INTRODUCTION

Top Abstract Introduction Materials & Methods Results Discussion References

IN PROLIFERATING CELLS, during the G₁ phase of the cell cycle, proliferative stimuli lead to the sequential activation of G₁ cyclins and associated cyclin-dependent kinases (CDKs), including D-type cyclins (D1, D2, and D3) complexed to cdk4 and cdk6 and cyclin E bound to cdk2 (21, 26). In addition to regulation by cyclin binding, CDK activity can also be regulated by interactions with the members of two families of CDK inhibitors: the p16^INK4a family proteins, which specifically inhibit the activity of cdk4/cdk6, and the p21^CIP1/WAF1/p27^KIP1 proteins, which inhibit the activities of cdk2 and cdk4/cdk6 (15, 22). G₁ CDKs are thought to target various cellular substrates, such as the retinoblastoma (Rb) gene product (pRb) family, phosphorylation of which is required for cell cycle progression (21, 26).

Cardiomyocytes undergo terminal differentiation soon after birth, irreversibly withdrawing from the cell cycle. Because of their inability to proliferate, postnatal cardiomyocytes grow by hypertrophy (25, 27). In adults, cardiac hypertrophy occurs in response to various stresses, such as increased afterload and ischemia, or as a mechanism to compensate for heart dysfunction. Elucidation of the mechanisms underlying cardiac hypertrophy is an important issue in cardiology, because severe heart dysfunction and arrythmias may occur as a result of hypertrophy. Cardiac hypertrophy, which is accompanied by the upregulation of protein synthesis and expression of fetal isoforms of sarcomeric proteins, such as skeletal -actin (SK-A) and -myosin heavy chain, is thought to be mediated by autocrine/paracrine mechanisms involving various growth and neurohumoral factors (6, 19). Also, atrial natriuretic factor (ANF) mRNA is considered to be a maker of cardiac hypertrophy (2). Serum (11, 12), basic fibroblast growth factor (16), insulin-like growth factors (7), endothelins (8), and ANG II (1) have been reported to induce cardiomyocyte hypertrophy in vitro. These stimuli activate multiple second messenger systems and induce various immediate-early genes, such as c-fos and c-jun, in cardiomyocytes (3, 18, 23). In view of similar responses being also observed in response to mitogenic stimuli in various cell types, it appears likely that hypertrophic and mitogenic stimuli share certain intracellular responses. Furthermore, recent experiments have indicated that serum stimulation upregulates the activities of G₁ cyclins and associated CDKs in terminally differentiated cardiomyocytes, similar to the observation in proliferating cells (17). We thus investigated whether G₁ CDK activity is involved in the process of hypertrophy in cultured rat neonatal cardiomyocytes.

	MATERIALS AND METHODS

Top Abstract Introduction Materials & Methods Results Discussion References

Production of recombinant adenoviruses. The replicate-deficient adenoviruses were prepared as described previously (14, 24). Briefly, rat p21^cip1/waf1 (Axp21) or human p16^INK4 (Axp16) cDNA was inserted into the cassetted-cosmid vector (pAdex1w) containing E₁- and E₃-deleted adenovirus sequences with the CAG (chicken -actin promotor + cytomegalovirus enhancer) promoter and rabbit -globin poly(A) signal sequences. The recombinant viruses were obtained by in vitro recombination in 293 cells. After separation of cell debris by centrifugation, the supernatant-containing virus particles were stored at 80°C until use. The titer of each viral stock was determined by plaque assay in 293 cells, and the titers consistently ranged between 10⁹ and 10¹¹ plaque-forming units/ml. Adenovirus without coding sequences (Ax1w1) was used as a control.

Cell culture and infection. Cardiomyocytes from 1- or 2-day postnatal Sprague-Dawley rats were isolated, subjected to Percoll gradient centrifugation, and cultured in vitro, as described previously (20); we routinely obtained cultures in which >95% of the cells were cardiomyocytes, as assessed by immunostaining with the mouse monoclonal antisarcomeric actin antibody (Dakopatts). Neonatal rat cardiomyocytes in culture were incubated in MEM (Flow Laboratories) with 5% FCS (Flow Laboratories) for 24 h. On the 2nd day, cardiomyocytes were infected with recombinant adenoviruses (50 plaque-forming units/cell) and further incubated with MEM without FCS for 48 h. Infected cells were analyzed after treatment with 10% FCS at 37°C for various periods. Nonmyocytes were prepared as controls, as described previously (4).

Bromodeoxyuridine incorporation. Cardiomyocytes/nonmyocytes that entered the S phase in response to serum stimulation were counted after incubation with bromodeoxyuridine (BrdU, 10 µmol/l) in the presence or absence of 10% FCS for 48 h at 37°C. To distinguish myocytes from nonmyocytes, cultures were first labeled with antisarcomeric actin antibody and then with 5 µg/ml of mouse monoclonal anti-BrdU antibody (Chemicon International). The distribution of labeled nuclei was visualized using an alkaline phosphatase-conjugated secondary antibody in conjunction with an ABC kit (Vector). The BrdU-labeling index is expressed as the number of BrdU-positive nuclei per 100 cells, with the total number of BrdU-labeled cells being the product of the labeling index and the mean cell number.

Northern blot analysis. Northern blot analysis and preparation of the probes were performed as previously described (2, 8).

Assay of cdk4 kinase. The nuclear extract of cardiomyocytes was immunoprecipitated with 1 µg of cdk4 polyclonal antibody (C-22, Santa Cruz Biotechnology), then the immunocomplexes were tested for cdk4 kinase activity using the glutathione S-transferase (GST)-Rb fusion protein as the substrate by a method described previously (13). In a control experiment, 1 µg of rat IgG was used for immunoprecipitation.

Evaluation of cardiac hypertrophy. To measure cell surface area, cells were fixed with 10% buffered Formalin after incubation in the presence or absence of 10% FCS for 48 h, then labeled with sarcomeric actin antibody. The mean surface area of 50 sarcomeric actin-positive cells was calculated (8).

	RESULTS

Top Abstract Introduction Materials & Methods Results Discussion References

Lack of DNA synthesis in cultured cardiomyocytes. First, we evaluated the entry of quiescent rat primary neonatal cardiomyocytes into the S phase of the cell cycle by determining the extent of BrdU incorporation. Rat cardiac fibroblasts (nonmyocytes) were used as positive controls. As summarized in Table 1, stimulation of cardiomyocytes with 10% serum led to no marked increase in the number of BrdU-positive cells (<2% above background); i.e., cell proliferation was, in effect, halted, whereas a 49% increase (P < 0.01) was noted among the corresponding control nonmyocytes. These results indicate that under our experimental conditions the addition of serum cannot stimulate cardiomyocyte DNA synthesis. Although serum stimulation had no effect on cardiomyocyte proliferation, as evidenced by the lack of DNA synthesis, it did lead to cardiomyocyte hypertrophy, causing a twofold increase in cell size, as determined by cell surface area measurements, upregulation of protein synthesis, and induction of SK-A transcription (see below).

Activation of G₁ CDKs in serum-stimulated cardiomyocytes. In agreement with the results of recent experiments (17), we detected upregulation of the expression of the mRNAs of cyclin D1 as well as cyclin A, which accumulates around S phase commencement in serum-stimulated cardiomyocytes. Although we detected cyclin E mRNA expression in unstimulated cardiomyocytes, it was not upregulated by serum stimulation (Fig. 1). Consistent with the upregulation of cyclin D1 mRNA expression, serum-stimulated cdk4 kinase activity was also increased (Fig. 2), as determined by kinase assays in which the GST-Rb fusion protein was used as the substrate. The samples immunoprecipitated with IgG did not phosphorylate GST-Rb (data not shown).

View larger version (56K): [in this window] [in a new window]   Fig. 1.   Time course of cyclin A, D1, and E mRNA induced by 10% FCS in cardiomyocytes. Neonatal rat cardiomyocytes were exposed to 10% FCS for indicated times. Northern blot hybridization (10 µg of total RNA/lane) was performed using 32P-labeled human cyclin A and E and mouse cyclin D1 cDNA probes. 28S, ribosomal RNA bands stained with ethidium bromide. Exposure times were 120 h for cyclin A and D1 and 192 h for cyclin E.

View larger version (46K): [in this window] [in a new window]   Fig. 2.   Effect of gene transfer of p21 on cdk4 kinase activity of cardiomyocytes. Lysates of cultured cardiomyocytes with or without 10% FCS and Axp21 were immunoprecipitated with anti-cdk4 antibody. Immune complexes were assayed for retinoblastoma gene product (pRb) kinase activity.

G₁ CDK activity is involved in cardiomyocyte hypertrophy. To determine whether endogenous D-type CDK activity is directly involved in cardiomyocyte hypertrophy, the cells were infected with a recombinant adenovirus Axp16, encoding a specific inhibitor of D-type CDK activity, p16, and Axp21, encoding a broad CDK inhibitor, p21. Ax1w1 was used as the control virus. Infection of both CDK inhibitor-expressing viruses, but not of the control Ax1w1 virus, inhibited the increase in the level of DNA synthesis in serum-stimulated nonmyocytes (Table 1). In addition, after infection of cardiomyocytes with Axp21, cdk4 kinase activity was effectively inhibited (Fig. 2), confirming that the proteins encoded by these viruses inhibit such activity. From cell surface area measurements (Fig. 3), it was determined that, with the exception of the control Ax1w1 virus (data not shown), serum-stimulated cardiomyocytes treated with Axp21 or Axp16 showed a substantial reduction in the extent of hypertrophy in comparison with those not possessing Axp21 or Axp16. Morphologically, the cells treated with Axp21 or Axp16 alone were similar to the control cells. The results of measurements of [³H]leucine incorporation, indicating the rate of protein synthesis, are summarized in Fig. 4. The degree of protein synthesis stimulated by 10% serum was inhibited by conjunction with Axp16 or Axp21 in a dose-dependent manner. On the other hand, serum and serum with Ax1w1 had significantly different effects on unstimulated control cardiomyocytes, indicating no increase in the level of protein synthesis in the inhibitor-infected serum-stimulated cardiomyocytes. Finally, we asked whether p21 and p16 can inhibit expression of markers of cardiac hypertrophy, SK-A and ANF. Each virus effectively inhibited serum-induced SK-A and ANF mRNA expression, whereas the control virus (Ax1w1) did not (Fig. 5). These results indicate that p21 and p16 suppress the development of cardiomyocyte hypertrophy.

View larger version (105K): [in this window] [in a new window]   Fig. 3.   Inhibition of cardiomyocyte hypertrophy by overexpression of p21 and p16. Micrographs (original magnification ×100) show morphologies of infected or uninfected cardiomyocytes stained with antibody for sarcomeric actin. Calculated cell surface area is shown below each micrograph. Surface area of cardiomyocytes was assessed by image-analyzer system (mean ± SE).

View larger version (26K): [in this window] [in a new window]   Fig. 4.   Inhibition of protein synthesis in serum-stimulated cardiomyocytes by Axp21 and Axp16 in cardiomyocytes. Stimulated cells were infected with or without indicated viruses, Ax1w1 [50 multiplicity of infection (moi)], Axp21 (20-50 moi), and Axp16 (20-50 moi), and assayed for protein synthesis by [3H]leucine incorporation. Bars, means; error bars, SE (n = 12). * Significantly different from control (P < 0.05);  significantly different from serum alone (P < 0.05).

View larger version (65K): [in this window] [in a new window]   View larger version (65K): [in this window] [in a new window]   Fig. 5.   Inhibition of expression of serum-induced skeletal -actin (SK-A) and atrial natriuretic factor (ANF) mRNA by Axp21 and Axp16 in cardiomyocytes. Cardiomyocytes were infected with Axp21 (A) or Axp16 (B) and treated with 10% FCS for 6 h for SK-A and ANF mRNA expression analysis. Northern blot hybridization (10 µg total RNA/lane) was performed as described in Fig. 1 legend. Exposure times were 72 h for SK-A, 48 h for ANF, and 24 h for glyceraldehyde3-phosphate dehydrogenase (GAPDH).

	DISCUSSION

Top Abstract Introduction Materials & Methods Results Discussion References

Consistent with the recent report by Sadoshima and co-workers (17), we showed that D-type CDK activity, which plays a key role in cell proliferation in various cell types, is promoted during hypertrophic processes of serum-stimulated cardiomyocytes in culture. To investigate the role of endogenous D-type CDK activity in the development of cardiac hypertrophy, we inhibited the activity by using the p21- and p16-expressing viruses and demonstrated that these viruses inhibited a variety of markers of cardiomyocyte hypertrophy. These observations, together with the fact that D-type CDKs are common targets of p21 and p16, suggest that the D-type CDK activity is a prerequisite for cardiomyocyte hypertrophy and provide direct evidence of a new role for G₁ CDKs in terminally differentiated cells in addition to their known critical functions in proliferating cells. A number of experiments demonstrated that multiple second messenger systems, such as the Ras-mitogen-activated protein kinase and the Janus protein-signal transduction and activator of transcription pathways, are also activated in response to mitogenic stimuli in various cell types and play critical roles in development of cardiac hypertrophy. Our results indicate that D-type CDK activity also contributes to the upregulation of protein synthesis for cardiomyocyte hypertrophy.

We used serum stimulation as an inducer of cardiomyocyte hypertrophy, because serum is a commonly used cell cycle activator in a variety of experiments. One may be concerned that the inhibition of hypertrophy seen in these experiments may be secondary to the inhibition of proliferation of the nonmyocytes in the culture. Thus we have performed the same experiments using cardiomyocytes pretreated with cytosine arabinose to prevent the proliferation of nonmyocytes (data not shown) and obtained data that p16 and p21 effectively suppressed the hypertrophy. These data indicate that the inhibition of hypertrophy is not secondary to effects on nonmyocytes and that p21 and p16 directly inhibit hypertrophy of cardiomyocytes. Furthermore, in our preliminary study, hypertrophy of cardiomyocytes induced by several other factors, such as ANG II and endothelin-1, which are known to play an important role in cardiac hypertrophy but not to cause fibroblast proliferation potently, was also inhibited by Axp21 or Axp16 (T. Nozato, M. Tamamori, and H. Ito, unpublished observations).

The mechanisms underlying stimulation of hypertrophy by D-type CDK activity remain unclear. One simple explanation is that E2F activity is also involved in the induction of hypertrophy. However, it appears unlikely that E2F activity contributes to the upregulation of SK-A and other sarcomeric genes that are activated during the hypertrophy process, since E2F sites have not been found within the promoter sequences of these genes. In fact, overexpression of E2F1 causes the opposite effect, i.e., suppression of the SK-A promoter activity in cardiomyocytes (9). Moreover, adenovirus E1A oncoprotein also suppresses SK-A transcription in a manner dependent on E1A domains required for binding to the Rb family members (10). Taken together, although the possibility of the Rb family-E2F pathways contributing to protein synthesis remains to be determined, these observations suggest that, unlike the control of cell proliferation, the Rb family-E2F pathway appears not to be the main pathway involved in the hypertrophic process.

Alternatively, D-type cyclins/cdk4 may directly activate unknown factors that bridge intracellular cascades of cardiac hypertrohy. Recent evidence suggests that D-type CDKs phosphorylate substrates other than the Rb family members, such as DMP1, an myb-like transcription factor (5). This suggests the interesting possibility that G₁ CDKs target substrate(s) other than that involved in the Rb family-E2F pathway for promoting hypertrophic responses in cardiomyocytes. Further studies to define the target substrates for D-type CDK activity in relation to cardiac hypertrophy would clarify the differences between hypertrophic and proliferative processes and provide important insights into the pathogenesis of cardiac hypertrophy.

Despite recent advances in the therapy of heart failure, it is likely that the incidence of heart failure will continue to increase. The limitation of pharmacological therapy for heart failure is, in part, due to the inability of the myocardium to reproduce once injured by severe ischemia or inflammation of the heart. We have offered several new insights into the understanding of the cell cycle in cardiomyocytes that may ultimately contribute to the development of new therapeutic approaches aimed at recovering the reproductive capacity of the myocardium. Moreover, in regard to cardiac hypertrophy, we have shown novel evidence of the involvement of cell cycle regulators in the hypertrophic process. The results of our study may be applicable to the prevention or regulation of cardiac hypertrophy in the near future, possibly using cell cycle inhibitors that block the G₁-to-S transition. In conclusion, we believe that the results of our study open new avenues for the development of therapeutic strategies for the hypertrophied and failing heart.