immunolocalization of the classical G1 phase cyclin, cyclin D3, are unknown. Using a newly developed polyclonal antibody against murine cyclin D3, we were able to determine protein levels and immunolocalization of cyclin D3 within the rat seminiferous tubule during testicular development.
FIG. 2. Northern blot analysis of cyclin D3 (top) and cyclin D1 (middle) mRNA expressions during normal testicular development. Parallel blots with total RNA (20 mg) from various developmental stages of testicular tissue preparations were examined. Testes were from 7- and 14-day-old and 1-, 2-, 3-, 4-, and 12-month-old rats. Radiolabeled cDNA probes of each cyclin were used for hybridization. Each blot was rehybridized with a GAPDH probe to verify equal loading of RNA in each lane (bottom). Results were similar from two experiments.
tated by spectrophotometry at 260 nm. For Northern analysis, RNA (20 mg) was denatured with glyoxal, separated by size on 1.2% agarose gels [11], and transferred to GeneScreen (NEN Research Product, Boston, MA). Probes were radiolabeled by random primer method according to the manufacturer’s instructions (Prime-a-Gene; Promega, Madison, WI). Specific activities of probes were typically 2–3 1 109 dpm/mg. Hybridizations were for 20 h at 657C in 41 SSC, 21 Denhardt’s, 0.1% sodium dodecyl sulfate (SDS), and 1 mg/ml salmon sperm DNA. Blots were washed at 657C in 21 SSC, 0.1% SDS, and signals were visualized by autoradiography at 0707C for 1–10 days with an intensifying screen. In all experiments, integrity, equivalent loading, and complete transfer of the RNA samples were established by UV shadowing of the blot prior to hybridization and glyceraldehyde-3-phosphate dehydrogenase (GAPDH) hybridization. Northern blot analysis was performed at least two times on each RNA sample. Generation of cDNA probes. Reverse transcriptase-polymerase chain reaction was employed to generate the cDNA probes used in this study. Oligodeoxynucleotides used for PCR amplification were as follows: Cyclin D3 [12, 13]: sense was 5*-ATGGAGCTGCTGTGTTGCGAAGGCACCCGG (1–30); antisense was 5*-CTACAGGTGTATGGCTGTGACATCTGTAGG (850–879). Cyclin D1 [13]: sense
MATERIALS AND METHODS Experimental animals and sampling. Animal experiments were conducted according to the ‘‘Recommendation from the Declaration of Helsinki and the Guiding Principles in the Care and Use of Animals.’’ Sprague–Dawley rats were obtained from our breeding colony. Male rats, 7 and 14 day and 1, 2, 3, 4, and 12 months old, were killed by decapitation and testes and thymus were harvested and were frozen in liquid nitrogen or fixed in Bouin’s solution overnight at room temperature. RNA isolation and Northern blot analysis. Total RNA from testes and thymus was extracted using TRI reagent (MRC, Cincinnati, OH) according to the manufacturer’s protocol. RNA samples were quanti-
AID
ECR 3590
/
6i23$$$182
06-09-97 12:30:36
FIG. 3. Western blot analysis of cyclin D3 (top), cyclin D1 (middle), and PCNA (bottom) during testicular development. Parallel blots with total protein (50 mg) from various developmental stages of testicular tissue preparations were examined. Testes were from 7- and 14-day-old and 1-, 2-, 3-, 4-, and 12-month-old rats. The antibodies used for each blot were described under Materials and Methods. Results were similar from two experiments.
eca
CYCLIN D3 DURING SPERMATOGENESIS
FIG. 4. Northern blot analysis of cyclin D3 (top) and cyclin D1 (middle) mRNA expressions in thymus during the aging process. Parallel blots with total RNA (20 mg) from various developmental stages of thymic tissue preparations were examined. Thymus were from 7- and 14-day-old and 1-, 2-, 3-, 4-, and 12-month-old rats. Radiolabeled cDNA probes of each cyclin were used for hybridization. Each blot was rehybridized with a GAPDH probe to verify equal loading of RNA in each lane (bottom). Results were similar from two experiments.
was 5*-ATGGAACACCAGCTCCTGTGCTGCGAAGTG (188–217); antisense was 5*-TCAGATGTCCACATCTCGCACGTCGGTGG (1046–1075). GAPDH [14]: sense was 5*-GGTGTGAACCACGAGAAATATGAC (477–500); antisense was 5*-ACCAGTGGATGCAGGGATGATGTT (690–713). Aliquots of the reverse transcriptase-polymerase chain reaction-amplified cDNAs were subcloned with the TA vector cloning system (Invitrogen, San Diego, CA) according to the manufacturer’s protocol. Identity of each cDNA probe was confirmed by sequence analysis. Generation of murine cyclin D3 antibody. Polyclonal anti-murine cyclin D3 antibody was produced by immunization of rabbit with murine cyclin D3 protein overexpressed in Escherichia coli. Briefly, the cyclin D3 cDNA fragment encoding C-terminal 236 amino acid residues was obtained RT-PCR from 1 mg total RNA of murine T-
FIG. 5. Western blot analysis of cyclin D3 (top), cyclin D1 (middle), and PCNA (bottom) in thymus during the aging process. Parallel blots with total protein (50 mg) from various developmental stages of thymic tissue preparations were examined. Thymus were from 7and 14-day-old and 1-, 2-, 3-, 4-, and 12-month-old rats. The antibodies used for each blot were described under Materials and Methods. Results were similar from two experiments.
AID
ECR 3590
/
6i23$$$182
06-09-97 12:30:36
29
FIG. 6. Schematic representation of seminferous tubules (stages VI–VII) in testis. To demonstrate the transitional characteristics of male germ cell, the cell nuclei are depicted. Spermatogonia (A), preleptene primary spermatocytes (PI), pachytene primary spermatocytes (P), differentiating round spermatids (R), and differentiating elongating spermatids (ES).
lymphocytes, and confirmed sequence by sequencing analysis. The RT-PCR product was inserted as a right open reading frame into NcoI/BamHI site of protein expression vector, pET 3d (Novagen, Madison, WI). The expression of cyclin D3 protein in E. coli host in the presence of 0.3 mM isoprophyl b-D-thiogalactopyranoside was performed essentially as described by Studier et al. [15]. The overexpressed cyclin D3 protein localized in inclusion body was denatured by guanidine hydrochloride and then renatured into a soluble form. Four hundred micrograms of cyclin D3 was mixed with an equal volume of complete Freund’s adjuvant and injected intramuscularly into the thigh and subcutaneously into the back neck area of a rabbit. For booster immunizations, the same amount of cyclin D3 protein mixed with incomplete Freund’s adjuvant was injected into the same rabbit with the same manner every 4 weeks. Bleeding was done 2 weeks after each immunization. Western blot. Testes samples were frozen with liquid nitrogen and stored at 0707C. Samples were homogenized directly in Nonidet P-40 buffer [150 mM NaCl, 5 mM EDTA, 50 mM Tris–HCl (pH 8.0), 1 mg/ml aprotinin, 1 mg/ml pepstatin, 1 mg/ml leupeptin, 50 mg/ml 1chloro-3-toxylamido-7-amino-2-heptanone, 50 mg/ml phenylmethylsulfonyl fluoride, 100 mg/ml L-1-tosylamino-2-phenylethyl chloromethyl ketone, 20 mM sodium fluoride, 20 mM glycerophosphate, 1% (v/v) Nonidet P-40]. Lysate protein was quantitated using a Bradford assay (Bio-Rad, Hercules, CA) with bovine serum albumin as a reference standard. Samples were boiled with 21 sample buffer for 10 min, separated by size on 12.5% polyacrylamide gel under SDS denaturing conditions, and electrotransferred to nitrocellulose membranes. The nitrocellulose membranes were stained with 0.1% naphthol blue black (Amido black) in 10% (v/v) acetic acid and 40% (v/v) methanol to assess the efficiency of transfer. Nonspecific binding was blocked by incubation in blocking buffer (5% nonfat dry milk, 3% bovine serum albumin, 0.1% Tween 20, 11 PBS) for 1 h at room temperature. Anti-cyclin D3 serum (1:200 dilution), anti-cyclin D1
eca
30
KANG ET AL.
FIG. 7. Light micrographs of portions of seminiferous tubules immunostained with anti-cyclin D3 antibody and PCNA antibody in a 7day-old newborn. In tissue sections, cyclin D3 appeared as intense reddish–brown punctated stains and was readily observed in the cytoplasm. PCNA appeared as intense reddish–brown punctated stains in cytoplasm and nuclei among a background of lighter hematoxylinstained nuclei. (A) Some epithelial stem cells showed cyclin D3 immunopositive staining. Magnification, 1200. (B) Cells that had bigger nuclei than surrounding cells had immunopositive cyclin D3 in cytoplasm. Magnification, 11000. (C) Nuclear dividing cells were visible, and some of these cells had abundant cyclin D3 in the cytoplasm. Magnification, 11000. (D, E) 80–90% of cells had immunopositive PCNA in their nuclei. No immunopositive staining of PCNA was observed in cells which had a bigger nucleus than surrounding cells and nuclear dividing cells. Magnification, 11000.
FIG. 8. Light micrographs of portions of seminiferous tubules (A, B, D, E, F) and thymus (C) that were immunostained with anti-cyclin D3 antibody (A, B, C, D, E) and anti-CDK4 antibody (F) in 30-day-old (A) and 3-month-old rats (B, C, D, E, F). (A) Strong cyclin D3 immunopositive staining was located mainly in the cytoplasm of luminal primary spermatocytes. There was no or barely detectable immunopositive cyclin D3 in some seminiferous tubules. Magnification, 1200. (B) A variable immunostaining pattern for cyclin D3 is visualized in different seminiferous tubules. Magnification, 1200. (C) Corticomedullary portion of thymus. Strong cyclin D3 immunopositive staining was located in cytoplasm and nuclei of 10–15% of thymocytes population. Magnification, 11000. (D, stage IV; E, stage IX) Cyclin D3 was abundant in cytoplasm of pachytene primary spermatocytes, but undetectable in elongated spermatids. Magnification, 11000. There was no cyclin D3 in spermatogonia cells or Sertoli cells. (F, stage V) CDK4 was abundant in nuclei of spermatogonia and spermatocytes, but not or barely detectable in cytoplasm of round spermatids. The stages of ST were determined according to Russell et al. (1990).
AID
ECR 3590
/
6i23$$3590
06-09-97 12:30:36
eca
CYCLIN D3 DURING SPERMATOGENESIS
AID
ECR 3590
/
6i23$$3590
06-09-97 12:30:36
eca
31
32
KANG ET AL.
monoclonal (72-13G; Santa Cruz Biotech, Santa Cruz, CA), and antiproliferating cell nuclear antigen (PCNA) monoclonal antibody (PC10; Santa Cruz Biotech) were used. Secondary anti-mouse IgG antibody which was conjugated to horseradish peroxidase (1:2,000) was used. Signals were visualized by the ECL detection method according to the manufacturer’s protocol (Amersham Int. plc, Buckinghamshire, United Kingdom). Immunohistochemistry. For paraffin sections, three testes in each age group were fixed in Bouin’s fixatives, dehydrated through graded alcohols, and infiltrated with paraffin under standard procedures. Tissue blocks were then sectioned at 4 mm. Sections were mounted on positively charged microscope slides (ProbeOn Plus, Fisher Scientific). Deparaffinized sections were hydrated, permeabilized in 0.1% Triton X-100, and blocked with phosphate-buffered saline containing 1% bovine serum albumin and 5% goat serum. Sections were then incubated with anti-cyclin D3 antiserum (1:100 dilution), anti-CDK4 polyclonal antibody (1:50 dilution, C-22, Santa Cruz Biotech) or anti-PCNA (1:100) at 47C overnight, and signals were developed with an avidin-biotin-peroxidase system (Biomeda) with 3-amino-9-ethyl-carbazole (for cyclin D3 and CDK4) or diaminobenzidine (for PCNA) as a chromogen. Sections were counterstained with hematoxylin. Either nonimmune rabbit serum (for polyclonal antibody) or irrelevant monoclonal antibodies were used for control slides. Slides were viewed and photographed with a microscope equipped with differential interface contrast optics (Axiskop with MC-80 camera, Zeiss).
RESULTS
Tissue Distribution and Temporal Levels of Cyclin D3 mRNA and Protein among Various Organs We examined the levels of cyclin D3 mRNA and protein in various organs of adult rats (Fig. 1). The level of cyclin D3 mRNA was the highest in thymus and was readily detectable in heart, spleen, and testis, but was barely detectable or undetectable in other organs including brain, liver, kidney, ovary, and skeletal muscle (Fig. 1, top). The level of cyclin D3 protein was highest in thymus, second highest in testis, third highest in spleen, but was barely detectable or undetectable in other organs including brain, heart, liver, kidney, ovary, and skeletal muscle (Fig. 1, bottom). Thus, rate of translation and degradation of cyclin D3 may differ among various organs. Temporal Levels of Cyclin D3 and Cyclin D1 mRNA and Protein in Testis and Thymus during Normal Aging Process The expression of cyclin D3 mRNA in 7-day testis was high, but the expression of cyclin D3 mRNA decreased, and then remained almost constant in 14-dayto 1-year-old rats (Fig. 2). A previous study reported an induction of cyclin D3 mRNA during testicular development [9]. In contrast to the fairly constant level of cyclin D3 mRNA, the level of cyclin D1 mRNA was high in the testis of 7- and 14-day-old rats and decreased gradually during testicular development (Fig. 2). This result was different from a previous study that reported a marked reduction of cyclin D1 mRNA during
AID
ECR 3590
/
6i23$$$182
06-09-97 12:30:36
testicular development in the mouse [9]. The protein levels of cyclin D3 in testes were high in 7-day-old newborns, remained almost constant until 3 months, and then gradually decreased thereafter (Fig. 3). In testes of 7- and 14-day-old newborns, the protein levels of cyclin D1 were high, but decreased and remained at almost constant levels thereafter (Fig. 3). As a proliferative indicator, the protein levels of PCNA [16] were measured in developing testes (Fig. 3). PCNA protein decreased gradually during testicular development. Overall, levels of cyclin D3, cyclin D1, and PCNA protein decreased at different rates during testicular development. The expression of cyclin D3 mRNA in 7-day thymus was high, remained almost constant until 2 months, and then decreased gradually thereafter (Fig. 4). In contrast to the decrease of cyclin D3 mRNA in the thymus, the level of cyclin D1 mRNA was almost constant from 7 day old to 12 month old rats (Fig. 4). The levels of cyclin D3, cyclin D1 and PCNA proteins were almost constant in thymus of 7-day-old to 12month-old rats (Fig. 5). Thus, the protein levels of cyclin D3 and D1 did not reflect changes of the corresponding mRNA levels in the testis and the thymus of in vivo animals. Discrepancies between levels of cyclin mRNA and protein have been observed by us [17] and others [18] and suggest that at different ages, some organs control the level of cyclins by rate of translation or degradation of proteins rather than by rate of transcription. Immunolocalization of Cyclin D3, CDK4, and PCNA during Testicular Development A cross-section of rat testis contains several hundreds seminferous tubules, which have been classified into 14 different stages. Each stage of seminferous tubule represents a different combination of dividing and differentiating germ cells (Fig. 6). Immunostaining of cyclin D3 protein in tissue sections appeared as intense reddish–brown punctated stains and was readily observed in the cytoplasm among a background of lighter hematoxylin-stained nuclei (Figs. 7 and 8). Immunostaining of PCNA protein in tissue sections also appeared as intense reddish–brown stains and was observed in the hematoxylin-stained nuclei (Figs. 7 and 9). A variable immunostaining pattern was visualized among different developmental periods and, in adult rats, different stages of seminiferous tubules (Figs. 7, 8, and 9). In the seminiferous tubules of 7-day-old newborns, cyclin D3 was mainly located in the cytoplasm of cells that had bigger nuclei than nuclei of surrounding cells (Figs. 7A and 7B). Also, the cells that seemed to be undergoing nuclear division had abundant cyclin D3 (Fig. 7C). Eighty to 90% of stem cells were PCNA immunopositive in 7-day-old newborns (Fig. 7D). The cells with bigger nuclei than surrounding
eca
CYCLIN D3 DURING SPERMATOGENESIS
cells, and nuclear dividing cells that stained positively for cyclin D3, did not stain for PCNA (Fig. 7E). In the 1month-old and older testes, a variable immunostaining pattern for cyclin D3 was visualized in different seminiferous tubules (Figs. 8A and 8B). In 1-month-old rats, some seminiferous tubules showed strong immunostaining of cyclin D3, while others totally lacked cyclin D3. The population of PCNA immunopositive staining cells in seminiferous tubules decreased during aging. Only spermatogonia contained immunopositive PCNA staining (Fig. 9A). In adult rats, cyclin D3 darkly stained primary spermatocytes, lightly stained late stage primary spermatocytes, but did not stain elongated spermatids (Figs. 8D and 8E). There was no cyclin D3 detected in Sertoli cells, interstitial cells, or fibroblasts and blood vessels within the interstitial matrix. The known cyclin D3 partner, CDK4, was localized mainly either in the nuclei of spermatogonia or in early stage primary spermatocytes (Fig. 8F). Detectable amounts (immunostaining positive signal was higher than background) of CDK4 were also localized in the cytoplasm of round spermatids. As a positive control, we examined the immunostaining pattern of cyclin D3 and CDK4 in the thymus. In the thymus, strong cyclin D3 immunopositive staining was located in the cytoplasm and nuclei of thymocytes (10–15% of population) (Fig. 8C). However, strong CDK4-immunopositive staining was located mainly in the nuclei of thymocytes (3–5% of the population; data not shown). The peripheral cells of seminiferous tubules showed strong PCNA immunopositive staining regardless of seminiferous tubule stage in adult rats (Fig. 9B). High magnification (11000) revealed that only spermatogonia and early stage primary spermatocytes showed PCNA immunopositive staining in nuclei, regardless of seminiferous tubule stage (Figs. 9C, 9D, and 9E). Thus, the percentage of PCNA immunopositive staining cells within seminiferous tubules decreased from 7-day-old to adult rats (7 days old, 86.2%; 30 days old, 22.3%; 3 months old, 12.6%). Controls slides, which were treated with either nonimmune rabbit serum (for polyclonal antibody) or irrelevant monoclonal antibodies, and immunofluorescence-treated slides were used to exclude the possibility of false positives by endogenous peroxidase, and were not immunoreactive (data not shown). DISCUSSION
In the present study, a tissue survey of rat tissues revealed that cyclin D3 mRNA was highly detectable in thymus, spleen, testis, and heart, but cyclin D3 protein was highly detectable only in thymus and testis. The levels of cyclin D3 protein in testes were high in 7-day-old newborns, remained almost constant until 3 months, and then gradually decreased thereafter. Surprisingly, cyclin D3 was located in the cytoplasm of
AID
ECR 3590
/
6i23$$$183
06-09-97 12:30:36
33
male germ cells. Cyclin D3 was abundant in primary spermatocytes, but was not present in either spermatogonia or elongated spermatids. CDK4 was localized mainly in nuclei of spermatogonia and early stage primary spermatocytes. PCNA was abundant only in the nuclei of spermatogonia and early stage primary spermatocytes. These results indicated that cyclin D3 is detectable in meiotically active male germ cells (PCNA-negative cells) but is absent from mitotically active spermatogonia (PCNA-positive cells). Ravnik et al. [9] reported that in mouse testis, the expression of cyclin D3 mRNA is highest in elongating spermatids, two of the three different sizes of cyclin D3 mRNA transcripts are germ cell specific, and the 2.3-kb cyclin D3 transcript increases while the expression of cyclin D1 mRNA decreases during aging. Moreover, mRNAs of cyclin D1 and D3 are expressed in nondividing rather than mitotically active germ cells. However, protein levels and immunolocalization of cyclin D3 have not been reported. Data from the present study differed from the report by Ravnik et al. (1995) in that (1) levels of cyclin D3 mRNA and protein decreased during testicular development, and (2) rat testis expressed only one size of cyclin D3 mRNA transcript. Species differences may account for these discrepancies. Similarities between the two reports are as follows: (1) expression of cyclin D1 mRNA decreased during testicular development and (2) cyclin D3 was expressed in nondividing, rather than mitotically active, male germ cells. Our most remarkable finding was the restricted presence of cyclin D3 in the cytoplasm of germ cells. Cyclin D3 is located in the nucleus of HT2T lymphocytes [19] and since cyclin D3 is necessary for the progression of the cell cycle in vitro, its location would be expected to be in the nucleus. In the thymus, we detected positive cyclin D3 staining in the nuclei and cytoplasm of thymocytes. However, using the same antibody, cyclin D3 was detected only in the cytoplasm of meiotically active male germ cells. The known cyclin D3 partner, CDK4, was localized mainly either in the nuclei of some spermatogonia cells or in some early stage primary spermatocytes. CDK4 was also detected in the cytoplasm of round spermatids. In proliferating melanoma cells, CDK4 is also present in nuclei and/or cytoplasm (20). Because of its location, a major portion of cyclin D3 may not interact with CDK4 in the nuclei of spermatocytes in vivo. Thus, the question arises, why is cyclin D3 present in the cytoplasm and not in the nucleus? An underlying mechanism for regulation of many processes in the cell is the localization of protein complexes within specific subcellular compartments. Some cyclin–CDK complexes are subject to this mode of regulation [21–23]. For example, the M-phase cyclin, cyclin B1, is translocated from the cytoplasm to the nucleus immediately prior to mitosis; the timing of this translocation may be a critical factor in determining the initia-
eca
34
AID
KANG ET AL.
ECR 3590
/
6i23$$3590
06-09-97 12:30:36
eca
CYCLIN D3 DURING SPERMATOGENESIS
tion of mitosis. In vivo, cyclin A is translocated from the cytoplasm to the nucleus of developing spermatocytes in testes of Drosophilia [24], but this translocation of cyclin A in spermatocytes does not occur in mice indicating species specificity [10]. Interestingly, cyclin D1 was shown to interact and to activate estrogen receptors independent of CDKs in breast epithelial cells [25] and cyclin D2 was shown to interact with the transcriptional factor, DMP-1 [26]. Thus, cyclin D3 may associate with yet unidentified protein kinases or interact with other proteins located in the cytoplasm and have functions other than G1-phase progression. D-type cyclins (D1, D2, D3) are expressed earlier in the G1 phase of the cell cycle than cyclin E, and they are known to link growth factor stimulation to the activation of S-phase genes [4, 13]. Furthermore, D-type cyclins are necessary and rate limiting for G1 progression during the mitotic cell cycle [4, 5], although the expression of cyclin D3 is quite different from expressions of cyclins D1 and D2. In 3T3 fibroblasts, a modest induction of cyclin D3 mRNA occurs near the G1/S boundary, 12 h after serum stimulation, while cyclin D1 mRNA increases 20- to 30-fold, peaking 6 h after addition of serum [27]. In human T lymphocytes, cyclin D2 mRNA accumulates in the early G1 phase, whereas cyclin D3 mRNA accumulates during the late G1 stage [28]. In cultured rat hepatocytes, cyclin D3 mRNA accumulates in the absence of mitogenic signals, yet mitogenic signals are necessary for the accumulation of cyclins D1 and D2 mRNAs [29]. Furthermore, when skeletal myoblasts are induced to differentiate with growth factors and serum deprivation, cyclin D3 mRNA increases, while mRNA levels of cyclin D1 and D2 rapidly decrease [30, and our unpublished observations]. Thus, although D-type cyclins regulate G1 phase progression, the functions of D-type cyclins may not be redundant. Moreover, there are accumulating reports that in vivo and in vitro expressions of D-type cyclins are quite different [9, 17, 31]. Unlike other members of the cyclin family, D-type cyclins are expressed in some nondividing cell types in vivo [9, 17, 18, 31]. For example, the heart and the kidney express D-type cyclins mRNAs after the cell cycle is arrested in young animals, while at the same time, cyclins A and B rapidly fall to undetectable levels [17, 31]. Genetically, overexpression of cyclin D1 is sufficient to alter the cell cycle in mammary tissue [32] but does not alter the cell cycle in lymphoid tissue [33, 34]. Thus, cyclin D1
35
is not the only limiting regulatory factor for proliferation and its effects may differ among tissues. Cyclin D3 was characterized by three groups [12, 35, 36] and the biological function of cyclin D3 was established by in vitro biochemical and cell culture studies. In the insect cell Sf9, cyclin D3 interacts with the serine threonine protein kinase, cdk2, and phosphorylates retinoblastoma protein (Rb) to relieve Rb’s inhibition of proliferation [37]. However, the in vivo function of cyclin D3 remains to be clarified. The testis provides a unique system in which to study cyclin D3 and the cell cycle because within a seminiferous tubule (1) mitotically active spermatogonia, meiotically active spermatocytes, and post-meiotic spermatids exist and (2) high levels of cyclin D3 protein are present. The seminiferous tubules of 7-day-old rats contain mitotically active epithelial stem cells. Cyclin D3 was located in cells with nuclei larger than the nuclei of surrounding cells. A positive signal was observed from the mitotic indicator, PCNA, in almost all stem cells except those with large nuclei. This indicates that cyclin D3 was not present in mitotically active cells. The cells with large nuclei may have been quiescient or undergoing the first stages of meiosis, but were not apoptotic as determined by the TUNEL method (data not shown). By 30 days postnatal, meiosis has already begun in spermatocytes, and the size of the epithelium is nearly that of the adult [38]. In 30-day-old rats, positive staining for cyclin D3 was detected in primary spermatocytes, but not in mitotically active spermatogonia. In contrast, positive staining for PCNA was detected in spermatogonia, but not in spermatocytes. This contrasting pattern of staining for PCNA and cyclin D3 was also observed in adult testis regardless of the stage of seminiferous tubules. Cyclin D3 associates with PCNA and with various CDKs in diploid lung fibroblasts [37, 39]. However, our in vivo data indicates that cyclin D3 and CDK4 were not present together in the same cells as PCNA. In the adult testis, cyclin D3 was most abundant in pachytene spermatocytes and gradually decreased in round and early stage elongating spermatids as they matured and moved toward the lumen. Since cyclin D3 was not detectable in elongated spermatids, cyclin D3 may not be needed for the final maturation of spermatids. In summary, we have developed an anti-murine cyclin D3 antibody that detects high levels of cyclin D3 in testis and thymus. Cyclin D3 is detectable in
FIG. 9. Light micrographs of portions of seminiferous tubules that were immunostained with PCNA antibody in 30-day-old (A) and 3month-old rats (B, C, D, E). (A) The population of PCNA-immunopositive-staining cells decreased. Spermatogonia cells showed PCNA immunopositive staining. Magnification, 11000. (B) Peripheral germ cells showed strong PCNA-immunopositive staining regardless of stage of seminiferous tubules. Magnification, 1200. (C, stage XI-XII; D (top), stage XII; D (bottom), stages II–III; E, stage VIII). High magnification revealed that only spermatogonia and early phase primary spermatocytes showed immunopositive PCNA. Magnification, 11000. The stages of seminiferous tubules were determined according to Russell et al. (1990).
AID
ECR 3590
/
6i23$$$183
06-09-97 12:30:36
eca
36
KANG ET AL.
meiotically active male germ cells and increases during development, but is conspicously absent from mitotically active spermatogonia. In contrast to in vitro reports, cyclin D3 is not located in the nucleus and is not present with CDK4 and PCNA in same cell, but rather is located in the cytoplasm of male germ cells in vivo. The presence and cytoplasmic location of cyclin D3 in spermatocytes lead us to speculate that cyclin D3 may have functions in male germ cells other than mitosis. We thank Ms. Keum Nim Koh, Ms. Hae Young Park, and Mr. Won Jin Cho for excellent technical assistance, This work was supported by grants from Korea Science and Engineering Foundation (95-040301-01-3) and from the Ministry of Education for promoting Genetic Engineering in 1994.
REFERENCES 1. Hunter, T., and Pines, J. (1991) Cell 66, 1071–1074. 2. Murray, A., and Hunt, T. (1993) in The Cell Cycle, Freeman, New York. 3. Hunter, T., and Pines, J. (1994) Cell 79, 573–582. 4. Sherr, C. J. (1993) Cell 73, 1059–1065. 5. Sherr, C. J. (1994) Cell 79, 551–555. 6. Morgan, D. O. (1995) Nature 374, 131–134. 7. Russell, L. D., Ettlin, R. A., Hikim, A. P. S., and Clegg, E. D. (1990) in Histological and Histopathological Evaluation of the Testis, Cache River Press, Clearwater, FL. 8. Chapman, D. L., and Wolgemuth, B. J. (1994) Dev. Biol. 165, 500–506. 9. Ravnik, S. E., Lee, K., and Wolgemuth, D. J. (1995) Dev. Genet. 16, 171–178. 10. Ravnik, S. E., and Wolgemuth, D. J. (1996) Dev. Biol. 173, 69– 78. 11. Sambrook, J., Fritsch, E. F., and Maniatis, T. (1989) in Molecular Cloning: A Laboratory Manual, 2nd ed., Cold Spring Harbor Laboratory, Cold Spring Harbor, NY. 12. Motokura, T., Keyomars, K., Kronenberg, M. H., and Arnold, A. (1992) J. Biol. Chem. 267, 20412–20415. 13. Matsushime, H., Roussel, M. F., Ashmun, R. A., and Sherr, C. J. (1991) Cell 65, 701–713. 14. Fort, P., Marty, L., Piechaczyk, M., Sabrouty, S. E., Dani, C., Jeanteur, P., and Blanchard, J. M. (1985) Nucleic Acids Res. 13, 1431–1442. 15. Studier, F. W., Rosenberg, A. H., Dunn, J. J., and Dubendorff, J. W. (1990) Methods Enzymol. 185, 60–89.
16. Bravo, R., Frank, R., Blundell, P. A., and MacDonald-Bravo, H. (1987) Nature 326, 515–517. 17. Park, S. K., Kang, S. K., Lee, D. Y., Kang, M. J., Kim, S. H., and Koh, G. Y. (1997) Kidney Int., in press. 18. Tamaru, T., Okada, M., and Nakagawa, H. (1994) Neurosci. Lett. 168, 229–232. 19. Vallance, S. J., Lee, H. M., Roussel, M. F., Shurtleff, S. A., Kato, J. Y., Strom, D. K., and Sherr, C. J. (1994) Hybridoma 13, 37– 44. 20. Wang, Y. L., Uhara, H., Yamazaki, Y., Nikaido, T., and Saida, T. (1996) Br. J. Dermatol. 134, 269–275. 21. Pines, J., and Hunter, T. (1991) J. Cell. Biol. 115, 1–17. 22. Bailly, E., Pines, J., Hunter, T., and Bornens, M. (1992) J. Cell. Sci. 101, 529–545. 23. Baldin, V., and Ducommun, B. (1995) J. Cell Sci. 108, 2425– 2432. 24. Eberhart, C. G., Maines, J. Z., and Wasserman, S. A. (1996) Nature 381, 783–785. 25. Zwijsen, R. M., Wientjens, E., Klompmaker, R., Sman, J. V. D., Bernards, R., and Michalides, R. J. A. M. (1997) Cell 88, 405– 415. 26. Hirai, H., and Sherr, C. J. (1996) Mol. Cell Biol. 16, 6457–6467. 27. Winston, J. T., and Pledger, W. J. (1993) Mol. Biol. Cell 4, 1133– 1144. 28. Ajchenbaumet, F., Ando, K., DeCaprio, J. A., and Griffin, J. D. (1993) J. Biol. Chem. 268, 4113–4119. 29. Loyer, P., Cariou, S., Glaise, D., Bilodeau, M., Baffet, G., and Guguen-Guillouzo, C. (1996) J. Biol. Chem. 271, 11484–11492. 30. Rao, S. S., and Kohtz, D. S. (1995) J. Biol. Chem. 270, 4093– 4100. 31. Yoshizumi, M., Lee, W. S., Hsieh, C. M., Tsai, J. C., Li, J., Perrella, M. A., Patterson, C., Endegr, W. O., Schlegel, R., and Lee, M. E. (1995) J. Clin. Invest. 95, 2275–2280. 32. Wang, T. C., Cardiff, R. D., Zukerberg, L., Lees, E., Arnold, A., and Schmidt, E. V. (1994) Nature 369, 669–671. 33. Lovec, H., Grzeschiczek, A., and Moroy, T. (1994) EMBO J. 13, 3487–3495. 34. Bodrug, S., Warner, B., Bath, M., Lindeman, G., Harris, A., and Adams, J. (1994) EMBO J. 13, 2124–2130. 35. Inaba, T., Matsushime, H., Valentine, M., Roussel, M. F., Sherr, C. J., and Look, A. T. (1992) Genomics 13, 565–674. 36. Xiong, Y., Menninger, J., Beach, D., and Ward, D. C. (1992) Genomics 13, 575–584. 37. Ewen, M. E., Sluss, H. K., Sherr, C. J., Matsushime, H., Kato, J., and Livingston, D. M. (1993) Cell 73, 487–497. 38. Hervieu, G., Segretain, D., and Nachon, J. L. (1996) Biol. Reprod. 54, 1161–1172. 39. Xiong, Y., Zhang, H., and Beach, D. (1992) Cell 71, 505–514.
Received November 6, 1996 Revised version received March 14, 1997
AID
ECR 3590
/
6i23$$$184
06-09-97 12:30:36
eca
234, 27–36 (1997)
EX973590
Cytoplasmic Localization of Cyclin D3 in Seminiferous Tubules during Testicular Development Min Jeong Kang,* Mi Kyung Kim,† Amy Terhune,* Jong Kwan Park,‡ Young Ho Kim,† and Gou Young Koh*,1 *Department of Physiology and Institute of Cardiovascular Research and ‡Department of Urology, Chonbuk National University Medical School, Chonju, 560-180, Republic of Korea; †Department of Microbiology, College of Natural Sciences, Kyungpook National University, Taegu 702-701, Republic of Korea
in the cytoplasm lead us to speculate that cyclin D3 may have functions in male germ cells other than mitosis. q 1997 Academic Press
Using a newly developed polyclonal antibody against murine cyclin D3, we have found that protein levels of cyclin D3 were highly detectable only in thymus and testis in rats. Since testis offer unique opportunities to examine the cell cycle in vivo, we examined the temporal and spatial expression of cyclin D3 and the DNA synthesis indicator, proliferating cell nuclear antigen (PCNA), in the rat testis during development. The protein levels of cyclin D3 protein in testis from 7 days to 3 months old were almost constant and then decreased gradually thereafter. The protein levels of cyclin D1 and PCNA were high in the testis of 7- and 14-day-old rats and decreased during testicular development. In the seminiferous tubules of 7-day-old newborns, cyclin D3 was surprisingly located in cytoplasm of stem cells that had bigger nuclei than the nuclei of surrounding cells. Interestingly, cyclin D3 immunopositive cells did not immunostain with PCNA in nuclei. In the adult testis, anti-cyclin D3 antibody strongly stained the cytoplasm of early stage primary spermatocytes, lightly stained pachytene spermatocytes, but did not stain elongated spermatids. There was no detectable cyclin D3 in Sertoli cells, interstitial cells, or fibroblasts within seminiferous tubules, or in blood vessels within the interstitial matrix. The known cyclin D3 partner, cyclin dependent kinase 4, was located mainly in nuclei of spermatogonia and in early stage primary spermatocytes. Strong PCNA immunopositive staining was located in the nuclei of spermatogonia in adult testis. These results indicate that cyclin D3 is detectable in meiotically active male germ cells (PCNA-negative cells), but is conspicuously absent from mitotically active spermatogonia (PCNA-positive cells). Moreover, in contrast to in vitro reports, cyclin D3 is not located in the nucleus, but rather in the cytoplasm of male germ cells in vivo. Taken together, the presence of cyclin D3 in spermatocytes and its location
INTRODUCTION
A number of cell-cycle-regulatory proteins have been identified and categorized as either cyclins, cyclin-dependent kinases (CDKs), or cyclin-dependent kinase inhibitors [1–3]. Cyclins are a family of proteins whose production oscillates during the cell cycle. Based on sequence similarities, they are subdivided into the classes A, B, D, E, F, and G [1–5]. Cyclins are also classified as G1 and S phase cyclins, or G2 and mitotic (M) phase cyclins, depending on the role and the protein abundance during the cell cycle. Each cyclin requires association with its specific CDK to regulate the progression of the cell cycle. Several CDKs, including cdc2, CDK2, CDK4, and CDK6 have been characterized with respect to their temporal activation and their cyclin partners [5, 6]. D-type cyclins are G1 phase cyclins that associate with CDK4 and CDK6, whereas, cyclin E associates with CDK2. Both types of cyclins allow the cell to proceed through the G1/S phase transition. Cyclin A associates with CDK2 or cdc2 and has regulatory effects in S and G2 phases. Cyclin B binds to cdc2 and controls entry into the mitotic phase. Cyclin/CDK function is regulated by cyclin availability and phosphorylation state of specific serine or threonine residues of CDKs. We have developed a polyclonal antibody using the murine form of cyclin D3 recombinant protein and have found that among brain, thymus, heart, liver, spleen, kidney, testis, ovary, and smooth muscle protein levels of cyclin D3 are the highest in thymus and are the second highest in testis of rats. The testis is composed of seminiferous epithelium where germ cells advance from the periphery to the lumen while transforming from spermatogonia to spermatozoa within 45 – 53 days in the rat [7]. This process, known as spermatogenesis, involves spermatogonia under-
1
To whom correspondence and reprint requests should be addressed at Department of Physiology and Institute of Cardiovascular Research, Chonbuk National University School of Medicine, San 220, Keum-Am-Dong, Chonju, 560-180, Republic of Korea. Fax: 82652-74-9892. E-mail: [email protected]. 27
AID
ECR 3590
/
6i23$$$181
06-09-97 12:30:36
0014-4827/97 $25.00 Copyright q 1997 by Academic Press All rights of reproduction in any form reserved.
eca
28
KANG ET AL.
FIG. 1. Tissue distribution of cyclin D3 mRNA (top) and protein (bottom) in adult rats. Total RNA (20 mg) and total lysate protein (50 mg) from various tissues: B (brain), Ty (Thymus), H (heart), L (liver), S (spleen), K (kidney), Te (testis), O (ovary), and Sm (skeletal muscle). Radiolabeled cDNA probe of cyclin D3 was used for hybridization (see Materials and Methods). The blot was rehybridized with a GAPDH probe to verify equal loading of RNA in each lane (middle). The dilution factor of cyclin D3 antibody and horseradish peroxidaseconjugated anti-rabbit IgG secondary antibody for Western blotting were described under Material and Methods. Results were similar from two experiments.
going mitosis, spermatocytes undergoing meiosis, and differentiating spermatids. Therefore, the cell cycle can be studied in this dynamic system. Wolgemuth’s group has recently reported that (1) cdc2 and cyclin B1-associated kinase activities are located in meiotically dividing pachytene spermatocytes, but are absent from postmeiotic spermatids [8], (2) cyclin A expression is predominant in spermatogonia, and highest levels are found in cells where premeiotic DNA synthesis occurs, preleptene spermatocytes [10], and (3) expression of cyclin D3 mRNA is present in germ cells and is highest in nondividing, haploid, round spermatids [9]. The presence and function of cyclins are currently being investigated, yet protein levels and immunolocalization of the classical G1 phase cyclin, cyclin D3, are unknown. Using a newly developed polyclonal antibody against murine cyclin D3, we were able to determine protein levels and immunolocalization of cyclin D3 within the rat seminiferous tubule during testicular development.
FIG. 2. Northern blot analysis of cyclin D3 (top) and cyclin D1 (middle) mRNA expressions during normal testicular development. Parallel blots with total RNA (20 mg) from various developmental stages of testicular tissue preparations were examined. Testes were from 7- and 14-day-old and 1-, 2-, 3-, 4-, and 12-month-old rats. Radiolabeled cDNA probes of each cyclin were used for hybridization. Each blot was rehybridized with a GAPDH probe to verify equal loading of RNA in each lane (bottom). Results were similar from two experiments.
tated by spectrophotometry at 260 nm. For Northern analysis, RNA (20 mg) was denatured with glyoxal, separated by size on 1.2% agarose gels [11], and transferred to GeneScreen (NEN Research Product, Boston, MA). Probes were radiolabeled by random primer method according to the manufacturer’s instructions (Prime-a-Gene; Promega, Madison, WI). Specific activities of probes were typically 2–3 1 109 dpm/mg. Hybridizations were for 20 h at 657C in 41 SSC, 21 Denhardt’s, 0.1% sodium dodecyl sulfate (SDS), and 1 mg/ml salmon sperm DNA. Blots were washed at 657C in 21 SSC, 0.1% SDS, and signals were visualized by autoradiography at 0707C for 1–10 days with an intensifying screen. In all experiments, integrity, equivalent loading, and complete transfer of the RNA samples were established by UV shadowing of the blot prior to hybridization and glyceraldehyde-3-phosphate dehydrogenase (GAPDH) hybridization. Northern blot analysis was performed at least two times on each RNA sample. Generation of cDNA probes. Reverse transcriptase-polymerase chain reaction was employed to generate the cDNA probes used in this study. Oligodeoxynucleotides used for PCR amplification were as follows: Cyclin D3 [12, 13]: sense was 5*-ATGGAGCTGCTGTGTTGCGAAGGCACCCGG (1–30); antisense was 5*-CTACAGGTGTATGGCTGTGACATCTGTAGG (850–879). Cyclin D1 [13]: sense
MATERIALS AND METHODS Experimental animals and sampling. Animal experiments were conducted according to the ‘‘Recommendation from the Declaration of Helsinki and the Guiding Principles in the Care and Use of Animals.’’ Sprague–Dawley rats were obtained from our breeding colony. Male rats, 7 and 14 day and 1, 2, 3, 4, and 12 months old, were killed by decapitation and testes and thymus were harvested and were frozen in liquid nitrogen or fixed in Bouin’s solution overnight at room temperature. RNA isolation and Northern blot analysis. Total RNA from testes and thymus was extracted using TRI reagent (MRC, Cincinnati, OH) according to the manufacturer’s protocol. RNA samples were quanti-
AID
ECR 3590
/
6i23$$$182
06-09-97 12:30:36
FIG. 3. Western blot analysis of cyclin D3 (top), cyclin D1 (middle), and PCNA (bottom) during testicular development. Parallel blots with total protein (50 mg) from various developmental stages of testicular tissue preparations were examined. Testes were from 7- and 14-day-old and 1-, 2-, 3-, 4-, and 12-month-old rats. The antibodies used for each blot were described under Materials and Methods. Results were similar from two experiments.
eca
CYCLIN D3 DURING SPERMATOGENESIS
FIG. 4. Northern blot analysis of cyclin D3 (top) and cyclin D1 (middle) mRNA expressions in thymus during the aging process. Parallel blots with total RNA (20 mg) from various developmental stages of thymic tissue preparations were examined. Thymus were from 7- and 14-day-old and 1-, 2-, 3-, 4-, and 12-month-old rats. Radiolabeled cDNA probes of each cyclin were used for hybridization. Each blot was rehybridized with a GAPDH probe to verify equal loading of RNA in each lane (bottom). Results were similar from two experiments.
was 5*-ATGGAACACCAGCTCCTGTGCTGCGAAGTG (188–217); antisense was 5*-TCAGATGTCCACATCTCGCACGTCGGTGG (1046–1075). GAPDH [14]: sense was 5*-GGTGTGAACCACGAGAAATATGAC (477–500); antisense was 5*-ACCAGTGGATGCAGGGATGATGTT (690–713). Aliquots of the reverse transcriptase-polymerase chain reaction-amplified cDNAs were subcloned with the TA vector cloning system (Invitrogen, San Diego, CA) according to the manufacturer’s protocol. Identity of each cDNA probe was confirmed by sequence analysis. Generation of murine cyclin D3 antibody. Polyclonal anti-murine cyclin D3 antibody was produced by immunization of rabbit with murine cyclin D3 protein overexpressed in Escherichia coli. Briefly, the cyclin D3 cDNA fragment encoding C-terminal 236 amino acid residues was obtained RT-PCR from 1 mg total RNA of murine T-
FIG. 5. Western blot analysis of cyclin D3 (top), cyclin D1 (middle), and PCNA (bottom) in thymus during the aging process. Parallel blots with total protein (50 mg) from various developmental stages of thymic tissue preparations were examined. Thymus were from 7and 14-day-old and 1-, 2-, 3-, 4-, and 12-month-old rats. The antibodies used for each blot were described under Materials and Methods. Results were similar from two experiments.
AID
ECR 3590
/
6i23$$$182
06-09-97 12:30:36
29
FIG. 6. Schematic representation of seminferous tubules (stages VI–VII) in testis. To demonstrate the transitional characteristics of male germ cell, the cell nuclei are depicted. Spermatogonia (A), preleptene primary spermatocytes (PI), pachytene primary spermatocytes (P), differentiating round spermatids (R), and differentiating elongating spermatids (ES).
lymphocytes, and confirmed sequence by sequencing analysis. The RT-PCR product was inserted as a right open reading frame into NcoI/BamHI site of protein expression vector, pET 3d (Novagen, Madison, WI). The expression of cyclin D3 protein in E. coli host in the presence of 0.3 mM isoprophyl b-D-thiogalactopyranoside was performed essentially as described by Studier et al. [15]. The overexpressed cyclin D3 protein localized in inclusion body was denatured by guanidine hydrochloride and then renatured into a soluble form. Four hundred micrograms of cyclin D3 was mixed with an equal volume of complete Freund’s adjuvant and injected intramuscularly into the thigh and subcutaneously into the back neck area of a rabbit. For booster immunizations, the same amount of cyclin D3 protein mixed with incomplete Freund’s adjuvant was injected into the same rabbit with the same manner every 4 weeks. Bleeding was done 2 weeks after each immunization. Western blot. Testes samples were frozen with liquid nitrogen and stored at 0707C. Samples were homogenized directly in Nonidet P-40 buffer [150 mM NaCl, 5 mM EDTA, 50 mM Tris–HCl (pH 8.0), 1 mg/ml aprotinin, 1 mg/ml pepstatin, 1 mg/ml leupeptin, 50 mg/ml 1chloro-3-toxylamido-7-amino-2-heptanone, 50 mg/ml phenylmethylsulfonyl fluoride, 100 mg/ml L-1-tosylamino-2-phenylethyl chloromethyl ketone, 20 mM sodium fluoride, 20 mM glycerophosphate, 1% (v/v) Nonidet P-40]. Lysate protein was quantitated using a Bradford assay (Bio-Rad, Hercules, CA) with bovine serum albumin as a reference standard. Samples were boiled with 21 sample buffer for 10 min, separated by size on 12.5% polyacrylamide gel under SDS denaturing conditions, and electrotransferred to nitrocellulose membranes. The nitrocellulose membranes were stained with 0.1% naphthol blue black (Amido black) in 10% (v/v) acetic acid and 40% (v/v) methanol to assess the efficiency of transfer. Nonspecific binding was blocked by incubation in blocking buffer (5% nonfat dry milk, 3% bovine serum albumin, 0.1% Tween 20, 11 PBS) for 1 h at room temperature. Anti-cyclin D3 serum (1:200 dilution), anti-cyclin D1
eca
30
KANG ET AL.
FIG. 7. Light micrographs of portions of seminiferous tubules immunostained with anti-cyclin D3 antibody and PCNA antibody in a 7day-old newborn. In tissue sections, cyclin D3 appeared as intense reddish–brown punctated stains and was readily observed in the cytoplasm. PCNA appeared as intense reddish–brown punctated stains in cytoplasm and nuclei among a background of lighter hematoxylinstained nuclei. (A) Some epithelial stem cells showed cyclin D3 immunopositive staining. Magnification, 1200. (B) Cells that had bigger nuclei than surrounding cells had immunopositive cyclin D3 in cytoplasm. Magnification, 11000. (C) Nuclear dividing cells were visible, and some of these cells had abundant cyclin D3 in the cytoplasm. Magnification, 11000. (D, E) 80–90% of cells had immunopositive PCNA in their nuclei. No immunopositive staining of PCNA was observed in cells which had a bigger nucleus than surrounding cells and nuclear dividing cells. Magnification, 11000.
FIG. 8. Light micrographs of portions of seminiferous tubules (A, B, D, E, F) and thymus (C) that were immunostained with anti-cyclin D3 antibody (A, B, C, D, E) and anti-CDK4 antibody (F) in 30-day-old (A) and 3-month-old rats (B, C, D, E, F). (A) Strong cyclin D3 immunopositive staining was located mainly in the cytoplasm of luminal primary spermatocytes. There was no or barely detectable immunopositive cyclin D3 in some seminiferous tubules. Magnification, 1200. (B) A variable immunostaining pattern for cyclin D3 is visualized in different seminiferous tubules. Magnification, 1200. (C) Corticomedullary portion of thymus. Strong cyclin D3 immunopositive staining was located in cytoplasm and nuclei of 10–15% of thymocytes population. Magnification, 11000. (D, stage IV; E, stage IX) Cyclin D3 was abundant in cytoplasm of pachytene primary spermatocytes, but undetectable in elongated spermatids. Magnification, 11000. There was no cyclin D3 in spermatogonia cells or Sertoli cells. (F, stage V) CDK4 was abundant in nuclei of spermatogonia and spermatocytes, but not or barely detectable in cytoplasm of round spermatids. The stages of ST were determined according to Russell et al. (1990).
AID
ECR 3590
/
6i23$$3590
06-09-97 12:30:36
eca
CYCLIN D3 DURING SPERMATOGENESIS
AID
ECR 3590
/
6i23$$3590
06-09-97 12:30:36
eca
31
32
KANG ET AL.
monoclonal (72-13G; Santa Cruz Biotech, Santa Cruz, CA), and antiproliferating cell nuclear antigen (PCNA) monoclonal antibody (PC10; Santa Cruz Biotech) were used. Secondary anti-mouse IgG antibody which was conjugated to horseradish peroxidase (1:2,000) was used. Signals were visualized by the ECL detection method according to the manufacturer’s protocol (Amersham Int. plc, Buckinghamshire, United Kingdom). Immunohistochemistry. For paraffin sections, three testes in each age group were fixed in Bouin’s fixatives, dehydrated through graded alcohols, and infiltrated with paraffin under standard procedures. Tissue blocks were then sectioned at 4 mm. Sections were mounted on positively charged microscope slides (ProbeOn Plus, Fisher Scientific). Deparaffinized sections were hydrated, permeabilized in 0.1% Triton X-100, and blocked with phosphate-buffered saline containing 1% bovine serum albumin and 5% goat serum. Sections were then incubated with anti-cyclin D3 antiserum (1:100 dilution), anti-CDK4 polyclonal antibody (1:50 dilution, C-22, Santa Cruz Biotech) or anti-PCNA (1:100) at 47C overnight, and signals were developed with an avidin-biotin-peroxidase system (Biomeda) with 3-amino-9-ethyl-carbazole (for cyclin D3 and CDK4) or diaminobenzidine (for PCNA) as a chromogen. Sections were counterstained with hematoxylin. Either nonimmune rabbit serum (for polyclonal antibody) or irrelevant monoclonal antibodies were used for control slides. Slides were viewed and photographed with a microscope equipped with differential interface contrast optics (Axiskop with MC-80 camera, Zeiss).
RESULTS
Tissue Distribution and Temporal Levels of Cyclin D3 mRNA and Protein among Various Organs We examined the levels of cyclin D3 mRNA and protein in various organs of adult rats (Fig. 1). The level of cyclin D3 mRNA was the highest in thymus and was readily detectable in heart, spleen, and testis, but was barely detectable or undetectable in other organs including brain, liver, kidney, ovary, and skeletal muscle (Fig. 1, top). The level of cyclin D3 protein was highest in thymus, second highest in testis, third highest in spleen, but was barely detectable or undetectable in other organs including brain, heart, liver, kidney, ovary, and skeletal muscle (Fig. 1, bottom). Thus, rate of translation and degradation of cyclin D3 may differ among various organs. Temporal Levels of Cyclin D3 and Cyclin D1 mRNA and Protein in Testis and Thymus during Normal Aging Process The expression of cyclin D3 mRNA in 7-day testis was high, but the expression of cyclin D3 mRNA decreased, and then remained almost constant in 14-dayto 1-year-old rats (Fig. 2). A previous study reported an induction of cyclin D3 mRNA during testicular development [9]. In contrast to the fairly constant level of cyclin D3 mRNA, the level of cyclin D1 mRNA was high in the testis of 7- and 14-day-old rats and decreased gradually during testicular development (Fig. 2). This result was different from a previous study that reported a marked reduction of cyclin D1 mRNA during
AID
ECR 3590
/
6i23$$$182
06-09-97 12:30:36
testicular development in the mouse [9]. The protein levels of cyclin D3 in testes were high in 7-day-old newborns, remained almost constant until 3 months, and then gradually decreased thereafter (Fig. 3). In testes of 7- and 14-day-old newborns, the protein levels of cyclin D1 were high, but decreased and remained at almost constant levels thereafter (Fig. 3). As a proliferative indicator, the protein levels of PCNA [16] were measured in developing testes (Fig. 3). PCNA protein decreased gradually during testicular development. Overall, levels of cyclin D3, cyclin D1, and PCNA protein decreased at different rates during testicular development. The expression of cyclin D3 mRNA in 7-day thymus was high, remained almost constant until 2 months, and then decreased gradually thereafter (Fig. 4). In contrast to the decrease of cyclin D3 mRNA in the thymus, the level of cyclin D1 mRNA was almost constant from 7 day old to 12 month old rats (Fig. 4). The levels of cyclin D3, cyclin D1 and PCNA proteins were almost constant in thymus of 7-day-old to 12month-old rats (Fig. 5). Thus, the protein levels of cyclin D3 and D1 did not reflect changes of the corresponding mRNA levels in the testis and the thymus of in vivo animals. Discrepancies between levels of cyclin mRNA and protein have been observed by us [17] and others [18] and suggest that at different ages, some organs control the level of cyclins by rate of translation or degradation of proteins rather than by rate of transcription. Immunolocalization of Cyclin D3, CDK4, and PCNA during Testicular Development A cross-section of rat testis contains several hundreds seminferous tubules, which have been classified into 14 different stages. Each stage of seminferous tubule represents a different combination of dividing and differentiating germ cells (Fig. 6). Immunostaining of cyclin D3 protein in tissue sections appeared as intense reddish–brown punctated stains and was readily observed in the cytoplasm among a background of lighter hematoxylin-stained nuclei (Figs. 7 and 8). Immunostaining of PCNA protein in tissue sections also appeared as intense reddish–brown stains and was observed in the hematoxylin-stained nuclei (Figs. 7 and 9). A variable immunostaining pattern was visualized among different developmental periods and, in adult rats, different stages of seminiferous tubules (Figs. 7, 8, and 9). In the seminiferous tubules of 7-day-old newborns, cyclin D3 was mainly located in the cytoplasm of cells that had bigger nuclei than nuclei of surrounding cells (Figs. 7A and 7B). Also, the cells that seemed to be undergoing nuclear division had abundant cyclin D3 (Fig. 7C). Eighty to 90% of stem cells were PCNA immunopositive in 7-day-old newborns (Fig. 7D). The cells with bigger nuclei than surrounding
eca
CYCLIN D3 DURING SPERMATOGENESIS
cells, and nuclear dividing cells that stained positively for cyclin D3, did not stain for PCNA (Fig. 7E). In the 1month-old and older testes, a variable immunostaining pattern for cyclin D3 was visualized in different seminiferous tubules (Figs. 8A and 8B). In 1-month-old rats, some seminiferous tubules showed strong immunostaining of cyclin D3, while others totally lacked cyclin D3. The population of PCNA immunopositive staining cells in seminiferous tubules decreased during aging. Only spermatogonia contained immunopositive PCNA staining (Fig. 9A). In adult rats, cyclin D3 darkly stained primary spermatocytes, lightly stained late stage primary spermatocytes, but did not stain elongated spermatids (Figs. 8D and 8E). There was no cyclin D3 detected in Sertoli cells, interstitial cells, or fibroblasts and blood vessels within the interstitial matrix. The known cyclin D3 partner, CDK4, was localized mainly either in the nuclei of spermatogonia or in early stage primary spermatocytes (Fig. 8F). Detectable amounts (immunostaining positive signal was higher than background) of CDK4 were also localized in the cytoplasm of round spermatids. As a positive control, we examined the immunostaining pattern of cyclin D3 and CDK4 in the thymus. In the thymus, strong cyclin D3 immunopositive staining was located in the cytoplasm and nuclei of thymocytes (10–15% of population) (Fig. 8C). However, strong CDK4-immunopositive staining was located mainly in the nuclei of thymocytes (3–5% of the population; data not shown). The peripheral cells of seminiferous tubules showed strong PCNA immunopositive staining regardless of seminiferous tubule stage in adult rats (Fig. 9B). High magnification (11000) revealed that only spermatogonia and early stage primary spermatocytes showed PCNA immunopositive staining in nuclei, regardless of seminiferous tubule stage (Figs. 9C, 9D, and 9E). Thus, the percentage of PCNA immunopositive staining cells within seminiferous tubules decreased from 7-day-old to adult rats (7 days old, 86.2%; 30 days old, 22.3%; 3 months old, 12.6%). Controls slides, which were treated with either nonimmune rabbit serum (for polyclonal antibody) or irrelevant monoclonal antibodies, and immunofluorescence-treated slides were used to exclude the possibility of false positives by endogenous peroxidase, and were not immunoreactive (data not shown). DISCUSSION
In the present study, a tissue survey of rat tissues revealed that cyclin D3 mRNA was highly detectable in thymus, spleen, testis, and heart, but cyclin D3 protein was highly detectable only in thymus and testis. The levels of cyclin D3 protein in testes were high in 7-day-old newborns, remained almost constant until 3 months, and then gradually decreased thereafter. Surprisingly, cyclin D3 was located in the cytoplasm of
AID
ECR 3590
/
6i23$$$183
06-09-97 12:30:36
33
male germ cells. Cyclin D3 was abundant in primary spermatocytes, but was not present in either spermatogonia or elongated spermatids. CDK4 was localized mainly in nuclei of spermatogonia and early stage primary spermatocytes. PCNA was abundant only in the nuclei of spermatogonia and early stage primary spermatocytes. These results indicated that cyclin D3 is detectable in meiotically active male germ cells (PCNA-negative cells) but is absent from mitotically active spermatogonia (PCNA-positive cells). Ravnik et al. [9] reported that in mouse testis, the expression of cyclin D3 mRNA is highest in elongating spermatids, two of the three different sizes of cyclin D3 mRNA transcripts are germ cell specific, and the 2.3-kb cyclin D3 transcript increases while the expression of cyclin D1 mRNA decreases during aging. Moreover, mRNAs of cyclin D1 and D3 are expressed in nondividing rather than mitotically active germ cells. However, protein levels and immunolocalization of cyclin D3 have not been reported. Data from the present study differed from the report by Ravnik et al. (1995) in that (1) levels of cyclin D3 mRNA and protein decreased during testicular development, and (2) rat testis expressed only one size of cyclin D3 mRNA transcript. Species differences may account for these discrepancies. Similarities between the two reports are as follows: (1) expression of cyclin D1 mRNA decreased during testicular development and (2) cyclin D3 was expressed in nondividing, rather than mitotically active, male germ cells. Our most remarkable finding was the restricted presence of cyclin D3 in the cytoplasm of germ cells. Cyclin D3 is located in the nucleus of HT2T lymphocytes [19] and since cyclin D3 is necessary for the progression of the cell cycle in vitro, its location would be expected to be in the nucleus. In the thymus, we detected positive cyclin D3 staining in the nuclei and cytoplasm of thymocytes. However, using the same antibody, cyclin D3 was detected only in the cytoplasm of meiotically active male germ cells. The known cyclin D3 partner, CDK4, was localized mainly either in the nuclei of some spermatogonia cells or in some early stage primary spermatocytes. CDK4 was also detected in the cytoplasm of round spermatids. In proliferating melanoma cells, CDK4 is also present in nuclei and/or cytoplasm (20). Because of its location, a major portion of cyclin D3 may not interact with CDK4 in the nuclei of spermatocytes in vivo. Thus, the question arises, why is cyclin D3 present in the cytoplasm and not in the nucleus? An underlying mechanism for regulation of many processes in the cell is the localization of protein complexes within specific subcellular compartments. Some cyclin–CDK complexes are subject to this mode of regulation [21–23]. For example, the M-phase cyclin, cyclin B1, is translocated from the cytoplasm to the nucleus immediately prior to mitosis; the timing of this translocation may be a critical factor in determining the initia-
eca
34
AID
KANG ET AL.
ECR 3590
/
6i23$$3590
06-09-97 12:30:36
eca
CYCLIN D3 DURING SPERMATOGENESIS
tion of mitosis. In vivo, cyclin A is translocated from the cytoplasm to the nucleus of developing spermatocytes in testes of Drosophilia [24], but this translocation of cyclin A in spermatocytes does not occur in mice indicating species specificity [10]. Interestingly, cyclin D1 was shown to interact and to activate estrogen receptors independent of CDKs in breast epithelial cells [25] and cyclin D2 was shown to interact with the transcriptional factor, DMP-1 [26]. Thus, cyclin D3 may associate with yet unidentified protein kinases or interact with other proteins located in the cytoplasm and have functions other than G1-phase progression. D-type cyclins (D1, D2, D3) are expressed earlier in the G1 phase of the cell cycle than cyclin E, and they are known to link growth factor stimulation to the activation of S-phase genes [4, 13]. Furthermore, D-type cyclins are necessary and rate limiting for G1 progression during the mitotic cell cycle [4, 5], although the expression of cyclin D3 is quite different from expressions of cyclins D1 and D2. In 3T3 fibroblasts, a modest induction of cyclin D3 mRNA occurs near the G1/S boundary, 12 h after serum stimulation, while cyclin D1 mRNA increases 20- to 30-fold, peaking 6 h after addition of serum [27]. In human T lymphocytes, cyclin D2 mRNA accumulates in the early G1 phase, whereas cyclin D3 mRNA accumulates during the late G1 stage [28]. In cultured rat hepatocytes, cyclin D3 mRNA accumulates in the absence of mitogenic signals, yet mitogenic signals are necessary for the accumulation of cyclins D1 and D2 mRNAs [29]. Furthermore, when skeletal myoblasts are induced to differentiate with growth factors and serum deprivation, cyclin D3 mRNA increases, while mRNA levels of cyclin D1 and D2 rapidly decrease [30, and our unpublished observations]. Thus, although D-type cyclins regulate G1 phase progression, the functions of D-type cyclins may not be redundant. Moreover, there are accumulating reports that in vivo and in vitro expressions of D-type cyclins are quite different [9, 17, 31]. Unlike other members of the cyclin family, D-type cyclins are expressed in some nondividing cell types in vivo [9, 17, 18, 31]. For example, the heart and the kidney express D-type cyclins mRNAs after the cell cycle is arrested in young animals, while at the same time, cyclins A and B rapidly fall to undetectable levels [17, 31]. Genetically, overexpression of cyclin D1 is sufficient to alter the cell cycle in mammary tissue [32] but does not alter the cell cycle in lymphoid tissue [33, 34]. Thus, cyclin D1
35
is not the only limiting regulatory factor for proliferation and its effects may differ among tissues. Cyclin D3 was characterized by three groups [12, 35, 36] and the biological function of cyclin D3 was established by in vitro biochemical and cell culture studies. In the insect cell Sf9, cyclin D3 interacts with the serine threonine protein kinase, cdk2, and phosphorylates retinoblastoma protein (Rb) to relieve Rb’s inhibition of proliferation [37]. However, the in vivo function of cyclin D3 remains to be clarified. The testis provides a unique system in which to study cyclin D3 and the cell cycle because within a seminiferous tubule (1) mitotically active spermatogonia, meiotically active spermatocytes, and post-meiotic spermatids exist and (2) high levels of cyclin D3 protein are present. The seminiferous tubules of 7-day-old rats contain mitotically active epithelial stem cells. Cyclin D3 was located in cells with nuclei larger than the nuclei of surrounding cells. A positive signal was observed from the mitotic indicator, PCNA, in almost all stem cells except those with large nuclei. This indicates that cyclin D3 was not present in mitotically active cells. The cells with large nuclei may have been quiescient or undergoing the first stages of meiosis, but were not apoptotic as determined by the TUNEL method (data not shown). By 30 days postnatal, meiosis has already begun in spermatocytes, and the size of the epithelium is nearly that of the adult [38]. In 30-day-old rats, positive staining for cyclin D3 was detected in primary spermatocytes, but not in mitotically active spermatogonia. In contrast, positive staining for PCNA was detected in spermatogonia, but not in spermatocytes. This contrasting pattern of staining for PCNA and cyclin D3 was also observed in adult testis regardless of the stage of seminiferous tubules. Cyclin D3 associates with PCNA and with various CDKs in diploid lung fibroblasts [37, 39]. However, our in vivo data indicates that cyclin D3 and CDK4 were not present together in the same cells as PCNA. In the adult testis, cyclin D3 was most abundant in pachytene spermatocytes and gradually decreased in round and early stage elongating spermatids as they matured and moved toward the lumen. Since cyclin D3 was not detectable in elongated spermatids, cyclin D3 may not be needed for the final maturation of spermatids. In summary, we have developed an anti-murine cyclin D3 antibody that detects high levels of cyclin D3 in testis and thymus. Cyclin D3 is detectable in
FIG. 9. Light micrographs of portions of seminiferous tubules that were immunostained with PCNA antibody in 30-day-old (A) and 3month-old rats (B, C, D, E). (A) The population of PCNA-immunopositive-staining cells decreased. Spermatogonia cells showed PCNA immunopositive staining. Magnification, 11000. (B) Peripheral germ cells showed strong PCNA-immunopositive staining regardless of stage of seminiferous tubules. Magnification, 1200. (C, stage XI-XII; D (top), stage XII; D (bottom), stages II–III; E, stage VIII). High magnification revealed that only spermatogonia and early phase primary spermatocytes showed immunopositive PCNA. Magnification, 11000. The stages of seminiferous tubules were determined according to Russell et al. (1990).
AID
ECR 3590
/
6i23$$$183
06-09-97 12:30:36
eca
36
KANG ET AL.
meiotically active male germ cells and increases during development, but is conspicously absent from mitotically active spermatogonia. In contrast to in vitro reports, cyclin D3 is not located in the nucleus and is not present with CDK4 and PCNA in same cell, but rather is located in the cytoplasm of male germ cells in vivo. The presence and cytoplasmic location of cyclin D3 in spermatocytes lead us to speculate that cyclin D3 may have functions in male germ cells other than mitosis. We thank Ms. Keum Nim Koh, Ms. Hae Young Park, and Mr. Won Jin Cho for excellent technical assistance, This work was supported by grants from Korea Science and Engineering Foundation (95-040301-01-3) and from the Ministry of Education for promoting Genetic Engineering in 1994.
REFERENCES 1. Hunter, T., and Pines, J. (1991) Cell 66, 1071–1074. 2. Murray, A., and Hunt, T. (1993) in The Cell Cycle, Freeman, New York. 3. Hunter, T., and Pines, J. (1994) Cell 79, 573–582. 4. Sherr, C. J. (1993) Cell 73, 1059–1065. 5. Sherr, C. J. (1994) Cell 79, 551–555. 6. Morgan, D. O. (1995) Nature 374, 131–134. 7. Russell, L. D., Ettlin, R. A., Hikim, A. P. S., and Clegg, E. D. (1990) in Histological and Histopathological Evaluation of the Testis, Cache River Press, Clearwater, FL. 8. Chapman, D. L., and Wolgemuth, B. J. (1994) Dev. Biol. 165, 500–506. 9. Ravnik, S. E., Lee, K., and Wolgemuth, D. J. (1995) Dev. Genet. 16, 171–178. 10. Ravnik, S. E., and Wolgemuth, D. J. (1996) Dev. Biol. 173, 69– 78. 11. Sambrook, J., Fritsch, E. F., and Maniatis, T. (1989) in Molecular Cloning: A Laboratory Manual, 2nd ed., Cold Spring Harbor Laboratory, Cold Spring Harbor, NY. 12. Motokura, T., Keyomars, K., Kronenberg, M. H., and Arnold, A. (1992) J. Biol. Chem. 267, 20412–20415. 13. Matsushime, H., Roussel, M. F., Ashmun, R. A., and Sherr, C. J. (1991) Cell 65, 701–713. 14. Fort, P., Marty, L., Piechaczyk, M., Sabrouty, S. E., Dani, C., Jeanteur, P., and Blanchard, J. M. (1985) Nucleic Acids Res. 13, 1431–1442. 15. Studier, F. W., Rosenberg, A. H., Dunn, J. J., and Dubendorff, J. W. (1990) Methods Enzymol. 185, 60–89.
16. Bravo, R., Frank, R., Blundell, P. A., and MacDonald-Bravo, H. (1987) Nature 326, 515–517. 17. Park, S. K., Kang, S. K., Lee, D. Y., Kang, M. J., Kim, S. H., and Koh, G. Y. (1997) Kidney Int., in press. 18. Tamaru, T., Okada, M., and Nakagawa, H. (1994) Neurosci. Lett. 168, 229–232. 19. Vallance, S. J., Lee, H. M., Roussel, M. F., Shurtleff, S. A., Kato, J. Y., Strom, D. K., and Sherr, C. J. (1994) Hybridoma 13, 37– 44. 20. Wang, Y. L., Uhara, H., Yamazaki, Y., Nikaido, T., and Saida, T. (1996) Br. J. Dermatol. 134, 269–275. 21. Pines, J., and Hunter, T. (1991) J. Cell. Biol. 115, 1–17. 22. Bailly, E., Pines, J., Hunter, T., and Bornens, M. (1992) J. Cell. Sci. 101, 529–545. 23. Baldin, V., and Ducommun, B. (1995) J. Cell Sci. 108, 2425– 2432. 24. Eberhart, C. G., Maines, J. Z., and Wasserman, S. A. (1996) Nature 381, 783–785. 25. Zwijsen, R. M., Wientjens, E., Klompmaker, R., Sman, J. V. D., Bernards, R., and Michalides, R. J. A. M. (1997) Cell 88, 405– 415. 26. Hirai, H., and Sherr, C. J. (1996) Mol. Cell Biol. 16, 6457–6467. 27. Winston, J. T., and Pledger, W. J. (1993) Mol. Biol. Cell 4, 1133– 1144. 28. Ajchenbaumet, F., Ando, K., DeCaprio, J. A., and Griffin, J. D. (1993) J. Biol. Chem. 268, 4113–4119. 29. Loyer, P., Cariou, S., Glaise, D., Bilodeau, M., Baffet, G., and Guguen-Guillouzo, C. (1996) J. Biol. Chem. 271, 11484–11492. 30. Rao, S. S., and Kohtz, D. S. (1995) J. Biol. Chem. 270, 4093– 4100. 31. Yoshizumi, M., Lee, W. S., Hsieh, C. M., Tsai, J. C., Li, J., Perrella, M. A., Patterson, C., Endegr, W. O., Schlegel, R., and Lee, M. E. (1995) J. Clin. Invest. 95, 2275–2280. 32. Wang, T. C., Cardiff, R. D., Zukerberg, L., Lees, E., Arnold, A., and Schmidt, E. V. (1994) Nature 369, 669–671. 33. Lovec, H., Grzeschiczek, A., and Moroy, T. (1994) EMBO J. 13, 3487–3495. 34. Bodrug, S., Warner, B., Bath, M., Lindeman, G., Harris, A., and Adams, J. (1994) EMBO J. 13, 2124–2130. 35. Inaba, T., Matsushime, H., Valentine, M., Roussel, M. F., Sherr, C. J., and Look, A. T. (1992) Genomics 13, 565–674. 36. Xiong, Y., Menninger, J., Beach, D., and Ward, D. C. (1992) Genomics 13, 575–584. 37. Ewen, M. E., Sluss, H. K., Sherr, C. J., Matsushime, H., Kato, J., and Livingston, D. M. (1993) Cell 73, 487–497. 38. Hervieu, G., Segretain, D., and Nachon, J. L. (1996) Biol. Reprod. 54, 1161–1172. 39. Xiong, Y., Zhang, H., and Beach, D. (1992) Cell 71, 505–514.
Received November 6, 1996 Revised version received March 14, 1997
AID
ECR 3590
/
6i23$$$184
06-09-97 12:30:36
eca