P38 Mitogen-Activated Protein Kinase Contributes to Cell Cycle Regulation by Camp in FRTL-5 Thyroid Cells

European Journal of Endocrinology (2005) 153 123–133 ISSN 0804-4643

EXPERIMENTAL STUDY p38 mitogen-activated protein kinase contributes to cell cycle regulation by cAMP in FRTL-5 thyroid cells C Corre`ze, J-P Blondeau and M Pome´rance Unite´ 486 INSERM-PARIS XI, Transduction Hormonale et Re´gulation Cellulaire, Faculte´ de Pharmacie, 92296 Chaˆtenay-Malabry, France (Correspondence should be addressed to M Pome´rance; Email: [email protected])

Abstract Objective: Thyrotropin activates the cAMP pathway in thyroid cells, and stimulates cell cycle progression in cooperation with insulin or insulin-like growth factor-I. Because p38 mitogen-activated protein kinases (p38 MAPKs) were stimulated by cAMP in the FRTL-5 rat thyroid cell line, we inves- tigated (i) the effect of the specific inhibition of p38 MAPKs on FRTL-5 cell proliferation and (ii) the mechanism of action of p38 MAPKs on cell cycle control, by studying the expression and/or the activity of several cell cycle regulatory proteins in FRTL-5 cells. Methods: DNA synthesis was monitored by incorporation of [3H]thymidine into DNA and the cell cycle distribution was assessed by fluorescence-activated cell sorter analysis. Expression of cell cycle regulatory proteins was determined by Western blot analysis. Cyclin-dependent kinase 2 (Cdk2) activity associated to cyclin E was immunoprecipitated and was measured by an in vitro kinase assay. Results: SB203580, an inhibitor of a and b isoforms of p38 MAPKs, but not its inactive analog SB202474, inhibited DNA synthesis and the G1-S transition induced by forskolin plus insulin. SB203580 inhibited specifically p38 MAPK activity but not other kinase activities such as Akt and p70-S6 kinase. Treatment of FRTL-5 cells with SB203580 decreased total and cyclin E-associated Cdk2 kinase activity stimulated with forskolin and insulin. However, inhibition of p38 MAPKs by SB203580 was without effect on total cyclin E and Cdk2 levels. The decrease in Cdk2 kinase activity caused by SB203580 treatment was not due to an increased expression of p21Cip1 or p27Kip1 inhibitory proteins. In addition, SB203580 affected neither Cdc25A phosphatase expression nor Cdk2 Tyr- 15 phosphorylation. Inhibition of p38 MAPKs decreased Cdk2-cyclin E activation by regulating the subcellular localization of Cdk2 and its phosphorylation on Thr-160. Conclusions: These results indicate that p38 MAPK activity is involved in the regulation of cell cycle progression in FRTL-5 thyroid cells, at least in part by increasing nuclear Cdk2 activity.

European Journal of Endocrinology 153 123–133

Introduction cyclin E partner. The resulting cyclin E–Cdk2 heterodi- mer remains inactive because of the presence of inhibi- Thyroid follicular cell proliferation is under the control tory phosphorylation of Tyr-15 and Thr-14 on Cdk2, of hormones and growth factors such as thyrotropin close to its ATP binding site. Tyr-15 is phosphorylated (TSH), insulin-like growth factor-I (IGF-I), and insulin by the mixed-lineage kinase Wee-1. Translocation of which also activates the IGF-I receptor when used at the cyclin E–Cdk2 complex in the nucleus contributes a high concentration. Increased proliferation of thyroid to the regulation of its activity. Activation of the cyclin cells is associated with several pathological states, E–Cdk2 complex is regulated by phosphorylation of including autoimmune diseases of the thyroid gland Cdk2 on a conserved threonine in the activation loop and thyroid carcinomas. (Thr-160) by Cdk activating kinase (CAK) and by sub- Cell cycle progression in mammalian cells is driven by sequent dephosphorylation of Tyr-15 and Thr-14 the sequential and periodic activation of a family of ser- amino acid residues by Cdc25A phosphatase. The ine/threonine kinases named cyclin-dependent kinases Cip/Kip family of cyclin-dependent kinase inhibitor (Cdks) (1, 2). Transition from G1 to S phase is regulated (CKI) proteins plays a crucial role in proliferation. It at several steps during cell cycle progression: cyclin D– was initially assumed that the Cip/Kip proteins Cdk4 and –Cdk6 complexes are active during the mid- (p27Kip1, p21Cip1, and p57Kip2) inhibit both Cdk4/6 G1 phase, whereas cyclin E–Cdk2 complexes are and Cdk2 (1, 2). However, it is now recognized that active at the G1-S transition. Activation of monomeric cyclin D-dependent kinases sequester Cip/Kip proteins, Cdk2 subunit requires as a ﬁrst step the binding of its thereby facilitating cyclin E–Cdk2 activation (3).

q 2005 Society of the European Journal of Endocrinology DOI: 10.1530/eje.1.01942 Online version via www.eje-online.org

Downloaded from Bioscientifica.com at 09/30/2021 11:59:15AM via free access 124 C Corre`ze and others EUROPEAN JOURNAL OF ENDOCRINOLOGY (2005) 153

Proliferation, differentiation and function of the fol- other signaling pathways that may be involved in the licular thyroid cells are regulated by TSH. Most of these regulation of FRTL-5 cell proliferation, such as the effects are mediated by a cAMP-mediated pathway that PI3K/Akt/S6K pathway. We show that inhibition of requires the activation of the cAMP-dependent protein p38 MAPKs decreased the G1-S phase progression kinase A (PKA). The TSH effects are mimicked by the and cyclin E–Cdk2 kinase activation in FRTL-5 cells diterpene forskolin which is a potent direct activator of stimulated with forskolin plus insulin. The decrease in adenylyl cyclase. The FRTL-5 rat thyroid cell line has Cdk2 activation could not be explained by the inhibi- been extensively used to study the mechanisms that con- tory phosphorylation of Cdk2 on Thr-160 or by trol G1-S progression. The effects of TSH and insuli- changes in the amounts of several key proteins involved n/IGF-I on DNA synthesis in FRTL-5 cells are either in cell cycle control. The mechanism by which p38 synergistic or additive (4, 5), depending on the cell sub- MAPKs interfere with G1-S phase progression may lines cultured in different laboratories. In PC Cl3 and involve, at least in part, the nucleocytoplasmic localiz- WRT rat thyroid cell lines, insulin/IGF-I is also a permiss- ation of Cdk2. ive factor for TSH stimulation of cell proliferation (4), although insulin itself acts as a mitogen in WRT cells (5). In our FRTL-5 cell line, forskolin alone does not Materials and methods stimulate DNA synthesis, in agreement with results Materials showing that TSH alone has no effect on DNA synthesis in WRT thyroid cells (6). The cAMP/PKA-dependent SB203580, SB202274, SKF86002, PD169316, pathway has been shown to activate p70-S6 kinase LY294002, calf thymus histone H1 and forskolin (S6K1) in thyroid cells (7), which is essential for cell were purchased from Calbiochem EMD Biosciences cycle progression in many cell types. Although the acti- Inc. (Darmstadt, Germany), propidium iodide was vation of PKA was necessary for cell cycle progression, from Sigma-Aldrich Co. (St Louis, MO, USA), it has been suggested that PKA-independent mechan- [methyl-3H]thymidine and [g-32P]ATP (specific activity: isms might be required in cAMP-mediated growth of 3000 Ci/mmol) were from PerkinElmer Inc. (Boston, thyroid cells (8, 9). Ras activity was required for DNA MA, USA) and protein G-Sepharose was from Amer- synthesis in TSH-stimulated WRT cells (10), but TSH sham Biosciences (Little Chalfont, Bucks, UK). Retino- did not activate directly the p42/p44 mitogen-activated blastoma (Rb) protein (sc4112) and antibodies protein kinase pathway in FRTL-5 cells and in dog thyr- directed against Cdk2 (sc-163), Cdk4 (sc-260), cyclin ocytes (11, 12). In addition, TSH or cAMP did not acti- E (sc-481), p21Cip (sc-397), p38 MAPK (sc-535) and vate directly the phosphatidyl inositol-3 kinase Akt1 (sc-1619) were purchased from Santa Cruz Bio- (PI3K)/Akt pathway in dog thyrocytes (13) or FRTL-5 technology Inc. (Santa Cruz, CA, USA). Monoclonal cells (14), but stimulated the formation of the PI3K– antibodies against p27kip2 (13231A) and b-tubulin Ras complex in the latter cell type (15). (60181A) were obtained from BD Biosciences Phar- Our previous studies established that p38 mitogen- Mingen (San Diego, CA, USA). Antibodies against activated protein kinases (p38 MAPKs) were expressed Cdc25A phosphatase, phospho-Akt1 (Ser-473), and in FRTL-5 cells and were activated via a PKA-depen- phosphotyrosine (PY20) were obtained from Upstate dent signaling pathway (16). Although such a mech- Inc. (Charlotteseville, VA, USA). Antibodies against anism was not found in primary cultures of dog and phospho-S6K1 (Thr-389), phospho-Cdk2 (Thr-160), human thyrocytes (17), we observed expression and and phospho-p38 MAPK (Thr180/Tyr182) were phosphorylation of p38 MAPKs in pathologic human obtained from Cell Signaling Technology Inc. (Beverly, thyroid tissues (M Pome´rance, paper in preparation). MA, USA). Antibodies against phospho-Rb (Ser-780) p38 MAPKs have been implicated in environmental were obtained from Immunotech (Marseille, France). stress responses, in the regulation of inflammation, Peroxidase-conjugated secondary antibodies against and in apoptosis (18). Recently, it was shown that mouse and rabbit IgG were obtained from Promega p38 MAPKs are also involved in proliferation and differ- Corp. (Madison, WI, USA); those against sheep and entiation of several cell types (19, 20). However, the goat IgG were from DakoCytomation (Glostrup, Den- exact mechanisms of p38 MAPK-mediated regulation mark). of cell proliferation remain largely unknown. p38 MAPK activity is inhibited by the pharmacologic Cell culture and transfections inhibitor SB203580, a pyridine-imidazole compound that specifically targets the a and b isoforms (21), Rat thyroid follicular FRTL-5 cells were kindly provided and by expression of dominant-negative mutants of by Dr M Eggo (Birmingham, UK). Cells were routinely p38 MAPK isoforms. In this study, we examined the cultured in Coon’s F-12 medium (Biochrom AG, role of p38 MAPKs in thyroid cell cycle regulation, Berlin, Germany) supplemented with GlutaMAX mostly using the inhibitory properties of SB203580. (2 mmol/l), and with 5% heat-inactivated newborn To exclude possible non-specific effects of SB203580, calf serum (Invitrogen), and three hormones: bovine we checked that this compound did not interfere with TSH (0.5 mU/ml), insulin (10 mg/ml), and transferrin

www.eje-online.org

Downloaded from Bioscientifica.com at 09/30/2021 11:59:15AM via free access EUROPEAN JOURNAL OF ENDOCRINOLOGY (2005) 153 p38 MAPK in FRTL-5 thyroid cell cycle regulation 125

(5 mg/ml) (all Sigma Chemical Co.). For starvation, sub- bated overnight with the appropriate primary antibody. confluent cells were incubated in basal medium with- The membranes were washed with TBS/T and then out serum and hormones for at least 72 h. Attempts incubated for 1 h at room temperature with a to obtain stable transfectants expressing dominant 1:25 000 dilution of appropriate horseradish negative p38aAGF and p38dAGF were performed as peroxidase-conjugated secondary antibody. Antibody follows: FRTL-5 cells grown in 60 cm2 plates under binding was revealed by the ECL system (Amersham standard conditions were transfected with 10 mgof Biosciences). For immunoprecipitation, cell lysates either the dominant negative p38aAGF or p38dAGF (200–300 mg protein) were incubated for 2 h at 4 8C cDNA expression plasmid (gifts from Dr H Enslen, with 2 mg Cdk2, Cdk4 or cyclin E antibody. The Paris, France) or the corresponding empty vector immune complexes were immunoprecipitated by pro- (pcDNA3), using Fugene reagent (Roche Molecular Bio- tein G-Sepharose. Immune complexes were pelleted by chemicals, Indianapolis, IN, USA). Stable transfectants centrifugation at 10 000 g and washed twice with were selected with 0.2 mg/ml G418 for 2 weeks. buffer A and a further three times with kinase assay buffer (20 mmol/l Hepes pH 7.4, 5 mmol/l MgCl2, Cell extracts 2 mmol/l EGTA, 1 mmol/l Na3VO4, 2 mmol/l dithiothreitol) prior to kinase assays. To prepare whole cell lysates, cells were collected and resuspended in cold lysis buffer A (10 mmol/l Tris/HCl, Measurement of DNA synthesis pH 7.5, 250 mmol/l NaCl, 10 mmol/l EDTA, 1% Noni- det P-40, 1% sodium deoxycholate, 0.05% SDS, Triplicate cultures of quiescent FRTL-5 cells 5 1 mmol/l Na3VO4, and a protease inhibitor mixture (0.5–1 £ 10 cells/ml) were pretreated with inhibitors (1 mmol/l 4-(2-aminoethyl)-benzenesulfonyl fluoride or vehicle (Me2SO) for 1 h and then stimulated with for- (AEBSF) 1 mmol/l benzamidine, 10 mg/ml soybean skolin (10 mmol/l), insulin (10 mg/ml) or both, and trypsin inhibitor, 10 mg/ml aprotinin, 1 mg/ml leupep- pulsed with 1026 mol/l thymidine and 1 mCi/ml tin, 1 mg/ml antipain, 1 mg/ml pepstatin A)). The [methyl-3H]thymidine 3 h prior to the end of the treat- lysates were centrifuged at 10 000 g for 10 min and ment protocol. The cells were washed twice with ice- assayed for protein using bicinchoninic acid-protein cold phosphate buffered saline (PBS), precipitated assay reagents (Biotech Inc, Rockford, IL, USA). For with ice-cold 5% trichloroacetic acid, and lysed with Western blot analysis with phospho-antibodies, cell 0.02 mol/l NaOH in 1% SDS. The radioactivity was extracts were prepared as described previously (16). counted by liquid scintillation spectrometry (Packard- To study the subcellular localization of Cdk2, nuclear Canberra, Meriden, CT, USA). and cytoplasmic extracts were prepared as follows. Cells were resuspended in buffer (20 mmol/l Hepes Flow cytometry pH 7.9, 1.5 mmol/l MgCl2, 200 mmol/l sucrose, 10 mmol/l sodium glycerophosphate, 10 mmol/l KCl, Following treatment of quiescent cells as above, cells 1 mmol/l dithiothreitol, 1 mmol/l Na3VO4, a protease were harvested by trypsinization, washed twice in PBS, inhibitor cocktail, 0.5% (w/v) Triton X-100), lysed for and then fixed with ice-cold 70% ethanol. Cells were 15 min on ice, and then spun at 1000 g for 10 min at washed twice with PBS and stained with 50 mg/ml pro- 4 8C. The nuclear pellets were washed with the same pidium iodide containing 100 mg/ml RNase for 30 min. buffer and the purity of the nuclei was checked by opti- Cell cycle analysis was performed using a fluorescence- cal microscopy. The nuclear pellets were resuspended in activated cell sorter (FACS) and Cell Quest software ver- buffer containing 20 mmol/l Hepes pH 7.9, 400 mmol/l sion 1.2 (BD Biosciences). At least 10 000 cells were ana- NaCl, 10 mmol/l sodium glycerophosphate, 1.5 mmol/l lyzed per sample. The cell cycle distribution was MgCl2, 1 mmol/l dithiothreitol, 1 mmol/l Na3VO4 and quantified using the ModFit LT software (Verity Software the protease inhibitor mixture, and then incubated for House Inc., Topsham, ME, USA). 30 min at 4 8C. The samples were centrifuged at 10 000 g for 15 min at 4 8C, and the supernatants In vitro kinase assays were removed and stored in aliquots at 280 8C. Cdk4 activity was assayed by incubating immunopreci- Western blot analysis and pitates with 2 mg Rb protein, used as a substrate, for immunoprecipitation 20 min at 30 8C in kinase assay buffer containing 40 mmol/l ATP and 5 mCi [g-32P]ATP. Cdk2 activity For Western blot analysis, proteins were denaturated by was assayed by incubating immunoprecipitates with boiling in Laemmli sample buffer, resolved by 12% SDS- 10 mg histone H1, used as a substrate, for 10 min at PAGE and transferred to nitrocellulose membranes. The 30 8C in kinase assay buffer containing 40 mmol/l membranes were blocked in Tris–buffered saline ATP and 5 mCi [g-32P]ATP. The reactions were stopped containing 0.1% Tween-20 (TBS/T) and 5% non-fat by the addition of Laemmli sample buffer. Samples were dry milk or 3% bovine serum albumin and then incu- subjected to SDS-PAGE. The gels were dried and phos-

www.eje-online.org

Downloaded from Bioscientifica.com at 09/30/2021 11:59:15AM via free access 126 C Corre`ze and others EUROPEAN JOURNAL OF ENDOCRINOLOGY (2005) 153 phorylated Rb protein or H1 histone was quantiﬁed by InstantImager analysis (Packard-Canberra).

Statistical analysis Statistical signiﬁcance was assessed using Student’s t-tests. P values less than 0.05 were considered to be signiﬁcant.

Results Inhibition of p38 MAPKs correlates with inhibition of G1–S transition in FRTL-5 thyroid cells FRTL-5 cells, forced to quiescence by hormonal and serum withdrawal, were pretreated for 1 h with SB203580, a specific p38 MAPK inhibitor, or with the vehicle alone (Me2SO), and then treated with 10 mg/ml insulin plus 10 mmol/l forskolin for different periods of time. Fig. 1A shows that 5 mmol/l SB203580 inhibited by 45–50% the increase in DNA synthesis induced by forskolin plus insulin, as followed by incorporation of [methyl-3H]thymidine into DNA. The inhibitory effect of SB203580 was maximal at 24–30 h after the stimulation, i.e. at the stage where the cells progressively reached G1–S transition. The results of five independent experiments (Fig. 1B) indicated that the effect of SB203580 was highly reproducible (46 and 41% inhibition at 24 and 30 h of treatment respectively). Fig. 1C shows that forskolin alone had no effect on cell cycle progression, while the effect of insulin alone was at least 10-fold lower than that of forskolin plus insulin. SB203580 had no effect on insulin-induced DNA synthesis. As a control, we observed that LY294002, a specific PI3K inhibitor, inhibited in a concentration-dependent manner the insulin-induced DNA synthesis (not shown). This suggests that p38 MAPK acts only when the cAMP pathway and the insulin/IGF-I pathway are activated simultaneously. To rule out a potential nonspecific effect of SB203580, we used SB202474, which has no effect on p38 MAPK activity, as a negative control; similar results were obtained whether SB202474 was present or not. As shown in Fig. 2, SB203580 inhibited Figure 1 Effect of SB203580 on [3H]thymidine incorporation into DNA synthesis induced by a 24h treatment with forsko- DNA. Quiescent FRTL-5 cells were pretreated without or with 5 mmol/l SB203580 for 1 h before stimulation with 10 mmol/l for- lin plus insulin in a dose-dependent manner. The half- skolin or 10 mg/ml insulin or both. Uptake of [3H]thymidine was maximum effect was obtained at 1 mmol/l, whereas the measured as described in the Materials and methods section. (A) inactive molecule had no significant effect. In addition, [3H]Thymidine incorporation was measured as a function of stimu- other pyridine-imidazole p38 MAPK inhibitors, such as lation time by forskolin plus insulin. Data obtained without (X)or W ^ SKF86002 and PD169316, also inhibited growth- with ( ) SB203580 are the means S.E.M. of triplicate wells. (B) [3H]Thymidine incorporation was measured at 24 and 30 h of stimulated DNA synthesis (data not shown). Then, we stimulation by forskolin plus insulin as in (A). Data obtained with- examined the cell cycle distribution by FACS analysis out (solid bars) or with (open bars) SB203580 are the means^ after 24 h and 30 h of stimulation of cells with forskolin S.E.M. of five independent experiments. (C) Cells were stimulated plus insulin in the presence or absence of SB203580. with forskolin (O, K) or insulin (B, A) without (solid symbols) or with (open symbols) SB203580. S.E.M. values, lower than 10% of Figure 3 shows the results of two independent exper- the mean (triplicate wells), were omitted for clarity. Similar results iments. SB203580 treatment reduced the number of were obtained in three independent experiments. *P , 0.05, cells in S phase by 34–54% at 24 h. This reduction **P , 0.005, ***P , 0.001 relative to control cells.

www.eje-online.org

Downloaded from Bioscientifica.com at 09/30/2021 11:59:15AM via free access EUROPEAN JOURNAL OF ENDOCRINOLOGY (2005) 153 p38 MAPK in FRTL-5 thyroid cell cycle regulation 127

Figure 3 Effect of SB203580 on cell cycle progression. Quiescent FRTL-5 cells were stimulated for 24 or 30 h with 10 mmol/l forsko- Figure 2 Dose-dependent inhibition of DNA synthesis by lin plus 10 mg/ml insulin in the absence (X, B) or presence (W, A) SB203580. [3H]Thymidine incorporation was stimulated by forsko- of 5 mmol/l SB203580. Cells were processed by FACS analysis to lin (10 mmol/l) plus insulin (10 mg/ml) for 24 h in the presence of quantitatively determine the percentage of cells in different increasing concentrations of SB203580 (W) or SB202474 (X). The phases of the cell cycle (G0/G1, S, G2/M). Data from two inde- values are the means (^S.E.M.) of three independent experiments. pendent experiments, distinguished by squares (B, A) and circles P,0.01, P , 0.001 relative to SB202474-treated cells. * ** (X, W), are shown on the graph whose Y-axis comprises two different scales. For each experiment, a line connects each value obtained without SB203580 to the corresponding one obtained was <26–28% after 30 h of stimulation. Concomi- with SB203580. tantly, cells in G0/G1 phase increased by 10–15%. Because FRTL-5 cells usually start synthesizing DNA suggesting that p38d MAPK is not involved in regu- 12 to 16 h after stimulation, only a few cells had lation of cell proliferation, in contrast to p38a MAPK. reached G2/M at the time of harvest. To exclude the possible nonspecific effects of Involvement of p38 MAPKs in Cdk2 activation SB203580 on other important signaling pathways by forskolin and insulin involved in FRTL-5 cell proliferation, we examined the effect of the drug on the PI3K/Akt pathway and the Because the inhibitory effect of SB203580 occured at the S6K1 pathway. As shown in Fig. 4A, using an anti- stage where cells reached G1–S transition, we studied phospho-Akt antibody we found that phosphorylated the immunoprecipitated activity of Cdk4 and Cdk2 Akt was detected in insulin-treated, but not in forsko- kinases in the presence or absence of SB203580 in lin-treated FRTL-5 cells. However, SB203580 had no FRTL-5 cells stimulated by forskolin plus insulin. No effect on Akt phosphorylation. SB203580 also had no effect of SB203580 was observed on Cdk4 activity effect on forskolin-induced phosphorylation of S6K1 measured using Rb protein as a substrate (not shown). when referred to the amount of S6K1 protein. We Accordingly, we did not find an effect of SB203580 on used rapamycin, a specific inhibitor of the mTOR pro- endogenous Rb phosphorylation, using an antibody tein kinase involved in S6K1 phosphorylation, as a directed against Ser-780, which is targeted by Cdk4. positive control of inhibition of S6K1 phosphorylation. On the other hand, immunoprecipitation with anti- As expected, rapamycin blocked forskolin-induced bodies directed against Cdk2 or cyclin E followed by S6K1 phosphorylation (Fig. 4B). We attempted to kinase assays on histone H1 showed an increase in express dominant negative forms of p38a and p38d endogenous total Cdk2 or cyclin E-associated kinase MAPK (in which the TGY motif in the activation loop activity, beginning 16 h after stimulation (Fig. 5A and is mutated to AGF) in FRTL-5 cells. All our attempts B). The activity was maximal between 24 and 30 h, at to obtain viable FRTL-5 cells expressing the p38aAGF which time cells initiated de novo DNA synthesis (Figs mutant were unsuccesful, whereas we did obtain 1A and 3). Treatment of FRTL-5 cells with SB203580 FRTL-5 cells that stably expressed p38dAGF. These (5 mmol/l) reduced total and cyclin E-associated Cdk2 latter cells showed a normal proliferative response to activity by 40–50%. The quantification of histone H1 forskolin plus insulin when studied by incorporation phosphorylation obtained in 3 to 5 independent exper- of [methyl-3H]thymidine into DNA (data not shown), iments (Fig. 5C) indicated that this inhibitory effect

www.eje-online.org

Downloaded from Bioscientifica.com at 09/30/2021 11:59:15AM via free access 128 C Corre`ze and others EUROPEAN JOURNAL OF ENDOCRINOLOGY (2005) 153

Figure 4 SB203580 has no effect on the PI3K and S6K1 pathways. (A) Cells were pretreated for 1 h without or with 5 mmol/l SB203580 before stimulation (1 or 3 h) with 10 mmol/l forskolin or 10 mg/ml insulin or both. Cell lysates were analyzed by SDS-PAGE and immunoblotting with antibodies specific for phospho (P)-Akt and total Akt. (B) Cells were pretreated without or with 5 mmol/l SB203580 or 20 nmol/l rapamycin, and unstimulated (0) or stimulated for 30 min by 10 mmol/l forskolin. Cell lysates were analyzed by SDS-PAGE and immunoblotting with antibodies specific for phospho (P)-S6K1 and total S6K1. Similar results were obtained in two separate experiments. Wb, Western blot. was highly reproducible. We checked that SB203580 forskolin plus insulin. Nevertheless, SB203580 did not had no direct inhibitory effect on Cdk2 kinase activity influence p21Cip1 protein levels (Fig. 6). Therefore, the measured in vitro (data not shown). p38 MAPK inhibitor did not modify the expression of Cdk2, cyclin E, and the CKIs under study.

Expression of several cell cycle regulatory p38 MAPK inhibition did not affect Cdk2 proteins upon SB203580 treatment Tyr-15 phosphorylation or Cdc25A expression We studied whether the decrease in Cdk2 activity We studied whether SB203580 affected the inhibitory induced by SB203580 could be due to its effects on phosphorylation of Cdk2 on Tyr-15 which maintains expression of Cdk2, cyclin E, and the CKIs p27Kip1 and Cip1 Cdk2 in an inactive form. Extracts from FRTL-5 cells p21 . Quiescent cells were stimulated with forskolin stimulated for 24 and 30 h with forskolin plus insulin plus insulin for 24 and 30 h in the presence or absence in the presence or absence of SB203580 were immuno- of SB203580, and Western blot analyses for Cdk2, Kip1 Cip1 precipitated with an anti-Cdk2 antibody and probed cyclin E, p27 and p21 were performed. Figure 6 with anti-phosphotyrosine antibody. Tyrosine phos- shows that anti-Cdk2 antibodies recognized two main phorylation increased in correlation with Cdk2 forms of Cdk2 proteins (33 and 34 kDa). The level of expression, but no significant change in phosphotyro- the 33 kDa activated form was very low in quiescent sine content was observed following SB203580 treat- cells and increased 24 h and 30 h after the treatment ment (Fig. 7A). As Cdc25A phosphatase regulates the of cells with forskolin plus insulin. The presence of activity of the Cdk2–cyclin E complex, we examined SB203580 during the cell stimulation did not affect sig- the effect of SB203580 on its expression. Cdc25A, nificantly the expression of the 33 and 34 kDa forms. which appeared as 3 bands ranging from 65 to Cyclin E was barely detectable in quiescent cells and 72 kDa on Western blots, was constitutively expressed increased 24 and 30 h after stimulation. However, the throughout the cell cycle and its expression did not presence of SB203580 did not affect the levels of cyclin undergo major changes in the presence of SB203580 E (Fig. 6). To determine whether SB203580 affected (Fig. 7B). Therefore, Cdk2 phosphorylation on Tyr-15 the expression of CKIs, their amounts were determined Kip1 and Cdc25A expression were unlikely to be involved by Western blot analysis. p27 expression was high in the p38 MAPK effect on Cdk2 activity. in quiescent cells and decreased at 24 and 30 h after stimulation (Fig. 6). p27Kip1 expression could not be detected at 48 h (data not shown). This downregulation Effect of SB203580 on subcellular coincides with the increase in Cdk2 kinase activity (see localization and Thr-160 phosphorylation Fig. 5) and in Cdk2 and cyclin E expression. However, of Cdk2 the level of p27Kip1 protein was not significantly affected by SB203580 treatment (Fig. 6). In contrast, p21Cip1 Cytosolic and nuclear extracts were prepared from expression was very low in quiescent cells and transi- FRTL-5 cells to study the effect of SB203580 on subcel- ently increased 24 h after the stimulation of cells with lular localization of Cdk2 and cyclin E. SB202474 was

www.eje-online.org

Downloaded from Bioscientifica.com at 09/30/2021 11:59:15AM via free access EUROPEAN JOURNAL OF ENDOCRINOLOGY (2005) 153 p38 MAPK in FRTL-5 thyroid cell cycle regulation 129

Figure 5 Inhibitory effect of SB203580 on total and cyclin E-associated Cdk2 activity. Cells were preincubated or not with 5 mmol/l SB203580 for 1 h, and stimulated with 10 mmol/l forskolin plus 10 mg/ml insulin. Immu- noprecipitated (IP) Cdk2 activity was measured in a kinase assay using histone H1 as a substrate and [g-32P]ATP. Histone H1 phosphorylation was quantified by Instant Ima- ger analysis. Cdk2 activity is expressed as fold-activation relative to that measured at zero time. (A) Lysates were immunoprecipitated with an anti-Cdk2 antibody, and Cdk2 activity was measured at the indicated times. (B) An identical set of samples was immunoprecipitated with an anti-cyclin E antibody. (C) Cdk2 activity was measured 24 h after stimulation, without (solid bars) or with (open bars) SB203580 pretreatment. The values are the means^S.E.M. of three to five independent experiments. *P , 0.05, **P , 0.001 relative to control cells. used as a negative control to validate the specificity reduction of the activating Thr-160 phosphorylation of SB203580. Nuclear and cytosolic extracts from of Cdk2 occurred in nuclei from SB203580-treated FRTL-5 cells stimulated with forskolin plus insulin in cells compared with nuclei of SB202474-treated cells. the presence of SB203580 or SB202474 were analyzed The same extracts were monitored for their Cdk2 by Western blot (Fig. 8). Beta-tubulin was used as a kinase activity. The treatment of the cells by cytoplasmic marker. Cyclin E was barely expressed in SB203580 decreased Cdk2 activity in the nuclei, com- nuclear and cytosolic fractions from quiescent cells, pared with that of SB202474-treated cells, with a time consonant with the results obtained with whole cell course similar to that of Thr-160 phosphorylation. extracts (see Fig. 6). After 16 h of stimulation with for- These results suggest that the nuclear translocation of skolin plus insulin, cyclin E was found in the cytosolic Cdk2, but not its activating phosphorylation, is specifi- and nuclear fractions. However, its amount was not sig- cally targeted by p38 MAPK inhibition. nificantly affected in either compartment by SB203580 or SB202474. Significant amounts of Cdk2 were detected in the nuclear fraction of stimulated cells, Discussion whereas Cdk2 was essentially cytosolic in quiescent cells. Furthermore, the level of Cdk2 within nuclei of Dependingon the cell type and stimulus, p38 MAPKs can SB203580-treated cells was repeatedly decreased com- have either a positive or a negative role in the regulation pared with that present in nuclei of SB202474-treated of cell cycle progression. p38 MAPKs have an inhibitory cells. Using an antibody specific for Thr-160-phos- role in cell cycle progression in various cell types (19, phorylated Cdk2, we observed that forskolin plus insu- 20). Conversely, there is accumulating evidence that lin induced the phosphorylation of Thr-160 within 16 activated p38 MAPKs are involved in positive control of to 30 h in the nuclei whereas negligible phosphoryl- cell proliferation. Thus, p38 MAPKs appear to be ation occurred in the cytosolic fraction. A marked required for the proliferation of Swiss 3T3 cells or

www.eje-online.org

Downloaded from Bioscientifica.com at 09/30/2021 11:59:15AM via free access 130 C Corre`ze and others EUROPEAN JOURNAL OF ENDOCRINOLOGY (2005) 153

Figure 6 Expression of cell cycle proteins upon SB203580 treatment. Quiescent cells were preincubated or not with 5 mmol/l SB203580 for 1 h, and stimulated with 10 mmol/l forskolin plus 10 mg/ml insulin. After 24 and 30 h of stimulation, cell lysates were Figure 7 Tyrosine phosphorylation of Cdk2 and expression of analyzed by Western blot (Wb) for Cdk2, cyclin E, p27kip1 and Cdc25A are not modified by SB203580 treatment. Quiescent p21Cip1 expression. A portion of the blots stained with Ponceau S FRTL-5 cells were preincubated or not with 5 mmol/l SB203580 for is shown as a loading control (lower panel). Representative 1 h, and stimulated with 10 mmol/l forskolin plus 10 mg/ml insulin experiments are shown. Similar results were obtained in three to for 24 and 30 h. (A) Cell lysates were immunoprecipitated (IP) six independent experiments. with an anti-Cdk2 antibody and immunoprecipitates were analyzed by Western blot (Wb) using an anti-phosphotyrosine (P-Tyr) antibody (upper panel). The membrane was stripped and airway smooth muscle cells stimulated by fibroblast reprobed with the anti-Cdk2 antibody (lower panel). Similar results growth factor-2 (22, 23), of hematopoietic cells stimu- were obtained in two independent experiments. (B) Expression of lated by cytokines (24, 25), and of vascular smooth Cdc25A was assessed by Western blot with an anti-Cdc25A antibody. The membrane was stripped and reprobed with the anti- muscle and pancreatic stellate cells stimulated by plate- Cdk2 antibody. Similar results were obtained in three independent let-derived growth factor (26, 27). p38 MAPKs are also experiments. required for CD40-induced proliferation of B-lymphocytes (28), and for cell cycle progression of breast the phosphorylation of Akt by insulin, and (iii) we cancer, melanoma and chondrosarcoma cell lines observed no effect of forskolin on the phosphorylation (29–31). However, the signaling pathways and identi- of Akt as previously reported (13, 14). Previous ties of signaling key intermediates, which mediate p38 studies reported that SB203580 does not affect the MAPK involvement in cell cycle progression, are poorly PI3K or the S6K1 pathways in non-thyroid cells characterized. (34). We also exclude an effect of SB203580 on In previous studies, we have reported that the p38 forskolin-stimulated S6K1 phosphorylation in FRTL-5 MAPK pathway was stimulated by forskolin in FRTL- cells. A similar result was also reported in feline cardi- 5 cells (16) and that forskolin-stimulated p38 MAPK omyocytes, where forskolin activates both p38 MAPKs phosphorylation was inhibited by SB203580 (32). and S6K1 (35). It has also been reported that a high Here, data obtained using thymidine incorporation concentration of SB203580 reduced the activity of and FACS analysis show that inhibition of p38 c-Jun N-terminal kinases (JNKs) in rat ventricular MAPKs decreased significantly DNA synthesis by myocytes and in transfected COS-7 cells (36, 37), FRTL-5 cells stimulated by forskolin plus insulin. but JNKs were not activated either by cAMP in We performed several control experiments that FRTL-5 cells (11) or by IGF-I in human thyrocytes argue in favor of a specific p38 MAPK-mediated (38). These results indicate that SB203580 inhibited effect of SB203580 on FRTL-5 cell proliferation. The G1-S transition in FRTL-5 cells by specifically targeting effect of SB203580 on DNA synthesis was not p38 MAPK isoform(s). obtained with SB202474, a structural pyridine-imida- The p38 family of MAPKs is composed of four mem- zole analog which does not inhibit p38 MAPK activity. bers, p38a, b, g, and d (18). In the human thyroid It has been reported that SB203580 reduced the phos- gland, the mRNAs of p38a and p38d MAPK isoforms phorylation and activation of Akt in T cells indepen- are predominantly expressed, compared with those of dently of p38 MAPK activity (33). Such a p38b and g isoforms (39). Only p38a and b MAPKs mechanism is unlikely to occur in FRTL-5 cells are inhibited by SB203580, thus our data suggest because: (i) SB203580 did not affect the slow prolif- that the a isoform is involved in the control of G1–S erative effect of insulin, (ii) SB203580 did not affect transition in FRTL-5 cells. Accordingly, we could not

www.eje-online.org

Downloaded from Bioscientifica.com at 09/30/2021 11:59:15AM via free access EUROPEAN JOURNAL OF ENDOCRINOLOGY (2005) 153 p38 MAPK in FRTL-5 thyroid cell cycle regulation 131

Figure 8 Effect of SB203580 on the subcellular localization of Cdk2. Cytosolic and nuclear extracts were prepared from FRTL-5 cells stimulated with 10 mmol/l forskolin plus 10 mg/ml insulin in the presence of 5 mmol/l SB203580 or SB202474. Western blot (Wb) analyses were performed using anti-cyclin E, anti- Cdk2 or anti-phospho (P)-Cdk2 antibodies. Beta-tubulin was used as a cytoplasmic marker. Cytoso- lic and nuclear extracts were immunoprecipitated with anti-Cdk2 antibody, and the immunoprecipitates were assayed for Cdk2 kinase activity using histone H1 as a substrate. Similar results were obtained in two independent experiments. obtain stable transfectants expressing the dominant S phase and to be positively regulated by Cdc25A negative form of the p38a isoform, whereas we phosphatase (1, 2). In this work, we did not detect obtained transfectants expressing the dominant nega- any significant effect of p38 MAPK inhibition on Tyr- tive form of the p38d isoform in FRTL-5 cells. These 15 phosphorylation of Cdk2. We found that Cdc25A results suggest that the dominant negative a isoform was constitutively expressed throughout the cell cycle interfered with FRTL-5 cell proliferation. in FRTL-5 cells. Furthermore, SB203580 treatment Cotreatment of FRTL-5 cells with TSH and insuli- had no effect on Cdc25A expression. We conclude n/IGF-I increases the expression of cyclins D and E, that the low kinase activity of cyclin E–Cdk2 complexes and thereby increases the corresponding associated in SB203580-treated cells is not due to inhibitory kinase activities (4). Our data indicate that phosphorylation of Thr-14 and Tyr-15 on Cdk2. SB203580 reduced the activity of Cdk2, but not that Anyhow, the role of Cdc25A in the control of Cdk2 of Cdk4. The levels of cyclin E or Cdk2 proteins were activity is now subject to controversy. not affected by SB203580 treatment. One mechanism Several studies have highlighted the importance of the of regulation of Cdk2–cyclin E activity is through its nucleocytoplasmic distribution of Cdks/cyclins com- inhibition by p27Kip1 and p21Cip1 (1, 2), but these pro- plexes in cell cycle control (40, 41), including dog thyroid teins are also shown to be essential activators of cyclin cells (42). Recently,it has been shown that the nucleocy- D complex formation (2). When we examined the effect toplasmic translocation of Cdk2 is under the control of of forskolin plus insulin stimulation on the expression p42/p44 MAPKs in fibroblasts (43, 44). Cytoplasmic of these proteins, only p27Kip1 showed a decrease mislocalization of Cdk2 has also been reported to be a 24 h after stimulation as previously observed (4), key event in vitamin D-mediated growth inhibition of whereas p21Cip1 was increased. To our knowledge, prostate cancer cells (45) and in irradiation-induced there was no previous report on the regulation of the apoptosis in mouse mesangial cells (46). In this study, expression of p21Cip1 in thyroid cells. We observed no we show that the subcellular localization of Cdk2 pro- effect of SB203580 on the regulation of p27Kip1 and vides a novel mechanism whereby p38 MAPKs could p21Cip1 expression by forskolin plus insulin. An effect regulate cell cycle progression. Our results show a of SB203580 on the stabilization of the Cdk2–cyclin decrease in the amount of Cdk2 within nuclei following E–CKI(s) complex is unlikely because p27Kip1 and p38 MAPK inhibition. Furthermore, cyclin E shuttled p21Cip1 expression is downregulated at 30 h of stimu- between the nucleus and the cytoplasm as reported in lation by forskolin plus insulin. other cell types (40, 41), but SB203580 had no effect Cdk2 activity has been shown to be negatively regu- on its translocation in FRTL-5 cells. This finding suggests lated by phosphorylation on Thr-14 and Tyr-15 during that the p38 MAPK pathway is not involved in the regu-

www.eje-online.org

Downloaded from Bioscientifica.com at 09/30/2021 11:59:15AM via free access 132 C Corre`ze and others EUROPEAN JOURNAL OF ENDOCRINOLOGY (2005) 153 lation of the subcellular localization of this cyclin in expression in dog thyroid cells. Molecular and Cellular Biology FRTL-5 cells. 1997 17 6717–6726. 9 Cass LA, Summers SA, Prendergast GV, Backer JM, In fibroblasts, it has been shown that p42/p44 Birnbaum MJ & Meinkoth JL. Protein kinase A-dependent and MAPKs are necessary for Cdk2 Thr-160 phosphoryl- -independent signaling pathways contribute to cyclic AMP-stimu- ation (47, 48), although it may be independent of lated proliferation. Molecular and Cellular Biology 1999 19 their impact on Cdk2 subcellular localization (48). 5882–5891. p42/p44 MAPKs could be involved in regulation of 10 Kupperman E, Wen W & Meinkoth JL. Inhibition of thyrotropin- stimulated DNA synthesis by microinjection of inhibitors of CAK or phosphatase activities. Here, we show that in cellular Ras and cyclic AMP-dependent protein kinase. Molecular FRTL-5 cells, p38 MAPKs have an impact on Thr-160 and Cellular Biology 1993 13 4477–4484. phosphorylation of Cdk2 and on its activity. In nuclei 11 Corre`ze C, Blondeau JP & Pome´rance M. The thyrotropin receptor from SB203580-treated cells, we observe a decrease is not involved in the activation of p42/p44 mitogen-activated in the amount of Cdk2, which likely accounts for the protein kinases by thyrotropin preparations in Chinese hamster ovary cells expressing the human thyrotropin receptor. Thyroid decrease in its phosphorylation on Thr-160, and there- 2000 10 747–752. fore for the decrease in its activity. Taken together, our 12 Van Keymeulen A, Roger PP, Dumont JE & Dremier S. TSH and results suggest that inhibition of p38 MAPK activity cAMP do not signal mitogenesis through Ras activation. Bio- reduces G1–S progression by decreasing nuclear chemical and Biophysical Research Communications 2000 273 154–158. Cdk2 available for cyclin E binding and for phosphoryl- 13 Coulonval K, Vandeput F, Stein RC, Kozma SC, Lamy F & ation by CAK of the resulting complex in rat thyroid Dumont JE. Phosphatidylinositol 3-kinase, protein kinase B and cells. Although the mechanisms underlying impaired ribosomal S6 kinases in the stimulation of thyroid epithelial cell Cdk2 translocation described in this study require proliferation by cAMP and growth factors in the presence of insu- further investigation, these data indicate a role for lin. Biochemical Journal 2000 348 351–358. p38 MAPKs in the regulation of cycle progression in 14 Saito J, Kohn AD, Roth RA, Noguchi Y, Tatsumo I, Hirai A, Suzuki K, Kohn LD, Saji M & Ringel MD. Regulation of FRTL-5 rat FRTL-5 thyroid cells. thyroid cell growth by phosphatidylinositol (OH) 3 kinase-dependent Akt-mediated signaling. Thyroid 2001 11 339–351. 15 Ciullo I, Diez-Roux G, Di Domenico M, Migliaccio A & Acknowledgements Avvedimento EV. cAMP signaling selectively influences Ras effec- tors pathways. Oncogene 2001 20 1186–1192. 16 Pome´rance M, Abdullah HB, Kamerji S, Corre`ze C & Blondeau JP. We are grateful to Jacqueline Breárd for help and stimu- Thyroid-stimulating hormone and cyclic AMP activate p38 mito- lating discussions on fluorescence-activated cell sorting gen-activated protein kinase cascade. Involvement of protein analysis. We appreciate the excellent assistance of kinase A, rac1, and reactive oxygen species. Journal of Biological Françoise Chantoux in cell culture. Chemistry 2000 275 40539–40546. 17 Vandeput F, Perpete S, Coulonval K, Lamy F & Dumont JE. Role of the different mitogen-activated protein kinase subfamilies in the stimulation of dog and human thyroid epithelial cell proliferation References by cyclic adenosine 50-monophosphate and growth factors. Endo- crinology 2003 144 1341–1349. 1 Grana X & Premkumar R. Cell cycle control in mammalian 18 Kyriakis JM & Avruch J. Mammalian mitogen-activated protein cells: role of cyclins, cyclin dependent kinases (CDKs), growth kinase signal transduction pathways activated by stress and suppressor genes and cyclin-dependent kinase inhibitors (CKIs). inflammation. Physiological Reviews 2001 81 807–869. Oncogene 1995 11 211–219. 19 Ono K & Han J. The p38 signal transduction pathway: activation 2 Lees E. Cyclin dependent kinase regulation. Current Opinion in Cell and function. Cellular Signalling 2000 12 1–13. Biology 1995 7 773–780. 20 Nebreda AR & Porras A. p38 MAP kinases: beyond the stress 3 Cheng M, Olivier P, Diehl JA, Fero M, Roussel MF, Roberts JM & response. Trends in Biochemical Sciences 2000 25 257–260. Sherr CJ. The p21(Cip1) and p27(Kip1) CDK ‘inhibitors’ are 21 Cuenda A, Rouse J, Doza YN, Meier R, Cohen P, Gallagher TF, essential activators of cyclin D-dependent kinases in murine Young PR & Lee JC. SB203580 is a specific inhibitor of a MAP fibroblasts. EMBO Journal 1999 18 1571–1583. kinase homologue which is stimulated by cellular stresses and 4 Medina DL & Santisteban P. Thyrotropin-dependent proliferation interleukin-1. FEBS Letters 1995 364 229–233. of in vitro rat thyroid cell systems. European Journal of Endocrin- 22 Maher P. p38 mitogen-activated protein kinase activation is ology 2000 143 161–178. required for fibroblast growth factor-2-stimulated cell prolifer- 5 Kimura T, Van Keymeulen A, Golstein J, Fusco A, Dumont JE & ation but not differentiation. Journal of Biological Chemistry Roger PP. Regulation of thyroid cell proliferation by TSH and 1999 274 17491–17498. other factors: a critical evaluation of in vitro models. Endocrine 23 Fernandes DJ, Ravenhall CE, Harris T, Tran T, Vlahos R & Reviews 2001 22 631–656. Stewart AG. Contribution of the p38MAPK signalling pathway to 6 Lewis A, Fikaris AJ, Prendergast GV & Meinkoth JL. Thyrotropin proliferation in human cultured airway smooth muscle cells is and serum regulate thyroid cell proliferation through differential mitogen-specific. British Journal of Pharmacology 2004 142 effects on p27 expression and localization. Molecular Endocrin- 1182–1190. ology 2004 18 2321–2332. 24 Rausch O & Marshall CJ. Cooperation of p38 and extracellular 7 Cass LA & Meinkoth JL. Differential effects of cyclic adenosine signal-regulated kinase mitogen-activated protein kinase path- 30,50-monophosphate on p70 ribosomal S6 kinase. Endocrinology ways during granulocyte colony-stimulating factor-induced 1998 139 1191–1998. hemopoietic cell proliferation. Journal of Biological Chemistry 8 Dremier S, Pohl V, Poteet-Smith C, Roger PP, Corbin J, 1999 274 4096–4105. Doskeland SO, Dumont JE & Maenhaut C. Activation of cyclic 25 Liu RY, Fan C, Liu G, Olashaw NE & Zuckerman KS. Activation AMP-dependent kinase is required but may not be sufficient to of p38 mitogen-activated protein kinase is required for tumor mimic cyclic AMP-dependent DNA synthesis and thyroglobulin necrosis factor-alpha-supported proliferation of leukemia and

www.eje-online.org

Downloaded from Bioscientifica.com at 09/30/2021 11:59:15AM via free access EUROPEAN JOURNAL OF ENDOCRINOLOGY (2005) 153 p38 MAPK in FRTL-5 thyroid cell cycle regulation 133

lymphoma cell lines. Journal of Biological Chemistry 2000 275 38 Shklyaev SS, Namba H, Mitsutake N, Alipov G, Nagayama Y, 21086–21093. Maeda S, Ohtsuru A, Tsubouchi H & Yamashita S. Transient 26 Yamaguchi H, Igarashi M, Hirata A, Susa S, Ohnuma H, activation of c-Jun NH2-terminal kinase by growth factors influ- Tominaga M, Daimon M & Kato T. Platelet-derived growth ences survival but not apoptosis of human thyrocytes. Thyroid factor BB-induced p38 mitogen-activated protein kinase 2001 11 629–636. activation causes cell growth, but not apoptosis, in vascular 39 Wang XS, Diener K, Manthey CL, Wang S, Rozenzweig B, Bray J, smooth muscle cells. Endocrine Journal 2001 48 433–442. Delaney J, Cole CN, Chan-Hui PY, Mantlo N, Lichenstein HS, 27 Masamune A, Satoh M, Kikuta K, Sakai Y, Satoh A & Zukowski M & Yao Z. Molecular cloning and characterization of Shimosegawa T. Inhibition of p38 mitogen-activated protein a novel p38 mitogen-activated protein kinase. Journal of Biological kinase blocks activation of rat pancreatic stellate cells. Journal of Chemistry 1997 272 23668–23674. Pharmacology and Experimental Therapeutics 2003 304 8–14. 40 Moore JD, Yang JY, Truant R & Kornbluth S. Nuclear import of 28 Craxton A, Shu G, Graves JD, Saklatvala J, Krebs EG & Clark EA. Cdk/cyclin complexes: identification of distinct mechanisms for p38 MAPK is required for CD40-induced gene expression and pro- import of Cdk2/cyclin E and Cdc2/cyclin B1. Journal of Cell Biology liferation in B lymphocytes. Journal of Immunology 1998 161 1999 144 213–224. 3225–3236. 41 Jackman M, Kubota Y, Den Elzen N, Hagting A & Pines J. Cyclin A- 29 Neve RM, Holbro T & Hynes NE. Distinct roles for phosphoinosi- and cyclin E-Cdk complexes shuttle between the nucleus and the tide 3-kinase, mitogen-activated protein kinase and p38 MAPK cytoplasm. Molecular Biology of the Cell 2002 13 1030–1045. in mediating cell cycle progression of breast cancer cells. Oncogene 42 Baptist M, Lamy F, Gannon J, Hunt T, Dumont JE & Roger PP. 2002 21 4567–4576. Expression and subcellular localization of CDK2 and cdc2 kinases 30 Recio JA & Merlino G. Hepatocyte growth factor/scatter factor and their common partner cyclin A in thyroid epithelial cells: activates proliferation in melanoma cells through p38 MAPK, comparison of cyclic AMP-dependent and -independent cell ATF-2 and cyclin D1. Oncogene 2002 21 1000–1008. cycles. Journal of Cellular Physiology 1996 166 256–273. 31 Halawani D, Mondeh R, Stanton LA & Beier F. p38 MAP kinase 43 Keenan SM, Bellone C & Baldassare JJ. Cyclin-dependent kinase 2 signaling is necessary for rat chondrosarcoma cell proliferation. nucleocytoplasmic translocation is regulated by extracellular Oncogene 2004 23 3726–3731. regulated kinase. Journal of Biological Chemistry 2001 276 32 Pome´rance M, Carapau D, Chantoux F, Mockey M, Corre`ze C, 22404–22409. Francon J & Blondeau JP. CCAAT/enhancer-binding protein-hom- 44 Blanchard DA, Mouhamad S, Auffredou MT, Pesty A, Bertoglio J, ologous protein expression and transcriptional activity are regu- Leca G & Vazquez A. Cdk2 associates with MAP kinase in vivo and lated by 30,50-cyclic adenosine monophosphate in thyroid cells. its nuclear translocation is dependent on MAP kinase activation Molecular Endocrinology 2003 17 2283–2294. in IL-2-dependent Kit 225 T lymphocytes. Oncogene 2000 19 33 Lali FV, Hunt AE, Turner SJ & Foxwell BMJ. The pyridinyl imida- 4184–4189. zole inhibitor SB203580 blocks phosphoinositide-dependent pro- 45 Yang ES & Burnstein KL. Vitamin D inhibits G1 to S progression tein kinase activity, protein kinase B phosphorylation, and in LNCaP prostate cancer cells through p27Kip1 stabilization and retinoblastoma hyperphosphorylation in interleukin-2-stimulated Cdk2 mislocalization to the cytoplasm. Journal of Biological Chem- T cells independently of p38 mitogen-activated protein kinase. istry 2003 278 46862–46868. Journal of Biological Chemistry 2000 275 7395–7402. 46 Hiromura K, Pippin JW, Blonski MJ, Roberts JM & Shankland SJ. 34 Davies SP, Reddy H, Caivano M & Cohen P. Specificity and mech- The subcellular localization of cyclin dependent kinase 2 deter- anism of action of some commonly used protein kinase inhibitors. mines the fate of mesangial cells: role in apoptosis and prolifer- Biochemical Journal 2000 351 95–105. ation. Oncogene 2002 21 1750–1758. 35 Iijima Y, Laser M, Shiraishi H, Willey CD, Sundaravadivel B, Xu L, 47 Chiariello M, Gomez E & Gutkind JS. Regulation of cyclin-depen- McDermott PJ & Kuppuswamy D. c-Raf/MEK/ERK pathway dent kinase (Cdk) 2 Thr-160 phosphorylation and activity by controls protein kinase C-mediated p70S6K activation in adult mitogen-activated protein kinase in late G1 phase. Biochemical cardiac muscle cells. Journal of Biological Chemistry 2002 277 Journal 2000 349 869–876. 23065–23075. 48 Lents NH, Keenan SM, Bellone C & Baldassare JJ. Stimulation of 36 Whitmarsh AJ, Yang SH, Su MS, Sharrocks AD & Davis RJ. Role of the Raf/MEK/ERK cascade is necessary and sufficient for acti- p38 and JNK mitogen-activated protein kinases in the activation vation and Thr-160 phosphorylation of a nuclear-targeted of ternary complex factors. Molecular and Cellular Biology 1997 17 CDK2. Journal of Biological Chemistry 2002 277 47469–47475. 2360–2371. 37 Clerk A & Sugden PH. The p38 MAPK inhibitor, SB203580, inhibits cardiac stress-activated protein kinases/c-Jun N-terminal Received 25 January 2005 kinases (SAPKs/JNKs). FEBS Letters 1998 426 93–96. Accepted 31 March 2005

www.eje-online.org

Downloaded from Bioscientifica.com at 09/30/2021 11:59:15AM via free access