Differential Requirements for Interleukin-2 Distinguish the Expression and Activity of the Cyclin-dependent Kinases Cdk4 and Cdk2 in Human T Cells*

We examined the expression and activity of Cdk4 and Cdk2 in resting, competent, and proliferating normal human T cells. Expression of Cdk4 but not of Cdk2 was induced in competent T cells independent of an IL-2 signal. This up-regulation of Cdk4 mRNA and protein was resistant to the immunosuppressant drugs cyclosporinA (CsA) and FK506. A further increase in Cdk4 expression was seen upon stimulation of competent T cells by IL-2, as was de nouo expression of Cdk2. Cyclin D2, a Cdk4 partner, showed similar patterns of regulation as Cdk4. The increases in Cdk4 and cyclin D2 expression seen in competent T cells were functionally significant since Cdk4 immunoprecipitates from these cells phosphoryl- ated recombinant RB protein in vitro. Despite the lack of an increase in the expression of Cdk2, a small pool of pre-existing Cdk2 protein detected in resting T cells could be activated upon induction of competence. These data demonstrate that 1) the signals that lead to induc- tion of competence in T cells stimulate an IL-2-inde-pendent and CsA-resistant phase of Cdk4 and cyclin D2 expression, Cdk4 kinase activity, and Cdk2 kinase activ- ity, and 2) IL-2 stimulates a second phase of Cdk4 and cyclin D2 expression human T cells.

Differential Requirements for Interleukin-2 Distinguish the Expression and Activity of the Cyclin-dependent Kinases Cdk4 and Cdk2 in Human T Cells* (Received for publication, May 24, 1994, and in revised form, October 12, 1994) Jaime F. ModianoS We examined the expression and activity of Cdk4 and Cdk2 in resting, competent, and proliferating normal human T cells. Expression of Cdk4 but not of Cdk2 was induced in competent T cells independent of an IL-2 signal. This up-regulation of Cdk4 mRNA and protein was resistant to the immunosuppressant drugs cyclosporinA (CsA) and FK506. A further increase in Cdk4 expression was seen upon stimulation of competent T cells by IL-2, as was de nouo expression of Cdk2. Cyclin D2, a Cdk4 partner, showed similar patterns of regulation as Cdk4. The increases in Cdk4 and cyclin D2 expression seen in competent T cells were functionally significant since Cdk4 immunoprecipitates from these cells phosphorylated recombinant RB protein in vitro. Despite the lack of an increase in the expression of Cdk2, a small pool of pre-existing Cdk2 protein detected in resting T cells could be activated upon induction of competence. These data demonstrate that 1) the signals that lead to induction of competence in T cells stimulate an IL-2-independent and CsA-resistant phase of Cdk4 and cyclin D2 expression, Cdk4 kinase activity, and Cdk2 kinase activity, and 2) IL-2 stimulates a second phase of Cdk4 and cyclin D2 expression and de nouo expression of Cdk2 in these cells. The data show that the expression and activity of these major cell cycle regulatory proteins are controlled differentially by growth factors and indicate a role for Cdk4 and cyclin D2 in T-cell cycle entry and/or early G, progression and for Cdk2 in later G, progression and G,/S transition. Activation of T cells by antigen or mitogens stimulates a series of nonlinear biochemical cascades that induce entry into the cell cycle. The earliest signal transduction events that follow T-cell activation have been extensively studied (1)(2)(3)(4)(5); however, those that regulate transit through early G, phase and, ultimately, cell-cycle progression are less well understood. Our aim was to delineate those regulatory events which are associated with cell cycle entry and early GI phase versus those that are necessary only for later cell cycle progression.
Two families of proteins, the cyclins and cyclin-dependent * This work was supported in part by National Institutes of Health Grants POI-HL-36577 and AI-26490 (to E. W. G.) and by American Cancer Society Grant IM-746 (to J. J. L.). The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. kinases (Cdk),' play a major role in cell cycle progression in eukaryotic cells (6)(7)(8)(9)(10). The expression of some of these proteins is controlled by growth factor-derived signals in hematopoietic or fibroblastoid cell lines (11,12). We and others have shown differential time courses of expression of Cdk and cyclin proteins in T cells during cell cycle progression ' (13, 14). This suggests that such proteins may respond to different environmental signals, and each may have specific regulatory functions at different stages of the mammalian cell cycle. Despite the characterization of the kinetics of expression of the Cdks and cyclins, little is known about the signaling mechanisms that control their expression and function. Our data show that induction of competence in normal T cells (by the transient activation of protein kinase C and increases in intracellular calcium concentrations) stimulates an IL-2-independent, cyclosporin A (CsA)-resistant phase of Cdk4 and cyclin D2 expression as well as Cdk4 kinase activity. Despite the lack of an increase in Cdk2 expression, induction of competence also can activate the pre-existing Cdk2 protein. Stimulation of competent cells by IL-2 then induces a further increase in the expression of Cdk4 and cyclin D2 and expression of new Cdk2 as these cells progress through G, and into S phase.

MATERIALS AND METHODS
Cell Cultures-Peripheral blood T cells were purified from plateletpheresis residues by Ficoll-Hypaque (1.077 g/ml) density gradient centrifugation followed by depletion of adherent cells from the peripheral blood mononuclear cells and E-rosetting on neuraminidase-treated sheep erythrocytes as described (15,161. The phenotype of the resulting T-cell population was 296% CD3+, 56% DR', 52% CD16+, and <1% CD20' or CD14'. The IL-2-dependent Kit-225 human T-cell leukemia line was a gift from Dr. T. Uchiyama (Kyoto, Japan, Ref. 17). Cells were maintained in culture in RPMI 1640 supplemented with 10% (v:v) heatinactivated fetal calf serum, 2 m M L-glutamine, 2 m M sodium pyruvate, and 10 m M HEPES, in a 5% CO, atmosphere at 37 "c. Kit-225 cell cultures were maintained in the presence of human recombinant IL-2 (2 nM). The proliferative response of T cells to mitogens was determined by the incorporation of trichloroacetic acid-precipitable t3Hlthymidine into DNA, 48 h after the onset of culture, as described (18). Phorbol 12,13dibutyrate (PDB) and ionomycin (Calbiochem) were dissolved at a concentration of 10 PM and 500 in dimethyl sulfoxide, respectively. Human recombinant IL-2 was obtained from Hoffman-La Roche, Inc. through the Biologic Response Modifiers Program, Division of Cancer Treatment, NCI. Cyclosporin A (CsA, Sandoz, Basel, Switzerland) and FK506 (Fujisawa Pharmaceutical Co., Ibaraki, Japan) were prepared as 1 mg/ml stock solutions in ethanol.
Induction of Competence-Competent T cells are those that exit Go but stop their passage through the cell cycle at a defined point before The abbreviations used are: Cdk, cyclin-dependent kinase; CsA, cyclosporin A , IL-2, interleukin-2; PDB, phorbol 12,13-dibutyrate; FITC, fluorescein isothiocyanate; PAGE, polyacrylamide gel electrophoresis.
Gelfand, unpublished data. entry into the S phase. These cells have been termed competent because the addition of a progression signal (such as IL-2) will induce the cells to progress into S phase and undergo cell division. Peripheral blood T cells were rested overnight in complete medium. These cells were rendered competent to proliferate as described (16). Briefly, following a 20-min incubation with PDB (10 m) and ionomycin (500 nM), cells were washed three times with phosphate-buffered saline solution a t 4 "C.
The competent cells were then divided into two groups. Anti-human recombinant IL-2 antibody (2 pg/ml) was added to one group, while the other group received human recombinant IL-2 (25 n~) for the duration of culture. In each experiment, a n equal number of T cells as were rendered competent were allowed to remain unstimulated (received vehicle) or were stimulated to proliferate by the addition of PDB and ionomycin throughout the culture period ("proliferating" T cells).
Synchronization of Kit-225 cells in early G, was accomplished by washing the cells extensively in PBS and incubating them in complete medium deprived of IL-2 for 72-96 h as described (19). During this period, the cells were washed in PBS and citrated saline solution (pH 6.5) daily to remove any residual IL-2 and re-incubated in fresh medium. The degree of synchrony in these cells was assessed for each experiment by measuring DNA synthesis. Northern Blotting-Cytosolic RNA was isolated by the method of Wilkinson (211. Ten to twenty pg of RNA were separated electrophoretically in 1% agarose, 5% formaldehyde denaturing gels and transferred to Nylon 66 membranes (MSI, Westboro, M A ) by capillary blotting.
Steady state levels of mRNA expression were assessed by hybridization of human Cdk4 cDNA (PSKJ3, 22), Cdk2 cDNA (23), cyclin D2 cDNA (24), IL-2 cDNA (25), or a human IL-2 receptor cDNA (26). cDNA encoding the human &-microglobulin gene (27) was used to ensure that the amounts of RNA present in each sample were approximately equivalent. The cDNA probes were labeled by random-primer extension and hybridized to the immobilized RNA as described (28). Blots were autoradiographed on XAR-5 film (Kodak) or Reflection film (Dupont NEN), and autoradiograms were quantitated densitometrically using the ScanAnalysis software in a Macintosh IIci computer.
Irnmunoblotting-Cells were lysed in a buffer containing 300 m M NaCl, 50 m~ Tris, pH 7.6, 0.5% Triton X-100, 1 mM N-ethylmaleimide, 0.03 p~ aprotinin, 0.5 1.1~ leupeptin. Protein content in the lysates was determined using the DC Protein Assay Kit (Bio-Rad Laboratories). Ten to fifty pg of cellular protein were separated by sodium dodecyl sulfate (SDS)-polyacrylamide gel electrophoresis (PAGE) in 10% or 12% minigels and transferred to nylon-re-enforced nitrocellulose membranes (Hybond-ECL, Amersham). The blots were probed with rabbit antihuman Cdk4 antibody (directed against the C-terminal 16 amino acids of the protein, Upstate Biotechnology, Inc. (Lake Placid, NY), antihuman Cdk2 antibody (directed against residues 287 to 298 of the protein, Upstate Biotechnology, Inc.), anti-human cyclin D2 antibody (directed against residues 104 to 123 of the protein, Santa Cruz Biotechnology, Santa Cruz, CA), or anti-human IL-2 antibody (Collaborative Biomedical, Bedford, MA), followed by secondary anti-rabbit antibody conjugated to horseradish peroxidase (Amersham Life Sciences) or alkaline phosphatase (Promega Biotech). Detection was performed for horseradish peroxidase by the enhanced chemiluminescence (ECL, Amersham Life Sciences) method, and for alkaline phosphatase using the Protoblot AP system (Promega) according to the manufacturers' instructions. Quantitation of the blots was performed densitometrically.
Cdk2 Kinase Assays"cdk2 kinase activity was determined as described (13). Briefly, unstimulated, competent, and proliferating T cells were cultured for 20 to 24 h and lysed in a buffer containing 250 m~ NaC1, 50 m~ Tris, pH 7.4, 5 m~ EDTA, 0.1% Nonidet P-40, 100 1.1~ Na3V0,, 50 mM NaF, 30 m aprotinin, and 500 I " leupeptin. The Cdk2associated complexes were immunoprecipitated with rabbit anti-human The samples were boiled for 5 min a t 95 "C, and the proteins were separated in 12% gels by SDS-PAGE. The labeled histone bands were stained with Coomassie Blue, excised, and quantitated by liquid scintillation counting. One unit/min of Cdk2-associated activity was defined as the incorporation of 1 fmol of phosphate/min into the substrate. We have demonstrated previously that the predominant Cdk2 activity present in activated T cells at this time is attributable to Cdk2,cyclin E c~mplexes.~

Progression of Competent T Cells through the Cell Cycle-
Resting peripheral blood T cells comprise a population of cells that are predominantly (>96%) in the Go phase of the cell cycle and lack significant expression of cytokine receptors (15,16,(32)(33)(34), and data not shown). These resting T cells can be rapidly stimulated to enter the cell cycle and proliferate in a synchronous manner upon continuous stimulation with PDB and ionomycin or they can be rendered competent without proliferating by a brief stimulation with PDB and ionomycin for 20 min followed by washing (16).
The inability of the competent cells to progress through G, phase is not due to cytotoxicity or cell death, but rather is due to their inability to sustain critical levels of lymphokine production (especially lL-2), despite the expression of cytokine receptors (15,16,(32)(33)(34). Fig. L4 shows that resting T cells did not express IL-2 mRNA, nor did they contain any detectable IL-2 protein (Fig. 1 B ) as determined by Northern blotting and immunoblotting, respectively. T cells stimulated continuously by PDB and ionomycin expressed high levels of IL-2 mRNA by 2 to 4 h that were sustained over 24 h and accumulated high levels of IL-2 protein in their cytoplasm (Fig. 1, A and B ) . The amounts of intracellular IL-2 found in these cells have been shown to be an accurate reflection of the levels of IL-2 secreted into culture supernatants (28). In contrast to the up-regulation of receptors for IL-2 and transferrin (Ref. 16 and data not shown), neither IL-2 gene expression nor IL-2 protein accumulation was detectable in competent T cells (Fig. 1, A and B). Also, the addition of exogenous IL-2 did not induce expression of the endogenous IL-2 gene in these cells (Fig. U). In terms of IL-2 production, competent T cells closely resemble T cells treated with the immunosuppressants cyclosporin A (CsA) or stimulated by IL-2 have heen described in detail previously (19). These cells arrest their growth in early G, when they are deprived of IL-2 for 72 h and re-enter the cell cycle synchronously upon the addition of exogenous IL-2. The proliferative response of these cells was independent of changes in cytosolic Ca" concentrations or protein kinase C activation (19). but was inhibited by anti-IL-2 antibody.
Both IL-2-indrprndrnt and IL-2-ckpendrnt Signals Regulate the Expression of Cdk4 in T Crlls-The systems described above (the competence/progression system in T cells, the Kit-225 cell line, and the treatment of mitogen-stimulated T cells with the immunosuppressive drugs CsA and FK506) provided us three models to analyze IL-2-dependent 1wrsu.s IL-2-independent signals that may be required for the expression of cyclin-dependent kinases (Cdk) in human T cells. Of the Cdk family members studied to date, Cdk4 is among the earliest Cdk whose expression is up-regulated in mitogen-stimulated T cells.' In the experiments described here, we first evaluated the steady state level of expression of the cdk4 gene and protein by Northern blotting and immunoblotting in resting, competent, or proliferating T cells. As shown in Fig. 2 A , the Cdk4 message was undetectable in resting cells (lanr 1 and was detectable in proliferating cells by 4 h (lane 6). The steady state levels of the message peaked shortly thereafter and remained at this level for at least 24 h (lane 7 ) . Induction of competence (20-min incubation with PDR and ionomycin) increased the level of cdh4 gene expression (lanr 2 ) , but the steady state mRNA levels of Cdk4 did not increase beyond those seen at 4 to 6 h in these cells ( l a m .? 1. The addition of exogenous IL-2 to comprtent T cells did not affect the initial c.xprc.ssion of Cdk4 I i.r. that seen 4 hours after stimulation; compare lnnr 4 ~' r r s~r s lnnr 2 I. but the Cdk4 mRNA levels were markcdly increasrd :It 24 h (lanr -5 ).
The treatment of mitogen-stimulated T crlls with C s A f o r FK506) resulted in a pattern of Cdk4 mRNAexpression similar to that seen in competent T cells: Cdk4 messagr in t h r w rrlls was only inhihited partially (Fig. 2R I , furthrr dclinc.nting the CsA-resistant or IL-2-independent component of Cdk4 mRNA expression. The addition ofexogenous IL-2 rrstored thr c'xpression of Cdk4 to levels comparahle to those sren in the ahsrncr of CsA (Fig. 2R ).
We also evaluated the exprrssion of Cdk4 in p-onth-arrrstrd Cdk4 protein also increascd and accumulated in small amounts following stimulation with PDR nlonr or with ionomycin alonr (Fig. 2F). Detectahle levels of Cdk4 protein \vrrv s r r n in competent T cells within 8 h following stimulation 1 Fig. 211) I. hut the steady state amount of Cdk4 protrin did not incrensr noticeably in these cells for the following 12 h (Fig. 2 0 I. A s was the case for mRNAexpression. the addition of 11,-2 to t h e w cells induced a further increase in the steady state Irvrls of Cdk4 protein (Fig. 2D). CsA and FK5O6 completely ahrogatrd the accumulation of intracellular 11, -2 in mitogrn-stimulatrd T cells, hut they only partially inhihited the accumulation of Cdk4 protein in these T cells (Fig. 2 F

The Exprrssion of Cdk2 in T Cclls Rryrrircs an II,-2-tlcprnrlent Signal-\Ve
showed previously that Cdk2 is expressrd tiuring mid-G, in proliferating T cells, at which time it associntw with cyclin E to form a functional complex.' \Ve c.valuatrtl t h r steady state expression of the cdk2 genr and protein in unstimulated, competent, or prolifrrating T cells to detrrminc. whether such expression could he stimulated hy simals that mediate the induction of competence. Fig. 3  lation (Fig. X ) . As was the case for unstimulated T cells, we had difficulty detecting cyclin D2 mRNA expression in competent T cells, hut there was an increase in the expression of hoth the slower and the faster mohility forms of cyclin D2 protein in these cells (Fig. X ) . The addition of IL-2 to these cells enhanced the expression of cyclin D2 protein, and, in fact, frequently resulted in the accumulation of a greater amount of this protein than that seen in the T cells stimulated by PDF3 and ionomycin (Fig. X ) .
We also assessed the relative requirement for IL-2-depende n t signals in the expression of cyclin D2 mRNA. As was reported previously (14). the expression of cyclin D2 mRNA in lysed at the indicated time. and Cdk4 cnmplexrn wrrr immunoprrclpitated using an anti-human Cdk4 antihody. In c.rfro klnaw :Ictivity of thc immunoprecipitated complexes was drterminrd using a rrcrlmhinant. truncated form of t h r RR protein ip56'"'1 :IS a suhstr:ltt*. S1rnll:lr rrsult.; were ohtained in four srparate rxpcmmrnts. proliferating T cells was inhihited markedly in the presrncrb o f CsA (Fig. 3 0 ). As was the case for Cdk4. the addition of exogenous IL-2 restored the expression of cyclin D2 (Fig. 311 I. T h e w data indicate that the activation of the cells by a hrief exposurr to PDR and ionomycin was sufficient to induce an IL-2-independent, CsA-resistant phase of Cdk4 and cyclin D2 cxprcssion. In addition, the expression of Cdk4 and cyclin D' L increased further following an IL-2-dependent sipnnlf s~.

Induction of Cornprtrnrr in Norrnnl T C'rlls R r s u l t s in
Actiuation of Cdk4 and CdhP-It was important to determine whether the IL-2-independent expression of Cdk4 and cyclin D2 was functionally significant. Therefore.
Cdk4 complexes were immunoprecipitated from resting, competent. or proliferating T cells 12 or 20 h following stimulation, and the activity of these complexes was examined in an in c-itro assay using recombinant, truncated RR protein (pFifiHh) a s :I suhstratr. The Cdk4 kinase activity present in these cells closely parallelrd the expression of the Cdk4 and cyclin D2 proteins. Fig. 4 shows that Cdk4 immunoprecipitates from resting T crlls contained little kinase activity (Ianr 1 ). In contrast. proliferating T crlls contained detectahle levels of Cdk4 kinase activity 12 h after stimulation ( / a m 4 ) that continued to increasr for at least 20 h.
Significant levels of Cdk4 kinase activit.v were dr.tc.ctahlr in competent T cells 20 h after stimulation (lnnr . ' I I . The Cdk4 kinase activity in these cells after 12 h was only slightly increased over that seen in the resting crlls. In other experiments, we determined that the faster mohility specirls of cyclin D2 was the predominant form of this protein present in association with Cdk4 in the active complcxcs rccovcrrd from competent or proliferating T cells. as determined hy immunohlot analyses of the Cdk4 immunoprecipitates (data not shown I.
As described ahove, even though no Cdk2 mRSA was drtectable in resting T cells, these cells expressed low levels of Cdk2 protein (Fig. 3 R ) . However, this Cdk2 protein Iackrd kinase activity as measured in ceitro using histone 111 a s a suhstrntr (Tahle I). Cdk2 kinase activity increased simificantly in T cc.lls stimulated hv PDR and ionomycin for 22 h (Tahle 11. Furthrrmore, despite the lack of an increase in rdk2 gene expressinn or Cdk2 protein accumulation in competent T crlls. the Cdk2 kinase activity increased on average to 5 5 5 of that seen in the proliferating cells 22 h after stimulation ITnhlc I I. Stimulation of competent T cells by IL-2 reconstitutrd the Cdk2 kinasr activity in these cells to an average of 8 K of t h r If.vc.ls seen in proliferating cells, presumably due. a t least in part, to the concurrent increase in Cdk2 protein levels (Tahle I I.

DIS('USSI0X
We compared the requirements for rdh2 and rdh4 gene expression. protein accumulation, and activation in mitogrn-activated normal human T cells. The data demonstratr the dif- Cdk2 kinase activity was determined in vitro in resting T cells, T cells that were rendered competent, competent T cells that were stimulated with IL-2, or T cells that were stimulated in the continuous pres- ferential requirements for the regulation of these proteins in terms of IL-2 dependence. Significant levels of Cdk4 message and Cdk4 protein with RB-kinase activity were inducible in activated T cells independent of a detectable IL-2-mediated signal. These findings were observed in cells that were rendered competent by a brief stimulation with PDB and ionomycin followed by extensive washing. Induction of competence, which does not lead to an increase in IL-2 mRNAor protein, but results in the up-regulation of IL-2 receptor expression, was associated with a n increase in Cdk4 mRNA and protein expression that ranged from 15 to 40% of the levels seen in cells cultured in the continuous presence of PDB and ionomycin. Furthermore, a comparable degree of Cdk4 kinase activity was associated with the Cdk4 protein that accumulated following the competence signals. The initial component of cdk4 gene expression and protein accumulation was resistant to CsA, a potent inhibitor of IL-2 gene transcription. This suggests that the transient activation of protein kinase C by PDB and calcium entry stimulated by ionomycin were sufficient to induce cdk4 gene expression and protein accumulation of the Cdk4 kinase in normal human T cells. We also showed that sustained stimulation of these cells by PDB alone induced the accumulation of small amounts of Cdk4 protein.' Thus, the transcription of the cdk4 gene may be regulated, at least partly, through protein kinase C-responsive transcription factors like NF-KB or AP-1 (37)(38)(39)(40).
It is noteworthy that the Cdk4 protein induced in competent T cells could phosphorylate RB in uitro, although the appearance of Cdk4 kinase activity in competent T cells was delayed as compared to that in proliferating T cells. It is most likely that this delay could be attributable to the smaller levels of Cdk4 present in the competent T cells at the earlier time points examined (8 to 12 h); however, we cannot exclude the possibility that there also was a qualitative difference in the signaling pathways that activate this protein following stimuli that render the cells competent or induce them to undergo proliferation.
Addition of IL-2 to competent T cells induced a further increase in the steady state levels of Cdk4 mRNA and protein, approaching those levels observed in T cells cultured in the presence of PDB and ionomycin. This second increase in Cdk4 expression was clearly IL-2-dependent since it was not detected in cells continuously exposed to PDB and ionomycin and cultured in the presence of CsA. This pattern of regulation is similar to that of the a chain of the IL-2 receptor, where stimulation by antigen or mitogen induces a n initial, IL-2-independent phase of IL-2 receptor gene expression that is followed by a second IL-2-dependent phase (30,41). This IL-2-dependent stimulation of IL-2 receptor gene expression is mediated through signals associated with the activation of non-receptor tyrosine kinases, rather than through the activation of protein kinase C or calcium entry (28,4244). Thus, it is possible that the Cdk4 promoter may share some of the structural features characteristic of the IL-2 receptor a promoter.
The results from the analysis of cdk4 gene expression in T cells were supported by our findings with the IL-2-dependent Kit-225 cells. Kit-225 cells synchronized by IL-2 deprivation express moderate levels of IL-2 receptor mRNA and protein (19). These cells also expressed low levels of Cdk4 mRNA. Following addition of IL-2, there was a marked increase in the levels of Cdk4 that were analogous to the levels seen for the IL-2 receptor gene (19).
The regulation of expression of G, cyclins has been analyzed previously in competent or proliferating fibroblasts stimulated by platelet-derived growth factor in the presence or absence of serum (11). Induction of competence in these cells stimulated the protein kinase C-dependent expression of cyclin Dl, but a progression signal was required for the expression of cyclin D3. The expression of Cdk4 or Cdk2 was not evaluated in these cells. T cells do not express cyclin D l (14); but appear to use cyclin D2 as a n early G, cyclin. In the present experiments, the expression of cyclin D2 appeared to be regulated concordantly with Cdk4. Our results show a CsA-resistant phase of cyclin D2 protein accumulation that followed T cell activation and was independent of IL-2 signaling. Furthermore, the cyclin D2 induced under these conditions associated with Cdk4 to form complexes that phosphorylated RB in vitro. As was the case for Cdk4, in addition to this initial, CsA-resistant phase of cyclin D2 induction, IL-2 stimulated a further increase in the expression of cyclin D2 mRNA and the accumulation of cyclin D2 protein. This second increase in cyclin D2 accumulation observed in T cells stimulated with PDB and ionomycin was abrogated in the presence of CsA, underscoring the role of IL-2 in triggering this phase of the response. Previous experiments also showed that CsA had no effect on cyclin D2 mRNA expression in the presence of anti-CD28 antibodies that activate a CsA-independent pathway of IL-2 production (14), and similar results were seen in mature murine IL-3-dependent hematopoietic cells that were transfected with functional, high affinity IL-2 receptor proteins and induced to proliferate in response to IL-2 (45).
In contrast to Cdk4 and cyclin D2, the expression of Cdk2 showed a dependence for IL-2-mediated signaling in the activated T cells. No aspect of the increase in Cdk2 expression observed in proliferating T cells could be detected in T cells simply rendered competent by the limited exposure to PDB and ionomycin or when CsA was included in the cultures. This suggests that the transcriptional (or post-transcriptional) regulation of Cdk2 is dependent on a n IL-2-responsive element(s), or alternatively, that the expression of Cdk2 is associated with one or several progression events (through G,) and is therefore only indirectly dependent on the IL-2 signal.
Despite the lack of an increase in Cdk2 protein accumulation, induction of competence induced the activation of the preexisting Cdk2 in normal T cells. This suggests that signals that lead to the expression and activation of Cdk4 are sufficient to activate Cdk2, and that these two events may be causally related. These data also are consistent with the findings of Ewen et al. (46) that showed transforming growth factor p l inhibited the expression of Cdk4 and the activation of Cdk2 in mink lung epithelial cells, and that overexpression of Cdk4 restored the activation of Cdk2 in these cells. Similar to the effects seen in the capacity of competent cells to proliferate in response to

IL-2-independent Expression and Activity of Cdk4 in T Cells
IL-2, this progression signal increased the Cdk2 kinase activity in competent T cells to levels comparable to those seen during late G, phase (20 to 24 h after stimulation) in the proliferating T cells.
Our results thus distinguish early cell cycle regulatory events that follow T cell activation from those that control later cell cycle progression. A model based on these data would predict that protein kinase C-dependent pathways, at least in part, induce the initial biochemical event($ necessary for cdk4 and cyclin D2 gene expression, and for Cdk4 activation that in turn may lead to an initial phase of Cdk2 activation, The interaction of IL-2 with its receptor initiates a second wave of biochemical events that activate increased expression of Cdk4 and cyclin D2, as well as de nouo expression of Cdk2 with a concomitant increase in the activity of both proteins. Whether or not the initial IL-2-independent increases in Cdk4 kinase activity are prerequisites for the subsequent IL-2-dependent increases in activity of this enzyme and of Cdk2.cyclin complexes remains t o be determined.