- Email: [email protected]
Cell Cycle Control with Minimal Participation of Cdk2 in a Murine Fibrosarcoma Clone Cultured in Protein-free Medium
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS ARTICLE NO.
232, 622–625 (1997)
RC976346
Cell Cycle Control with Minimal Participation of Cdk2 in a Murine Fibrosarcoma Clone Cultured in Protein-free Medium Yoh Dobashi,* Masaki Chigira,† Mitsuhiko Shoji,* Yoko Wakata,* Masatoshi Kitagawa,‡ Hidekazu Tamauchi,§ Tetsu Akiyama,Ø and Toru Kameya*,1 *Department of Pathology and §Department of Microbiology, Kitasato University School of Medicine, Sagamihara, Kanagawa, Japan; †Department of Orthopedic Surgery, Gunma University School of Medicine, Maebashi, Gunma, Japan; ‡Tsukuba Research Institute Banyu Pharmaceutical Co. Ltd., Tsukuba, Ibaragi, Japan; and Ø Department of Oncogene Research, Research Institute for Microbial Diseases, Osaka University, Osaka, Japan
Received February 14, 1997
The differences in the protein expression of cyclins, cyclin-dependent kinases (cdks), and their inhibitors and cdk kinase activities were examined in serum dependent (SD) and independent (PF) clones of the murine fibrosarcoma cell line, Gc-4. The expression of cyclin A in SD was minimal in contrast to PF. Furthermore, cdk2 kinase activity in PF was remarkably lower than that in SD, yet the G1/S transition in PF appeared normal. PF was also resistant against the selective inhibitor of cdk2, butyrolactone I. These findings suggest that tumor cell proliferation and tumor progression can be promoted by the activation of a molecule(s) downstream of cdk2. q 1997 Academic Press
In response to extracellular signals acting through cell surface receptors, many effectors within the cells function to promote cell proliferation (1). These include the G1 cyclins as well as their catalytic partners, the cdks, which are proving to be integrators of growth factor-mediated signals driving the cell cycle through the G1 restriction point (2,3). We have previously established a cell line (Gc4-PF) which exhibits exponential growth in a protein-free medium and is suitable as an in vitro model for the analysis of autocrine circuits (4). This cell, line which survives under protein-free (PF) conditions was derived from its parental serumdependent (SD) clone, Gc-4SD, a murine fibrosarcoma (4). Several differences in the biological characteristics of PF and SD cells have been demonstrated: PF cells are resistant to the protein kinase inhibitors, K-252 1
and staurosporine whereas SD cells exhibit apoptosis without internucleosomal cleavage when transferred to protein-free medium (5,6). It seemed most likely that PF cells produced some sort of autocrine proliferation signal which is not present in a parental cell line, SD. Therefore, we examined differences in the expression of cyclins, cdks and their inhibitors as well as cyclindependent kinase activities to identify the critical differences in the regulation of the cell cycle in SD and PF cells. MATERIALS AND METHODS Cell culture and assay of cell growth. SD and PF clones of Gc-4 were established as previously described (4). SD was maintained in RPMI supplemented with 5% FBS. PF was maintained in defined medium which consisted of RPMI supplemented with 10 mM L-glutamine, 100 mg/l sodium pyruvate, 10 mM HEPES and 2.0 g/l sodium hydrocarbonate. Cell growth was assayed in the absence or presence (30 or 60 mg/ml) of Butyrolactone I (BL-I) as described (5,7,8). Immunoblotting analysis. Cells were lysed in high-salt lysis buffer (8) and 80 mg of protein was used for immunoblotting. For immunoprecipitation followed by immunoblotting, cells were lysed in NP-40 lysis buffer (9). Lysates (200 mg of protein) were incubated with anti-cdk2 antibody immobilized on Sepharose, or with p13suc1Sepharose for 4 hours at 47C. The precipitates were used for immunoblotting analysis of cyclin A, cyclin E, p21 and p27. Antibodies. Polyclonal antibodies used were as follows; cyclin A (1:100, Upstate Biochemicals Incorporated; UBI), cyclin E (1:150, UBI), p53 (1:500, Novocastra), cdk4 (1:120, Santa Cruz). Polyclonal antibodies against cyclin D1 (1:150), cdk2 (1:150) and p21 (1:100) were prepared by immunizing rabbits with synthetic peptides (8). A polyclonal antibody against p27 was a generous gift from Dr. Matsushime (University of Tokyo). In vitro kinase reactions. For cdk2 complex kinase assays, cells were lysed in solubilizing buffer (8). For cdk4 complex kinase assays, cells were lysed in IP buffer followed by sonication (10). Lysates (500
Correspondence. Fax: 81-427-78-8441.
622
0006-291X/97 $25.00 Copyright q 1997 by Academic Press All rights of reproduction in any form reserved.
AID
BBRC 6346
/
6924$$$161
03-11-97 06:29:16
bbrcg
AP: BBRC
Vol. 232, No. 3, 1997
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
FIG. 1. Protein levels of cell cycle regulators expressed in Gc-4 SD and PF cells. Lysates (80 mg of protein) were subjected to immunoblotting analysis to detect cyclins A, D1, and E, cdk2, cdk4, p21, p27, and p53.
mg of protein) were used for cdk2 and cdk4 kinase assay as described previously (8-10). Densitometric analysis. Densitometric quantification of the data was done with a dual-wave length flying-spot scanner (Shimadzu Ltd.). Fluorescence-activated cell sorter (FACS) analysis. One million exponentially-growing cells of each cell line were trypsinized and utilized for propidium iodide (PI) staining (12). The PI fluorescence of individual nuclei was analysed using a FACScan flow cytometer (Becton & Dickinson).
inhibitor of cdk2, BL-I. Although, in the presence of BL-I, the growth of SD cells declined significantly in a dose-dependent manner, PF cells did not show a significant reduction in cell growth even in the presence of 60 mg/ml of BL-I (Fig. 3). Immunoblotting analysis of cdk-associated cyclin A, cyclin E, p21, and p27. To evaluate the levels of cyclin
RESULTS AND DISCUSSION Expression of cyclins, cdks, and cdk-inhibitors. SD and PF cells were lysed and the extracts were subjected to immunoblotting analysis. There was no difference in the expression of cyclins D1, E, cdk2, cdk4, p21, p27 or p53 (Fig. 1). However, a significantly lower level of cyclin A expression (20-fold lower) was observed in SD compared to PF (Fig. 1). Kinase activity of cdk2 and cdk4. Cdk2 and cdk4 kinase activities were examined from the lysates of SD and PF cells. PF extract exhibited 15-fold lower cdk2 activity than SD extract as determined by densitometric analysis (Fig. 2). There was no remarkable difference in cdk4 activities between these cell lines. Effect of butyrolactone I on the cell growth. To determine whether the kinase activity of cdk2 is dispensable for PF, the growth of SD and PF cells were examined by culturing these cell lines in the presence of selective
FIG. 2. Kinase activities associated with CDK2 and CDK4. Kinase activities associated with cdk2 and cdk4 in Gc-4 SD and PF cells are shown. Lysates prepared from these two cell lines (500 mg of protein) were immunoprecipitated with anti-cdk2 or anti-cdk4 antibodies. The immunocomplexes were assayed for kinase activity using GST-RB fusion protein as a substrate.
623
AID
BBRC 6346
/
6924$$$162
03-11-97 06:29:16
bbrcg
AP: BBRC
Vol. 232, No. 3, 1997
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
FIG. 3. Growth rates of cell lines. Growth rates of SD (a) and PF (b) cells in the absence or presence of Butyrolactone I (30 or 60 mg/ ml). Cell numbers represent the mean values of triplicate experiments.
A, cyclin E, p21 and p27 associated with cdk2, we subjected p13suc1-precipitates and cdk2-immunoprecipitates prepared from these 2 clones to immunoblotting analysis. As shown in Fig. 4, the amount of cyclin A associated with p13suc1-sepharose-bound cdk2 was very low but similar between SD and PF. This is in contrast
FIG. 4. Levels of cdk-associated cyclin A, cyclin E, and p21. Lysates prepared from Gc-4 SD and PF cells (200 mg of protein) were precipitated with p13suc1 (for cyclin A and cyclin E) or immunoprecipitated with anti-cdk2 antibody (for p21 and p27). The immunocomplexes were subjected to immunoblotting analysis with the corresponding polyclonal antibodies.
to the previous result showing that the total level of cyclin A is much lower in SD (Fig. 1). This suggests that the level of the catalytically active form of cyclin A is almost identical in both clones. With regards to cyclin E, p21 and p27, there was no significant difference in their respective cdk2-associated fractions between SD and PF (Fig. 4). FACS analysis. Because cdk2 promotes the cell cycle transition from G1 to S through the G2 phase in association with cyclins E and A (2,3), we analysed the cell cycle distribution of these clones by FACS. As shown in Fig. 5, 29.5% of proliferating SD cells were found to be in the S phase, compared to 40.7% of the PF cells. Gc-4 PF cells transferred to protein-free chemically-defined medium do not undergo growth arrest, but rather continue to proliferate exponentially (4). In contrast, SD cells undergo apoptosis when transferred into protein-free medium. These findings led us to hypothesize that some cell-cycle regulators in PF may have become autonomously activated. The present study confirms this hypothesis. The most surprising result was that cdk2 kinase activity in PF was significantly lower than that in SD. The reason for this lower kinase activity is unclear because we detected an almost identical amount of cyclin E, cyclin A, p21 and p27 associated with cdk2. However, this low cdk2 kinase activity did not prevent the G1/ S transition, since FACS analysis showed that PF cells have a higher ratio of cells in the S-phase compared to SD cells. Staurosporine (ST) has been recently reported to cause G1 arrest by affecting phos-
624
AID
BBRC 6346
/
6924$$$162
03-11-97 06:29:16
bbrcg
AP: BBRC
Vol. 232, No. 3, 1997
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
FIG. 5. DNA histogram of Gc-4 SD and PF cells. Nuclei of Gc-4 SD and PF cells were stained with propidium iodide and analysed by flow cytometry. The data shown below indicate the percentage of cells in the different phases of the cell cycle.
phorylation of retinoblastoma protein (pRB) and other molecules which lie upstream of cdk2 (13). ST as well as a more potent and selective inhibitor of cdk2, BL-1, did not inhibit cell growth in PF (5). From these accumulated data and notions, we hypothesize that a molecule(s) downstream of cdk2 might be aberrantly activated in PF cells. Alternatively, there may exist a growth-promoting pathway independent of cdk2 which is activated. However, we did not detect any significant difference in the expression or phosphorylation state of pRB, one of the substrates for cdk2, by immunoblotting analysis using several kinds of anti-pRB antibodies (data not shown). Probably, pRB is phosphorylated by cdk4 which is identical in the protein expression and kinase activity in both clones. In conclusion, the minimal level of cdk2 kinase activity observed in PF suggests to us the novel idea that in some cases activation of molecules downstream of cdk2 may cause abnormal cell proliferation, and possibly tumor progression, in the absence of significant cdk2 activity.
REFERENCES 1. 2. 3. 4. 5. 6. 7.
8. 9. 10. 11. 12. 13.
Ullrich, A., and Schlessinger, J. (1990) Cell 61, 203–212. Hunter, T., and Pines, J. (1994) Cell 79, 573–582. Sherr, C. J. (1994) Cell 79, 551–555. Watanabe, H., Arita, S., and Chigira, M. (1991) In Vitro Cell Develop. Biol. 27A, 267–269. Chigira, M., Noda, K., and Watanabe, H. (1993) Int. J. Oncol. 2, 67–72. Watanabe, H., Kanbe, K., Shinozaki, T., Hoshino, H., and Chigira, M. (1995) Int. J. Cancer 62, 191–198. Kitagawa, M., Okabe, T., Ogino, H., Matsumoto, H., Takahashi, I., Kokubo, T., Higashi, H., Saito, S., Taya, Y., Yasuda, H., Ohba, Y., Nishimura, S., Tanaka, N., and Okuyama, A. (1993) Oncogene 8, 2425–2432. Dobashi, Y., Kudoh, T., Matsumine, A., Toyoshima, K., and Akiyama, T. (1995) J. Biol. Chem. 270, 23031–23037. Xiong, Y., Zhang, H., and Beach, D. (1992) Cell 71, 505–514. Akiyama, T., Ohuchi, T., Sumida, S., Matsumoto, K., and Toyoshima, K. (1992) Proc. Natl. Acad. Sci. USA 89, 7900–79049. Matsushime, H., Quelle, D. E., Shurtleff, S. A., Shibuya, M., Sherr, C. J., and Kato, J. (1994) Mol. Cell. Biol. 14, 2066–2076. Nicoletti, I., Migliorati, G., Pagliacci, M. C., Grignani, F., and Riccardi, C. (1991) J. Immunol. Methods 139, 271–279. Schnier, J. B., Nishi, K., Goodrich, D. W., and Bradbury, E. M. (1996) Proc. Natl. Acad. Sci. USA 93, 5941–5946.
625
AID
BBRC 6346
/
6924$$$162
03-11-97 06:29:16
bbrcg
AP: BBRC
232, 622–625 (1997)
RC976346
Cell Cycle Control with Minimal Participation of Cdk2 in a Murine Fibrosarcoma Clone Cultured in Protein-free Medium Yoh Dobashi,* Masaki Chigira,† Mitsuhiko Shoji,* Yoko Wakata,* Masatoshi Kitagawa,‡ Hidekazu Tamauchi,§ Tetsu Akiyama,Ø and Toru Kameya*,1 *Department of Pathology and §Department of Microbiology, Kitasato University School of Medicine, Sagamihara, Kanagawa, Japan; †Department of Orthopedic Surgery, Gunma University School of Medicine, Maebashi, Gunma, Japan; ‡Tsukuba Research Institute Banyu Pharmaceutical Co. Ltd., Tsukuba, Ibaragi, Japan; and Ø Department of Oncogene Research, Research Institute for Microbial Diseases, Osaka University, Osaka, Japan
Received February 14, 1997
The differences in the protein expression of cyclins, cyclin-dependent kinases (cdks), and their inhibitors and cdk kinase activities were examined in serum dependent (SD) and independent (PF) clones of the murine fibrosarcoma cell line, Gc-4. The expression of cyclin A in SD was minimal in contrast to PF. Furthermore, cdk2 kinase activity in PF was remarkably lower than that in SD, yet the G1/S transition in PF appeared normal. PF was also resistant against the selective inhibitor of cdk2, butyrolactone I. These findings suggest that tumor cell proliferation and tumor progression can be promoted by the activation of a molecule(s) downstream of cdk2. q 1997 Academic Press
In response to extracellular signals acting through cell surface receptors, many effectors within the cells function to promote cell proliferation (1). These include the G1 cyclins as well as their catalytic partners, the cdks, which are proving to be integrators of growth factor-mediated signals driving the cell cycle through the G1 restriction point (2,3). We have previously established a cell line (Gc4-PF) which exhibits exponential growth in a protein-free medium and is suitable as an in vitro model for the analysis of autocrine circuits (4). This cell, line which survives under protein-free (PF) conditions was derived from its parental serumdependent (SD) clone, Gc-4SD, a murine fibrosarcoma (4). Several differences in the biological characteristics of PF and SD cells have been demonstrated: PF cells are resistant to the protein kinase inhibitors, K-252 1
and staurosporine whereas SD cells exhibit apoptosis without internucleosomal cleavage when transferred to protein-free medium (5,6). It seemed most likely that PF cells produced some sort of autocrine proliferation signal which is not present in a parental cell line, SD. Therefore, we examined differences in the expression of cyclins, cdks and their inhibitors as well as cyclindependent kinase activities to identify the critical differences in the regulation of the cell cycle in SD and PF cells. MATERIALS AND METHODS Cell culture and assay of cell growth. SD and PF clones of Gc-4 were established as previously described (4). SD was maintained in RPMI supplemented with 5% FBS. PF was maintained in defined medium which consisted of RPMI supplemented with 10 mM L-glutamine, 100 mg/l sodium pyruvate, 10 mM HEPES and 2.0 g/l sodium hydrocarbonate. Cell growth was assayed in the absence or presence (30 or 60 mg/ml) of Butyrolactone I (BL-I) as described (5,7,8). Immunoblotting analysis. Cells were lysed in high-salt lysis buffer (8) and 80 mg of protein was used for immunoblotting. For immunoprecipitation followed by immunoblotting, cells were lysed in NP-40 lysis buffer (9). Lysates (200 mg of protein) were incubated with anti-cdk2 antibody immobilized on Sepharose, or with p13suc1Sepharose for 4 hours at 47C. The precipitates were used for immunoblotting analysis of cyclin A, cyclin E, p21 and p27. Antibodies. Polyclonal antibodies used were as follows; cyclin A (1:100, Upstate Biochemicals Incorporated; UBI), cyclin E (1:150, UBI), p53 (1:500, Novocastra), cdk4 (1:120, Santa Cruz). Polyclonal antibodies against cyclin D1 (1:150), cdk2 (1:150) and p21 (1:100) were prepared by immunizing rabbits with synthetic peptides (8). A polyclonal antibody against p27 was a generous gift from Dr. Matsushime (University of Tokyo). In vitro kinase reactions. For cdk2 complex kinase assays, cells were lysed in solubilizing buffer (8). For cdk4 complex kinase assays, cells were lysed in IP buffer followed by sonication (10). Lysates (500
Correspondence. Fax: 81-427-78-8441.
622
0006-291X/97 $25.00 Copyright q 1997 by Academic Press All rights of reproduction in any form reserved.
AID
BBRC 6346
/
6924$$$161
03-11-97 06:29:16
bbrcg
AP: BBRC
Vol. 232, No. 3, 1997
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
FIG. 1. Protein levels of cell cycle regulators expressed in Gc-4 SD and PF cells. Lysates (80 mg of protein) were subjected to immunoblotting analysis to detect cyclins A, D1, and E, cdk2, cdk4, p21, p27, and p53.
mg of protein) were used for cdk2 and cdk4 kinase assay as described previously (8-10). Densitometric analysis. Densitometric quantification of the data was done with a dual-wave length flying-spot scanner (Shimadzu Ltd.). Fluorescence-activated cell sorter (FACS) analysis. One million exponentially-growing cells of each cell line were trypsinized and utilized for propidium iodide (PI) staining (12). The PI fluorescence of individual nuclei was analysed using a FACScan flow cytometer (Becton & Dickinson).
inhibitor of cdk2, BL-I. Although, in the presence of BL-I, the growth of SD cells declined significantly in a dose-dependent manner, PF cells did not show a significant reduction in cell growth even in the presence of 60 mg/ml of BL-I (Fig. 3). Immunoblotting analysis of cdk-associated cyclin A, cyclin E, p21, and p27. To evaluate the levels of cyclin
RESULTS AND DISCUSSION Expression of cyclins, cdks, and cdk-inhibitors. SD and PF cells were lysed and the extracts were subjected to immunoblotting analysis. There was no difference in the expression of cyclins D1, E, cdk2, cdk4, p21, p27 or p53 (Fig. 1). However, a significantly lower level of cyclin A expression (20-fold lower) was observed in SD compared to PF (Fig. 1). Kinase activity of cdk2 and cdk4. Cdk2 and cdk4 kinase activities were examined from the lysates of SD and PF cells. PF extract exhibited 15-fold lower cdk2 activity than SD extract as determined by densitometric analysis (Fig. 2). There was no remarkable difference in cdk4 activities between these cell lines. Effect of butyrolactone I on the cell growth. To determine whether the kinase activity of cdk2 is dispensable for PF, the growth of SD and PF cells were examined by culturing these cell lines in the presence of selective
FIG. 2. Kinase activities associated with CDK2 and CDK4. Kinase activities associated with cdk2 and cdk4 in Gc-4 SD and PF cells are shown. Lysates prepared from these two cell lines (500 mg of protein) were immunoprecipitated with anti-cdk2 or anti-cdk4 antibodies. The immunocomplexes were assayed for kinase activity using GST-RB fusion protein as a substrate.
623
AID
BBRC 6346
/
6924$$$162
03-11-97 06:29:16
bbrcg
AP: BBRC
Vol. 232, No. 3, 1997
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
FIG. 3. Growth rates of cell lines. Growth rates of SD (a) and PF (b) cells in the absence or presence of Butyrolactone I (30 or 60 mg/ ml). Cell numbers represent the mean values of triplicate experiments.
A, cyclin E, p21 and p27 associated with cdk2, we subjected p13suc1-precipitates and cdk2-immunoprecipitates prepared from these 2 clones to immunoblotting analysis. As shown in Fig. 4, the amount of cyclin A associated with p13suc1-sepharose-bound cdk2 was very low but similar between SD and PF. This is in contrast
FIG. 4. Levels of cdk-associated cyclin A, cyclin E, and p21. Lysates prepared from Gc-4 SD and PF cells (200 mg of protein) were precipitated with p13suc1 (for cyclin A and cyclin E) or immunoprecipitated with anti-cdk2 antibody (for p21 and p27). The immunocomplexes were subjected to immunoblotting analysis with the corresponding polyclonal antibodies.
to the previous result showing that the total level of cyclin A is much lower in SD (Fig. 1). This suggests that the level of the catalytically active form of cyclin A is almost identical in both clones. With regards to cyclin E, p21 and p27, there was no significant difference in their respective cdk2-associated fractions between SD and PF (Fig. 4). FACS analysis. Because cdk2 promotes the cell cycle transition from G1 to S through the G2 phase in association with cyclins E and A (2,3), we analysed the cell cycle distribution of these clones by FACS. As shown in Fig. 5, 29.5% of proliferating SD cells were found to be in the S phase, compared to 40.7% of the PF cells. Gc-4 PF cells transferred to protein-free chemically-defined medium do not undergo growth arrest, but rather continue to proliferate exponentially (4). In contrast, SD cells undergo apoptosis when transferred into protein-free medium. These findings led us to hypothesize that some cell-cycle regulators in PF may have become autonomously activated. The present study confirms this hypothesis. The most surprising result was that cdk2 kinase activity in PF was significantly lower than that in SD. The reason for this lower kinase activity is unclear because we detected an almost identical amount of cyclin E, cyclin A, p21 and p27 associated with cdk2. However, this low cdk2 kinase activity did not prevent the G1/ S transition, since FACS analysis showed that PF cells have a higher ratio of cells in the S-phase compared to SD cells. Staurosporine (ST) has been recently reported to cause G1 arrest by affecting phos-
624
AID
BBRC 6346
/
6924$$$162
03-11-97 06:29:16
bbrcg
AP: BBRC
Vol. 232, No. 3, 1997
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
FIG. 5. DNA histogram of Gc-4 SD and PF cells. Nuclei of Gc-4 SD and PF cells were stained with propidium iodide and analysed by flow cytometry. The data shown below indicate the percentage of cells in the different phases of the cell cycle.
phorylation of retinoblastoma protein (pRB) and other molecules which lie upstream of cdk2 (13). ST as well as a more potent and selective inhibitor of cdk2, BL-1, did not inhibit cell growth in PF (5). From these accumulated data and notions, we hypothesize that a molecule(s) downstream of cdk2 might be aberrantly activated in PF cells. Alternatively, there may exist a growth-promoting pathway independent of cdk2 which is activated. However, we did not detect any significant difference in the expression or phosphorylation state of pRB, one of the substrates for cdk2, by immunoblotting analysis using several kinds of anti-pRB antibodies (data not shown). Probably, pRB is phosphorylated by cdk4 which is identical in the protein expression and kinase activity in both clones. In conclusion, the minimal level of cdk2 kinase activity observed in PF suggests to us the novel idea that in some cases activation of molecules downstream of cdk2 may cause abnormal cell proliferation, and possibly tumor progression, in the absence of significant cdk2 activity.
REFERENCES 1. 2. 3. 4. 5. 6. 7.
8. 9. 10. 11. 12. 13.
Ullrich, A., and Schlessinger, J. (1990) Cell 61, 203–212. Hunter, T., and Pines, J. (1994) Cell 79, 573–582. Sherr, C. J. (1994) Cell 79, 551–555. Watanabe, H., Arita, S., and Chigira, M. (1991) In Vitro Cell Develop. Biol. 27A, 267–269. Chigira, M., Noda, K., and Watanabe, H. (1993) Int. J. Oncol. 2, 67–72. Watanabe, H., Kanbe, K., Shinozaki, T., Hoshino, H., and Chigira, M. (1995) Int. J. Cancer 62, 191–198. Kitagawa, M., Okabe, T., Ogino, H., Matsumoto, H., Takahashi, I., Kokubo, T., Higashi, H., Saito, S., Taya, Y., Yasuda, H., Ohba, Y., Nishimura, S., Tanaka, N., and Okuyama, A. (1993) Oncogene 8, 2425–2432. Dobashi, Y., Kudoh, T., Matsumine, A., Toyoshima, K., and Akiyama, T. (1995) J. Biol. Chem. 270, 23031–23037. Xiong, Y., Zhang, H., and Beach, D. (1992) Cell 71, 505–514. Akiyama, T., Ohuchi, T., Sumida, S., Matsumoto, K., and Toyoshima, K. (1992) Proc. Natl. Acad. Sci. USA 89, 7900–79049. Matsushime, H., Quelle, D. E., Shurtleff, S. A., Shibuya, M., Sherr, C. J., and Kato, J. (1994) Mol. Cell. Biol. 14, 2066–2076. Nicoletti, I., Migliorati, G., Pagliacci, M. C., Grignani, F., and Riccardi, C. (1991) J. Immunol. Methods 139, 271–279. Schnier, J. B., Nishi, K., Goodrich, D. W., and Bradbury, E. M. (1996) Proc. Natl. Acad. Sci. USA 93, 5941–5946.
625
AID
BBRC 6346
/
6924$$$162
03-11-97 06:29:16
bbrcg
AP: BBRC