Portal de Periódicos da CAPES

Differential Myeloma Cell Responsiveness to Interferon-α Correlates with Differential Induction of p19INK4d and Cyclin D2 Expression

1998; Elsevier BV; Volume: 273; Issue: 19 Linguagem: Inglês

10.1074/jbc.273.19.11799

1083-351X

Taruna Arora, Diane F. Jelinek,

Multiple Myeloma Research and Treatments

Interferon-α (IFN-α) has been used as therapy for the treatment of a variety of viral diseases and malignancies including multiple myeloma. The effectiveness of interferon-α in treating multiple myeloma, however, has been somewhat variable, and the mechanism(s) accounting for this is not well understood. As a means to examine the basis for the differential effectiveness of this cytokine, we have analyzed IFN-α-mediated modulation of the cell cycle in two human myeloma cell lines. These two cell lines, ANBL-6 and KAS-6/1, display dramatically different outcomes in response to this cytokine. Although IFN-α inhibited the growth of ANBL-6 cells by blocking cell cycle progression from G0/G1 to S phase, IFN-α stimulated cell cycle progression in KAS-6/1 cells. Moreover, the effects of IFN-α on cell cycle progression correlated with the phosphorylation status of the retinoblastoma protein. Of interest, IFN-α increased cyclin D2 expression and cyclin-dependent kinase activity in the KAS-6/1 cells but not in the ANBL-6 cells. To determine whether the differential effects of IFN-α on myeloma cell cycle progression could also result from differences in the expression of cyclin-dependent kinase inhibitors, we examined the effects of IFN-α on the induction of cyclin-dependent kinase inhibitors with broad regulatory function (p21 and p27) and those with specificity for G1-associated cyclin-cyclin-dependent kinase complexes (p15, p16, p18, and p19). Although we failed to detect an effect of IFN-α on expression levels of p21, p15, p16, or p18, IFN-α treatment of the ANBL-6 cell line resulted in induction of p19 expression, whereas it was without effect on the KAS-6/1 cell line. These results suggest that heterogeneity in IFN-α-mediated growth effects in myeloma cells correlates with differential induction of cyclin D2 and p19INK4d expression. Interferon-α (IFN-α) has been used as therapy for the treatment of a variety of viral diseases and malignancies including multiple myeloma. The effectiveness of interferon-α in treating multiple myeloma, however, has been somewhat variable, and the mechanism(s) accounting for this is not well understood. As a means to examine the basis for the differential effectiveness of this cytokine, we have analyzed IFN-α-mediated modulation of the cell cycle in two human myeloma cell lines. These two cell lines, ANBL-6 and KAS-6/1, display dramatically different outcomes in response to this cytokine. Although IFN-α inhibited the growth of ANBL-6 cells by blocking cell cycle progression from G0/G1 to S phase, IFN-α stimulated cell cycle progression in KAS-6/1 cells. Moreover, the effects of IFN-α on cell cycle progression correlated with the phosphorylation status of the retinoblastoma protein. Of interest, IFN-α increased cyclin D2 expression and cyclin-dependent kinase activity in the KAS-6/1 cells but not in the ANBL-6 cells. To determine whether the differential effects of IFN-α on myeloma cell cycle progression could also result from differences in the expression of cyclin-dependent kinase inhibitors, we examined the effects of IFN-α on the induction of cyclin-dependent kinase inhibitors with broad regulatory function (p21 and p27) and those with specificity for G1-associated cyclin-cyclin-dependent kinase complexes (p15, p16, p18, and p19). Although we failed to detect an effect of IFN-α on expression levels of p21, p15, p16, or p18, IFN-α treatment of the ANBL-6 cell line resulted in induction of p19 expression, whereas it was without effect on the KAS-6/1 cell line. These results suggest that heterogeneity in IFN-α-mediated growth effects in myeloma cells correlates with differential induction of cyclin D2 and p19INK4d expression. Interferon-alpha (IFN-α) 1The abbreviations used are: IFN-α, interferon-α; CDK, cyclin-dependent kinase; CKI, CDK inhibitor; MM, multiple myeloma; PAGE, polyacrylamide gel electrophoresis; PCR, polymerase chain reaction; pRb, retinoblastoma protein; STAT, signal transducer and activator of transcription; IL-6, interleukin-6; mAb, monoclonal antibodies; DTT, dithiothreitol; bp, base pair(s); GST, glutathione S-transferase. is a natural therapeutic agent that has been used to treat a variety of viral diseases and malignancies, including multiple myeloma (MM), a B cell malignancy characterized by the clonal expansion of plasma cells (1Brenning G. Ahre A. Nilsson K. Scand. J. Haematol. 1985; 35: 543-549Crossref PubMed Scopus (52) Google Scholar, 2Oken M.M. Mayo Clin. Proc. 1994; 69: 781-786Abstract Full Text Full Text PDF PubMed Scopus (37) Google Scholar, 3Pestka S. Langer J. Zoon K.C. Samuel C. Annu. Rev. Biochem. 1987; 56: 727-777Crossref PubMed Scopus (1604) Google Scholar). Although IFN-α typically has an anti-proliferative effect on a wide variety of cell types, thereby providing the rationale for its clinical use, IFN-α treatment of MM patients has not been universally effective. Indeed, several reports have demonstrated that IFN-α therapy may actually result in disease aggravation and stimulation of plasma cell growth (4Blade J. Lopez-Guillermo A. Tassies D. Montserrat E. Rozman C. Br. J. Haematol. 1991; 79: 523-525Crossref PubMed Scopus (30) Google Scholar, 5Sawamura M. Murayama K. Ui G. Matsushima T. Tamura J. Murakami H. Naruse T. Tsuchiya J. Br. J. Haematol. 1992; 82: 631Crossref PubMed Scopus (19) Google Scholar). Understanding the mechanisms by which some myeloma cells are adverse to the cell cycle inhibitory effects of IFN-α is essential to permit the effective use of IFN-α in the treatment of MM. Cell cycle progression is regulated by proteins called cyclins and cyclin-dependent kinases (CDKs) that associate with each other in a cell cycle phase-specific manner (reviewed in Ref. 6Sherr C.J. Cell. 1993; 73: 1059-1065Abstract Full Text PDF PubMed Scopus (1993) Google Scholar). The process of cyclin-dependent activation of CDKs is counterbalanced by CDK inhibitors (CKIs). The first family of CKIs, referred as the p21cip1 family, includes at least three proteins, and each member of this family binds to several cyclin-CDK complexes (7Xiong Y. Hannon G.J. Zhang H. Casso D. Kobayashi R. Beach D. Nature. 1993; 366: 701-704Crossref PubMed Scopus (3179) Google Scholar). The second family of CKIs, referred to as the INK4 family, specifically binds to CDK4 and CDK6 and hence is specific for regulation of the G1 phase (8Hirai H. Roussel M.F. Kato J.-Y. Ashmun R.A. Sherr C.J. Mol. Cell. Biol. 1995; 15: 2672-2681Crossref PubMed Scopus (586) Google Scholar, 9Okuda T. Hirai H. Valentine V.A. Shurtleff S.A. Kidd V.J. Lahti J.M. Sherr C.J. Downing J.R. Genomics. 1995; 132: 309-315Google Scholar, 10Sherr C.J. Roberts J.M. Genes Dev. 1995; 9: 1149-1163Crossref PubMed Scopus (3220) Google Scholar). Numerous studies have demonstrated the importance of the INK4 family in overall cell growth control (11Hirama T. Koeffler H.P. Blood. 1995; 86: 841-854Crossref PubMed Google Scholar, 12Hunter T. Pines J. Cell. 1994; 79: 573-582Abstract Full Text PDF PubMed Scopus (2159) Google Scholar, 13Kamb A. Gruis N.A. Weaver-Feldhaus J. Liu Q. Harshman K. Tavitigian S.V. Stockert E. Day III, R.S. Johnson B.E. Skolnick M.H. Science. 1994; 264: 436-440Crossref PubMed Scopus (2819) Google Scholar, 14Nobori T. Miura K. Wu D.J. Lois A. Takabayashi K. Karson D.A. Nature. 1994; 368: 753-756Crossref PubMed Scopus (1722) Google Scholar, 15Ogawa S. Hirano N. Sato N. Takahashi T. Hangaishi A. Tanaka K. Kurokawa M. Tanaka T. Mitani K. Yazaki Y. Hirai H. Blood. 1994; 84: 2431-2435Crossref PubMed Google Scholar, 16Serrano M. Lee H. Chin L. Cordon-Cardo C. Beach D. DePinho R.A. Cell. 1996; 85: 27-37Abstract Full Text Full Text PDF PubMed Scopus (1413) Google Scholar). The CDK inhibitors function by inhibiting CDK-mediated phosphorylation of the retinoblastoma protein (pRb), a known tumor suppressor (17Kato J.H. Hiebert S.W. Ewen M.E. Sherr C.J. Genes Dev. 1993; 7: 331-342Crossref PubMed Scopus (1093) Google Scholar, 18Bartek J. Bartkowa J. Lukas J. Curr. Opin. Cell Biol. 1996; 8: 805-814Crossref PubMed Scopus (368) Google Scholar). Thus, during G1 phase, phosphorylation of pRb is regulated by the periodic expression of INK4 CKIs and cyclin Ds, which determines whether active cyclin D-CDK or inactive inhibitor-CDK complexes will be assembled. Growth arrest of hematopoietic cell lines by IFN-α has been shown to correlate with inhibition of pRb phosphorylation (19Kumar R. Atlas I. Proc. Natl. Acad. Sci. U. S. A. 1992; 89: 6599-6603Crossref PubMed Scopus (92) Google Scholar,20Resnitzsky D. Tiefenbrun N. Berissi H. Kimchi A. Proc. Natl. Acad. Sci. U. S. A. 1992; 89: 402-406Crossref PubMed Scopus (116) Google Scholar). Considerable effort has been directed at the analysis of IFN-α signaling; however, the role of CKIs in mediating inhibition of pRb hyperphosphorylation and cell cycle progression by IFN-α remains to be elucidated. Although p27kip1 has been shown to be induced following IFN-α treatment in some cell lines, its kinetics of expression are not consistent with a role for this inhibitor in the regulation of early G1 phase arrest (21Yamada H. Ochi K. Nakada S. Takahara S. Nemato T. Sekikawa T. Horiguchi-Yamada J. J. Mol. Biochem. 1995; 152: 149-158Crossref Scopus (25) Google Scholar). Induction of p21Cip1 expression has also been suggested to underlie IFN-α-mediated growth arrest of Daudi cells, and this occurred in a pRb-dependent manner (22Sangfelt O. Erickson S. Einhorn S. Grander D. Oncogene. 1997; 14: 415-423Crossref PubMed Scopus (101) Google Scholar, 23Subramaniam P.S. Johnson H.W. J. Interferon Cytokine Res. 1997; 17: 11-15Crossref PubMed Scopus (45) Google Scholar). However, p21 expression was also induced in IFN-α growth-arrested pRb−/− U-266 cells and in IFN-α-resistant H9 cells (22Sangfelt O. Erickson S. Einhorn S. Grander D. Oncogene. 1997; 14: 415-423Crossref PubMed Scopus (101) Google Scholar). In contrast, Yamada et al.(21Yamada H. Ochi K. Nakada S. Takahara S. Nemato T. Sekikawa T. Horiguchi-Yamada J. J. Mol. Biochem. 1995; 152: 149-158Crossref Scopus (25) Google Scholar) did not observe p21 induction in Daudi cells by IFN-α treatment. Thus, it remains unclear whether the anti-proliferative effects of IFN-α are indeed mediated by p21 or whether IFN-α acts instead via induction of other cell cycle regulatory protein(s). In addition, the role of the INK4 family members has not been clearly defined in IFN-α-mediated cell cycle regulation. Although IFN-α treatment of the U-266 cell line was reported to increase p15 and p16 mRNA expression, these genes were not induced in IFN-α growth-arrested Daudi cells (22Sangfelt O. Erickson S. Einhorn S. Grander D. Oncogene. 1997; 14: 415-423Crossref PubMed Scopus (101) Google Scholar). As the U-266 cell line is pRb−/−, the role of these CKIs in mediating the anti-proliferative effects of IFN-α is not understood. Finally, in the same study, IFN-α had no effect on the other two members of the INK4 family, p18 and p19 (22Sangfelt O. Erickson S. Einhorn S. Grander D. Oncogene. 1997; 14: 415-423Crossref PubMed Scopus (101) Google Scholar). We have previously described the establishment of a panel of interleukin-6 (IL-6) responsive human myeloma cell lines and have shown that one of the cell lines, KAS-6/1, is uniquely growth stimulated by IFN-α (24Jelinek D.F. Aagaard-Tillery K.M. Arendt B.K. Arora T. Tschumper R.C. Westendorf J.J. J. Clin. Invest. 1997; 99: 447-456Crossref PubMed Scopus (44) Google Scholar). We have also shown that the differential growth responsiveness of MM cell lines to IFN-α does not seem to result from differential activation of STAT factors (24Jelinek D.F. Aagaard-Tillery K.M. Arendt B.K. Arora T. Tschumper R.C. Westendorf J.J. J. Clin. Invest. 1997; 99: 447-456Crossref PubMed Scopus (44) Google Scholar). To gain more insight into the differential growth regulation of myeloma cells by IFN-α, in this study we have focused on the analysis of the molecular events regulating cell cycle progression. The myeloma cell lines ANBL-6 and KAS-6/1 were derived in our laboratory and are maintained with IL-6 (kindly provided by Immunex Corp., Seattle, WA) as described previously (25Jelinek D.F. Ahmann G.J. Greipp P.R. Jalal S.M. Westendorf J.J. Katzmann J.A. Kyle R.A. Lust J.A. Cancer Res. 1993; 53: 5320-5327PubMed Google Scholar, 26Westendorf J.J. Ahmann G.J. Greipp P.R. Witzig T.E. Kyle R.A. Lust J.A. Jelinek D.F. Leukemia. 1996; 10: 866-876PubMed Google Scholar). Recombinant purified IFN-α2b was purchased from Schering Corp. and was used at a concentration of 1000 units/ml. Monoclonal antibodies (mAb) to pRb and CDK4 as well as polyclonal Abs to cyclin D were purchased from Upstate Biotechnology Inc. Rabbit polyclonal antisera against the carboxyl-terminal region of human p33CDK2 was a generous gift of Dr. R. T. Abraham (Mayo Clinic, MN). GST-Rb fusion protein, mouse mAb to p21, and rabbit polyclonal Abs to p15, p16, p18, and p19 were purchased from Santa Cruz Biotechnology, Inc. Before analyzing the effects of IFN-α on cell cycle progression, cells were starved of IL-6 for 48 h in medium supplemented with 0.5% bovine serum albumin. Cells were then washed, recultured in media supplemented with bovine serum albumin at a density of 1 × 106 per ml, and stimulated with 1 ng/ml IL-6 or 500 units/ml IFN-α. Cell cycle analysis was subsequently performed using propidium iodide staining as described previously (27Jelinek D.F. Witzig T.E. Arendt B.K. J. Immunol. 1997; 159: 487-496PubMed Google Scholar). As described above, myeloma cells were cultured in IL-6 free media for 48 h prior to addition of the indicated cytokines. Cells (0.5–1.0 × 107) were then lysed in lysis buffer containing 50 mm Tris-HCl (pH 8.0), 100 mm NaF, 10% glycerol, 0.1 mm EDTA, 200 mm NaCl, 0.5% Nonidet P-40, 1 mm Na3VO4, 10 μg/ml leupeptin, 5 μg/ml pepstatin, 1 mm DTT, 0.2 mm phenylmethylsulfonyl fluoride, and 5 μg/ml aprotinin for 30 min on ice. For direct immunoblot analysis, cell lysates containing 50 μg of total protein were resolved by electrophoresis through 7.5% SDS-polyacrylamide gels and then transferred onto Immobilon-P membranes (Millipore Corp. For immunoprecipitation, cell lysates were first precleared with protein A-Sepharose (Pharmacia, Uppsala, Sweden) and then immunoprecipitated with the indicated Abs at 4 °C for 2 h. The immune complexes were collected by adsorption to protein A-Sepharose beads by overnight agitation at 4 °C. The immunoprecipitates were washed twice with lysis buffer and once with cold phosphate-buffered saline. The complexes were then boiled in 30 μl of 2 × SDS sample buffer and resolved on an SDS-PAGE gel followed by transfer to Immobilon-P membranes. The membranes were blocked (25 mm Tris-Cl (pH 7.2), 150 mm NaCl, 0.05% Tween 20, 2% bovine serum albumin) and subsequently probed with the indicated Abs. For reprobing the same blot, the membranes were stripped with 7 m guanidine, renatured, and blocked before immunoblotting with the indicated Abs. Immunoreactive proteins were detected using an enhanced chemiluminescence detection system (Amersham). Cells were lysed on ice in H1 kinase lysis buffer (50 mm Tris (pH 7.4), 150 mm NaCl, 0.5% Triton X-100, 5 mm EDTA, 5 mm NaF, 0.1 mm Na3VO4, 1 mm DTT, and protease inhibitors) and then immunoprecipitated with CDK2 antisera and protein A-Sepharose for 1 h at 4 °C. Immunoprecipitates were washed twice with lysis buffer followed by two washes with kinase buffer (20 mm HEPES, pH 7.0, 10 mm β-glycerophosphate, 5 mmMgCl2, 1 mm DTT, and 10 μg/ml leupeptin). Kinase activity of the immunoprecipitated complex was assessed using histone H1 as a substrate in a reaction buffer consisting of 0.6 mg/ml histone H1 (Sigma) and 5 μCi [γ-32P]ATP/reaction and 5 μm cold ATP. Following a 10-min incubation at 30 °C, the reaction was stopped by addition of ice-cold 20 mmEDTA, pH 8.0. Duplicate aliquots of each sample supernatant were spotted onto phosphocellulose paper. The papers were immersed briefly in 1% H3PO4 solution containing 10 mm Na4P2O7. After three additional 15-min washes in 1% H3PO4, radioactivity incorporated in filter-bound histone was quantitated by liquid scintillation counting. Cells were lysed on ice in lysis buffer (50 mm HEPES, pH 7.4, 150 mm NaCl, 1 mm EDTA, 2.5 mm EGTA, 0.1% Tween 20, 10% glycerol, 1 mm DTT, 10 μg/ml leupeptin, 5 μg/ml pepstatin, 10 mm β-glycerophosphate, 0.1 mmsodium orthovanadate) and immunoprecipitated with CDK4 Abs (UBI) and protein A-Sepharose for 2 h. Immunoprecipitates were washed twice with lysis buffer and twice with kinase buffer (50 mmHEPES, pH 7.5, 1 mm NaF, 2.5 mm EGTA, 1 mm DTT, 10 mm β-glycerophosphate, 0.1 mm Na3VO4, 10 mmMgCl2). The kinase reaction was carried out for 30 min at 30 °C in a reaction buffer containing 1 μg/ml GST-pRb, 10 μCi [γ-32P]ATP/reaction, and 20 μm cold ATP. SDS-PAGE sample buffer was added, and proteins were eluted by boiling and resolved by 10% SDS-PAGE. The phosphorylated substrate was detected by autoradiograph followed by immunoblotting for CDK4. Total RNA was isolated using the TRIZOL reagent (Life Technologies, Inc.) and reverse transcribed using a first strand cDNA synthesis kit (Pharmacia). PCR reactions contained 1.5 mmMgCl2, 1 mm dNTPs, 1 μm each of sense (S) and antisense (AS) oligonucleotides, and 2.5 unitsTaq DNA polymerase (Promega). The primer sequences and their expected sizes are as follows: β-actin, S-5′-gacttcgagcaagagatggccac-3′ and AS-5′-caatgccagggtatggtggtg-3′ (265 bp); p19, S-5′-ctcacacccttggagctgg-3′ and AS-gcagcctgaggcgcagaag-3′ (354 bp); p18, S-5′-ctgcaggttatgaaacttgg-3′ and AS-5′-ttattgaagatttgtggctcc-3′ (382 bp) (28Miller C.W. Aslo A. Campbell M.J. Kawamata N. Lampkin B.C. Koeffler H.P. Cancer Genet. Cytogenet. 1996; 86: 136-142Abstract Full Text PDF PubMed Scopus (103) Google Scholar); p15, S-5′-ctgcgcgtctgggggctgc-3′ and AS-5′-cctcccgaaacggttgactc-3′ (165 bp) (28Miller C.W. Aslo A. Campbell M.J. Kawamata N. Lampkin B.C. Koeffler H.P. Cancer Genet. Cytogenet. 1996; 86: 136-142Abstract Full Text PDF PubMed Scopus (103) Google Scholar), p16, S-5′-gccactctcacccgacc-3′ and AS-5′-ctacgaaagcggggtgg-3′ (351 bp). PCR products were visualized by electrophoresis on 2% agarose gels and staining with ethidium bromide after 30 cycles (denaturation at 95 °C for 1 min, annealing at 60 °C for 2 min, and extension at 72 °C for 3 min). In experiments requiring visualization of bands, computer-generated images were obtained by scanning the original autoradiograms or digitalized images of ethidium bromide stained bands. In all cases, the images are representative of the original data. The experiments undertaken in this study were designed to analyze IFN-α-regulated cell cycle control in myeloma cells. The first experiments in this study therefore analyzed the effects of IFN-α-stimulated cell cycle progression following serum and cytokine starvation. As expected, IL-6 stimulated cell cycle progression by both cell lines after 24 and 48 h as revealed by the increased numbers of cells with a DNA content suggestive of cells in S + G2/M phases of the cell cycle (Fig. 1,A and B; Table I). Addition of IFN-α to the ANBL-6 cells failed to stimulate cell cycle progression at any of the time points beyond that observed in control cultures. In striking contrast, KAS-6/1 cells progressed into both the S and the G2/M phases of the cell cycle in response to IFN-α with increasing lengths of stimulation (Fig. 1 B). Moreover, IFN-α did not inhibit IL-6-stimulated KAS-6/1 cell entry into either S or G2/M phases. These data suggest that IFN-α-mediated growth control of myeloma cells is coupled to regulation of cell cycle progression.Table IIFN-α has contrasting effects on myeloma cell cycle progressionCell lineIncubation length (h)Cell cycle phaseStimulusNilIL-6IFN-αIL-6+IFN-α% of cellsANBL-624G0/ G179.160.780.283.5S11.734.58.39.8G2 /M9.14.711.46.648G0 /G180.868.587.885.9S12.410.77.96.4G2 /M6.820.64.17.5KAS-6/124G0/G181.644.045.447.1S8.955.153.652.0G2 /M9.50.91.00.848G0 /G183.540.050.045.9S6.630.914.026.6G2 /M9.829.035.327.3 Open table in a new tab We next examined the possibility that the differential effects of IFN-α on the growth of these cell lines resulted from differential phosphorylation of pRb. As shown in Fig.2 A, IL-6 induced the hyperphosphorylation of pRb in both cell lines as revealed by the reduced rate of pRb electrophoretic migration. When the effects of IFN-α on the ANBL-6 cell line were examined, it was first noted that a 12-h stimulation with IFN-α failed to inactivate pRb by hyperphosphorylation in this cell line and second that IFN-α profoundly inhibited IL-6-stimulated hyperphosphorylation of pRb. By contrast, the hyperphosphorylated form of pRb predominated in IFN-α-stimulated KAS-6/1 cells. The level of pRb hyperphosphorylation stimulated by the combined addition of IL-6 and IFN-α was comparable with that observed in cells stimulated with IL-6 alone (results not shown). It is likely that alterations in the expression of cyclin D or CDK2 and CDK4 might contribute to either the induction or inhibition of pRb phosphorylation by IFN-α. To test this possibility, we next analyzed expression of cyclin D and CDKs in cells cultured with IL-6 or IFN-α for 6 h. As may be seen in Fig. 2 B, IL-6 stimulated the increased expression of cyclin D in both cell lines. Densitometric scanning of multiple experiments revealed that IL-6 increased cyclin D expression by 6.0 ± 0.8-fold in the ANBL-6 cells (n = 2) and by 9.4 ± 0.5-fold in the KAS-6/1 cells (n = 3). By contrast, IFN-α treatment of the ANBL-6 cell line resulted in markedly reduced levels of cyclin D expression (0.2 ± 0.1 relative to control cell levels,n = 2). Furthermore, the IL-6-induced increase in cyclin D expression was antagonized by coculturing ANBL-6 cells with IFN-α (0.6 ± 0.1 relative to control cell levels,n = 2). IFN-α stimulation of the KAS-6/1 cells, however, resulted in an increase in cyclin D expression (6.7 ± 1.3-fold induction over unstimulated cells, n = 3). When the blot was reprobed for the expression of CDK4, no apparent modulation in expression was observed in control or IFN-α-stimulated cells (Fig. 2 C). Similar results were obtained when CDK2 levels were quantitated (results not shown). To characterize the induction of cyclin D expression in greater detail, we next used a polyclonal antisera to immunoprecipitate cyclin D from the KAS-6/1 cells following cytokine stimulation (Fig.3). To determine the identity of the immunoprecipitated cyclin D, we compared its electrophoretic migration with cyclin D immunoprecipitates prepared from control cell lines known to express cyclin D1, D2, and D3 (U2OS) or cyclin D1 and D3 (MCF-7). As may be seen in Fig. 3, both IL-6 and IFN-α were observed to rapidly (6 h) induce cyclin D expression in the KAS-6/1 cell line, and the migration position suggests that these cells primarily express cyclin D2. When cyclin D2 levels were quantitated after 12 h of cytokine stimulation by densitometry, IL-6 caused a 5.1 ± 0.1-fold induction, whereas IFN-α caused a 4.8 ± 0.1-fold induction of cyclin D2 expression (n = 2). When KAS-6/1 cells were stimulated with both cytokines, a 7.9 ± 1.1-fold induction was observed. The ANBL-6 cells were similarly observed to express cyclin D2 upon IL-6 but not IFN-α stimulation (data not shown). pRb phosphorylation during early and late G1 phase is coordinated by the activation of CDK4 and CDK2, respectively (17Kato J.H. Hiebert S.W. Ewen M.E. Sherr C.J. Genes Dev. 1993; 7: 331-342Crossref PubMed Scopus (1093) Google Scholar). To confirm the role of CDK4 and CDK2 in the time course of pRb phosphorylation, we next assayed kinase activity in CDK4 and CDK2 immunoprecipitates from ANBL-6 and KAS-6/1 cell lysates (Fig.4). At 8 h poststimulation, IL-6 stimulated CDK4 activity in ANBL-6 cells as indicated by a 4.4 ± 0.1-fold (n = 2) increase in phosphorylation of a GST-Rb fusion protein. Although IFN-α treatment of the ANBL-6 cells did not alter CDK4 activity over that observed in unstimulated cells (0.7 ± 0.2 relative to control cell levels, n = 2), it suppressed IL-6-induced CDK4 activation to the level of activity observed in unstimulated cells. In contrast, CDK4 kinase activity was increased 4.1 ± 0.2-fold over control cells (n = 2) in response to IFN-α as well as IL-6 (7.0 ± 1.6-fold increase over control cells, n = 2) in KAS-6/1 cells. When CDK2 activity was assayed in the ANBL-6 cells (Fig.5), a 24-h IL-6 stimulation increased CDK2 activity by 8–9-fold over that of unstimulated cells. Although IFN-α itself did not affect basal CDK2 activity, it inhibited the IL-6-mediated increase in CDK2 activity. In contrast, both IL-6 and IFN-α induced a significant increase in CDK2 activity in the KAS-6/1 cell line when assayed after 24 h. Unlike the rapid alterations in the expression of cyclin D, modulations in CDK2 activity were observed only after the effects on cell cycle progression and pRb phosphorylation became obvious.Figure 5Analysis of CDK2 kinase activity by histone H1 kinase assay in ANBL-6 and KAS-6/1 cells. II-6 (solid bars), IFN-α- (open bars), and IL-6 + IFN-α (stippled bars)-stimulated cells were lysed and immunoprecipitated with anti-CDK2 Abs. Immunoprecipitates were assayed for H1 kinase, and the data are displayed as the -fold increase in H1 phosphorylation over that observed in unstimulated cells. Results are representative of multiple experiments.View Large Image Figure ViewerDownload (PPT) Because IFN-γ has been shown to induce p21 expression (29Chin Y.E. Kitagawa M. Su W.-C.S. Yu Z.-H. Iwamoto Y. Fu X.-Y. Science. 1996; 272: 719-722Crossref PubMed Scopus (732) Google Scholar), we next asked whether IFN-α stimulation could also affect p21 expression in myeloma cells. As shown in Fig. 6, neither IL-6 nor IFN-α significantly affected p21 expression levels in either the ANBL-6 cells or KAS-6/1 cells. Similarly, p27kip1 expression was not modulated by IL-6 or IFN-α in either cell line (data not shown). Thus, expression levels of p21Cip1 or p27kip1 do not correlate with pRb phosphorylation status. Because IFN-α treatment of myeloma cell lines affected CDK4 activity in direct correlation with pRb phosphorylation status, we next evaluated the ability of IFN-α to induce expression of p15, p16, p18, and p19 mRNA. As shown in Fig. 7, neither IL-6 nor IFN-α altered p15, p16, or p18 mRNA levels in either cell line as assessed by reverse transcriptase-PCR. Densitometric scanning of three independent experiments carried out on three independent preparations of RNA confirmed the lack of effect of these cytokines on p15, p16, and p18 expression. Indeed, the levels of expression in both cell lines were found to be 1.0 ± 0.1 relative to control levels, regardless of the length of stimulation or whether the cells were stimulated with IL-6 or IFN-α. By contrast, IFN-α treatment of ANBL-6 cells resulted in a significant induction of p19 mRNA over the basal levels as early as 2 h poststimulation. These levels remained high with continued IFN-α stimulation. Densitometric scanning of multiple experiments revealed that IFN-α caused -fold increases in p19 expression in the ANBL-6 cells of 7.2 ± 0.4, 8.2 ± 1.4, 9.5 ± 1.7, and 9.2 ± 1.9 following stimulation for 2, 4, 6, and 12 h, respectively (n = three independent preparations of RNA). IFN-α treatment of the KAS-6/1 cell line, however, did not affect basal mRNA expression of p19 in KAS-6/1 cells over the entire time period of 12 h (-fold increases of 1.0 ± 0.3, 0.8 ± 0.1, 1.0 ± 0.1, and 0.8 ± 0.1 after 2, 4, 6, and 12 h, respectively; n = 2). IFN-α stimulation also increased p19 protein expression in ANBL-6 cells, whereas IL-6 stimulation had no effect on p19 expression (Fig.8). Thus, analysis of multiple experiments indicated the ability of IFN-α to increase p19 protein levels by 4.4 ± 0.5, 6.8 ± 1.6, and 8.5 ± 1.7-fold after stimulation for 2, 4, and 6 h, respectively. Similarly, as observed for p19 mRNA, IL-6 or IFN-α did not affect p19 protein expression in KAS-6/1 cells. Rather, IFN-α slightly decreased p19 levels (0.7 ± 0.1 relative to control cell levels after 4 h,n = 3). In contrast with the effects of IFN-α on p19 protein expression, p15 levels were not altered by either cytokine in either cell line (Fig. 8). In similar studies, p18 protein levels were similarly unaffected, whereas p16 protein expression could not be detected (data not shown).Figure 8Protein expression of p15 and p19 following IFN-α treatment in ANBL-6 and KAS-6/1. Cytokine-stimulated cells were lysed, resolved by 15% SDS-PAGE, and immunoblotted with anti-p15 or anti-p19 Abs. Results are representative of three experiments.View Large Image Figure ViewerDownload (PPT) The mechanisms that underlie the variable response of patient myeloma cells to IFN-α treatment have remained elusive primarily as a result of a lack of a suitable model system that is representative ofin vivo heterogeneity. Our panel of myeloma cell lines, however, circumvents this difficulty and provides a physiologically relevant model system in which to study the inherent heterogeneity in IFN-α responsiveness. Moreover, the KAS-6/1 cell line, which uniquely proliferates in response to IFN-α, provides us with a tool to understand “nonclassical” IFN-α responses in a human cell line. Our previous results showed that despite the contrasting effects of IFN-α on the growth of the myeloma cell lines ANBL-6 and KAS-6/1, no differential activation of STAT factors was observed nor was any autocrine IL-6 expression induced in the cells that were growth stimulated by IFN-α (24Jelinek D.F. Aagaard-Tillery K.M. Arendt B.K. Arora T. Tschumper R.C. Westendorf J.J. J. Clin. Invest. 1997; 99: 447-456Crossref PubMed Scopus (44) Google Scholar). These results prompted us to examine other mechanisms that may account for the differential biological responsiveness to IFN-α. The studies described herein demonstrate that one of the key mechanisms underlying this heterogeneity may involve regulation of genes that encode for cell cycle regulatory molecules. Our results show that IFN-α stimulates the differential induction of cyclin D2 and p19 INK4d. Our analysis of cell cycle regulatory molecules was prompted by a series of observations. We have demonstrated that IFN-α-stimulated proliferation was accompanied by hyperphosphorylation of pRb and that this did not occur in the growth-inhibited cell line, ANBL-6. Moreover, IFN-α was capable of antagonizing IL-6-stimulated pRb phosphorylation in the same cell line. The finding that IFN-α inhibits IL-6-stimulated pRb phosphorylation in the ANBL-6 cells and directly stimulates pRb hyperphosphorylation in the KAS-6/1 cells is interesting, as these events precede the appearance of alterations in cell cycle progression and thymidine incorporation. This suggested that modulation of the phosphorylation status of pRb may be a key event in determining whether IFN-α stimulates growth or is growth inhibitory. Our results are consistent with other reports in which sensitivity to growth inhibition by IFN-α correlated with suppression of pRb hyperphosphorylation (19Kumar R. Atlas I. Proc. Natl. Acad. Sci. U. S. A. 1992; 89: 6599-6603Crossref PubMed Scopus (92) Google Scholar, 20Resnitzsky D. Tiefenbrun N. Berissi H. Kimchi A. Proc. Natl. Acad. Sci. U. S. A. 1992; 89: 402-406Crossref PubMed Scopus (116) Google Scholar). However, the KAS-6/1 cells differ from cells resistant to the inhibitory effects of IFN-α in that these cells are growth responsive to IFN-α. IFN-α stimulates pRb inactivation, which may allow KAS-6/1 cells to pass this critical G1 to S phase checkpoint. One striking observation in our study was the ability of IFN-α to up-regulate cyclin D2 protein expression in the growth-responsive KAS-6/1 cells but not in the growth-inhibited ANBL-6 cells. In results not shown, we have also demonstrated that IFN-α up-regulated cyclin D2 mRNA in the KAS-6/1 cells but not in the ANBL-6 cells. The cyclin D2 promoter contains potential binding sites for transcription factors of the AP1 family, CTF/NFY, MyB, and C/EBP (30Brooks A.R. Shiffman D. Chan C.S. Brooks E.E. Milner P.G. J. Biol. Chem. 1996; 271: 9090-9099Abstract Full Text Full Text PDF PubMed Scopus (102) Google Scholar). Therefore, it is possible that IFN-α activates one or more of these transcription factors in the KAS-6/1 cells via a mitogenic signaling pathway. We are currently investigating this possibility. We have also shown that IFN-α stimulates an increase in CDK4 and CDK2 activity, which correlated with G1 to S phase progression of the KAS-6/1 cells. Because the activation of CDKs requires association with cyclins, cyclin D2 may therefore play an important role in the regulation of G1 progression by activating CDK4. Another means of cell cycle regulation is provided by the Cip1/Kip1 and INK4 family CKIs. p21Cip1 has been shown to be involved in IFN-γ-mediated growth arrest via STAT1-dependent transcription of p21 waf1/cip1 (29Chin Y.E. Kitagawa M. Su W.-C.S. Yu Z.-H. Iwamoto Y. Fu X.-Y. Science. 1996; 272: 719-722Crossref PubMed Scopus (732) Google Scholar). This suggested the possibility that the differential effects of IFN-α on myeloma cell proliferation may result from differential regulation of p21. However, we did not observe any difference in the expression of p21 with IFN-α stimulation. As we have previously observed STAT1 activation in both cell lines following IFN-α stimulation (24Jelinek D.F. Aagaard-Tillery K.M. Arendt B.K. Arora T. Tschumper R.C. Westendorf J.J. J. Clin. Invest. 1997; 99: 447-456Crossref PubMed Scopus (44) Google Scholar), our observations are consistent with the notion that p21 transcription following STAT activation may not differ between the ANBL-6 and KAS-6/1 cell lines. Our results also do not support a role for p27kip1 in initiating growth arrest in response to IFN. Our analysis of INK4 family expression levels did not show any alterations in p15, p16, and p18 mRNA and protein expression in myeloma cells in response to IFN-α. Although p18 has been shown to be a potential CKI in terminal differentiation of myoblasts (31Franklin D.S. Xiong Y. Mol. Cell. Biol. 1996; 7: 1587-1599Crossref Scopus (105) Google Scholar) or EBV-transformed B cells (32Morse L. Chen D. Franklin D.S. Xiong Y. Chen-Kiang S. Immunity. 1997; 6: 47-56Abstract Full Text Full Text PDF PubMed Scopus (134) Google Scholar), it does not seem to function as a primary CKI in the growth arrest of undifferentiated cycling cells. Our data support these observations as myeloma cells differ from normal plasma cells because their terminal differentiation is deregulated. Of interest, p19 expression was induced at both RNA and protein levels by IFN-α in the ANBL-6 cells by as early as 2 h poststimulation, whereas in the IFN-α growth-responsive KAS-6/1 cells, p19 levels remained unaltered following stimulation by IFN-α. These data suggest that IFN-α may regulate growth of myeloma cell lines by inducing transcription of p19 as an early response gene prior to the G1 arrest. As IFN-α also induced p19 expression in the presence of IL-6 in the ANBL-6 cells (data not shown), it suggests that one of the mechanisms by which IFN-α may inhibit IL-6-dependent proliferation is by inducing p19 expression. Recent work by Chan et al. (33Chan F.K.M. Zhang J. Chen L. Shapiro D.N. Winoto A. Mol. Cell. Biol. 1995; 15: 2682-2688Crossref PubMed Scopus (336) Google Scholar) demonstrated the ability of p19 to interact with the Nur77 orphan steroid receptor using a yeast two-hybrid system. Although an in vivo interaction between Nur77 and p19 has not been formally demonstrated, these data raise the possibility that p19 might directly bind to targets other than cyclin D-dependent kinases. It is conceivable, therefore, that the basis for p19 specificity resides within its ability to simultaneously contact secondary targets such as transcription factors. Thus, p19 induction in response to IFN-α may lead to functional modulation of a transcription factor besides pRb. The mechanisms underlying the induction of p19 expression in the ANBL-6 cells by IFN-α, as well as the mechanisms underlying the failure of IFN-α to induce p19 in the KAS-6/1 cells, are currently under investigation. In summary, we have demonstrated that IFN-α-stimulated myeloma cell proliferation or growth arrest correlates with induction of cyclin D2 or p19, or lack thereof. Because of the important role that both of these proteins play in up- or down-regulation of the cell cycle, these results are highly suggestive that these molecules play a direct role in determining cellular outcome in response to IFN-α. Although these results are very interesting, it is important to acknowledge that our studies have not yet identified the signals that lie upstream of these effects, which presumably lie in close proximity to the IFN-α receptor. In this regard, a serine/threonine kinase PKR has been shown to down-regulate c-Myc expression in growth-arrested cells (34Raveh T. Hovanessian A.G. Meurs E.F. Sonenberg N. Kimchi A. J. Biol. Chem. 1996; 271: 25479-25484Abstract Full Text Full Text PDF PubMed Scopus (59) Google Scholar). However, we have not observed alterations in the expression of c-Myc in growth inhibitory or proliferative responses of myeloma cells to IFN-α (35Jelinek D.F. Arora T. Curr. Top. Microbiol. Immunol. 1997; 224: 261-268PubMed Google Scholar). It will be interesting to determine whether PKR or additional signaling pathways may control induction of p19 or cyclin D2. Finally, it remains possible that genetic variations in myeloma cell lines may lead to IFN-α-dependent activation of different transcription factors, which ultimately results in positive or negative regulation of p19 and cyclin D2 and hence in differential growth control. We thank Drs. Robert T. Abraham and Gregory J. Brunn for the generous gift of CDK2 antisera and helpful discussions. We also acknowledge the technical expertise of Renee C. Tschumper and Bonnie K. Arendt.