INTRODUCTION

Top Abstract Introduction Materials & Methods Results Discussion References

LEFT VENTRICULAR HYPERTROPHY (LVH) is a compensatory mechanism triggered in response to an increased working pressure of the heart, e.g., as a result of hemodynamic overload associated with hypertension. Although the process of LVH initially is compensatory, the adaptive growth in the size of cardiomyocytes leads to contractile failure if the increase in LV pressure persists. LVH remains one of the major causes of mortality and morbidity in humans (17, 29). Although the molecular and biochemical mechanisms responsible for the hypertrophic growth of cardiomyocytes have been investigated intensively (4), many of the molecules involved in the process of LVH after pressure overload still remain undetermined. However, recent studies from our laboratory (12, 13) have suggested that cell cycle regulatory molecules may be implicated in this process.

Normal cellular growth is controlled by the sequential formation, activation, and subsequent inactivation of a series of cyclin/cyclin-dependent kinase (CDK) complexes (16, 18). The complexes formed by CDK4 or CDK6 and three D-type cyclins (D1, D2, and D3) have been implicated in the control of G₁ phase progression of the cell cycle, and CDK2 associated with cyclins E and A has been implicated in the G₁/S and S/G₂ phase transitions, respectively. (16, 18). Mammalian cardiomyocytes proliferate actively during fetal development and grow by both hyperplasia (cell division) and hypertrophy (increase in cell size). However, the ability of cardiomyocytes to divide ceases completely soon after birth, with all subsequent growth of the heart being due to cardiac hypertrophy (see reviews, Refs. 14 and 22). Furthermore, chronic stimulation of mature, adult cardiomyocytes, e.g., as a consequence of hypertension or increased cardiac workload, results only in hypertrophic growth of myocytes rather than any increase in myocyte cell number. Currently, it is unclear which molecular switch controls cardiomyocyte proliferation and hypertrophy. However, an identification of the cell cycle regulatory molecules involved in hypertrophic growth would extend our understanding of the cell cycle machinery within adult cardiomyocytes and could provide an opportunity for developing therapeutic strategies for improving the prognosis of cardiac hypertrophy.

Recently, we have reported (12, 13) a transient, but significant, downregulation of the CDK inhibitors (CDKIs) p21 and p27 during the development of pressure overload-induced LVH. We postulated that a transient downregulation of these CDKIs soon after the imposition of pressure overload may result in the activation of G₁/S phase cyclins and CDKs, subsequently leading to G₁ phase progression and hypertrophic growth of cardiomyocytes. To test this hypothesis, we have examined the protein expression of cyclins (A, E, D1, D2, and D3) and CDKs (CDK2, CDK4, CDK5, and CDK6) in LV tissues and in isolated cardiomyocytes during the development of pressure overload-induced LVH in rats. We also have determined the in vitro kinase activities of CDK2, CDK4, and CDK6 and the formation of cyclin-CDK complexes in cardiomyocytes freshly isolated at different time points following the imposition of pressure overload. Finally, we have investigated the cell cycle profile of cardiomyocyte nuclei during the development of hypertrophy using fluorescence-activated cell sorter (FACS) analysis. Our results provide evidence for the first time that certain G₁ cyclins and CDKs are upregulated during the early stages of pressure overload-induced hypertrophy, which may contribute to the adaptive growth of cardiomyocytes during LVH in rats.

	MATERIALS AND METHODS

Top Abstract Introduction Materials & Methods Results Discussion References

Materials. [-³²P]ATP (3,000 Ci/mmol) and enhanced chemiluminescence (ECL) Western blotting reagents were purchased from Amersham International (Amersham, UK). DMEM with Glutamax was obtained from GIBCO (Paisley, Scotland). FCS was obtained from Globepharm (Esher, UK). Collagenase CLS1 was purchased from Worthington (Freehold, NJ). BSA (type V), pepsin, and propidium iodide (PI) were purchased from Sigma Chemical (Poole, UK). Rabbit polyclonal antibodies against cyclins A, E, and D1-3 and CDK2, CDK4, CDK5, and CDK6; the corresponding peptides to the various cyclins and CDKs; goat anti-rabbit IgG horseradish peroxidase-conjugated antibody; and glutathione-S-transferase-retinoblastoma protein (GST-pRb) fusion protein were obtained from Santa Cruz Biotechnology (Santa Cruz, CA). Goat polyclonal antibody to cardiac troponin I was a gift from Dr. P. Cummins, University of Birmingham (Birmingham, UK). Histone H1 substrate protein was purchased from Boehringer Mannheim (Mannheim, Germany), and protein A Sepharose beads were obtained from Pharmacia Biotech (Uppsala, Sweden). All other chemicals used were of the highest grade available commercially.

Cell culture. Murine NIH/3T3 fibroblasts were maintained in DMEM containing 10% FCS. All cultures were maintained in a humidified atmosphere containing 95% O₂-5% CO₂ at 37°C.

Animals. Adult male Wistar rats (starting weight: 166 ± 3 g) were obtained from Binton and Kingman (Hull, UK). Animals were killed by an approved method in accordance with the Home Office Guidance on the Operation of Animals (Scientific Procedures) Act 1986.

Induction of LVH and cardiac tissue preparation. Chronic pressure overload was produced in rats by subtotal suprarenal constriction of the abdominal aorta as described previously (9). Briefly, under anesthesia with Hypnorm (0.3 ml/kg ip) and diazepam (2.5 mg/kg ip), the aorta was carefully exposed through an abdominal incision. A uniform degree of constriction (0.45 mm in diameter) was achieved by fixing a titanium clip around the aorta (11). Age-matched sham-operated (SH) rats underwent the same operation but without aortic constriction (AC). Both AC and SH rats were housed and fed under identical conditions. Seven rats from each group were killed at days 1, 3, 7, 14, 21, and 42 after surgery. In addition, seven rats that did not undergo any operation were used as day 0 controls. Body weight was recorded from each rat on both the day of operation and the day of death, when the heart was excised and weighed. The atria were carefully dissected from the ventricles, and LV tissue was separated from right ventricular (RV) tissue and weighed. Heart tissue was frozen immediately in liquid N₂ until required for analysis. For myocyte isolation, six rats from AC or SH groups were killed at days 7, 14, and 28, and six rats that did not undergo any operation were used as day 0 controls.

Cardiomyocyte isolation. Myocytes from LV tissues were isolated according to a modification of an established method described previously (25, 26, 32). Briefly, AC and SH rats were anesthetized with ether, and the heart was rapidly excised and perfused in a retrograde manner through the aorta (37°C) using the Langendorff technique. The heart was first perfused for 5 min with Tyrode solution (137 mM NaCl, 5.4 mM KCl, 1.8 mM CaCl₂, 0.5 mM MgCl₂, 10 mM HEPES, and 10 mM glucose, adjusted to pH 7.4 at 34°C with NaOH) and then with nominally Ca²⁺-free Tyrode solution (135 mM NaCl, 5.4 mM KCl, 0.33 mM NaH₂PO₄, 1.0 mM MgCl₂, 10 mM HEPES, and 10 mM glucose, adjusted to pH 7.2 at 34°C with NaOH) for 5 min, followed by perfusion with nominally Ca²⁺-free Tyrode solution containing 0.05% collagenase CLS1 for 10 min. The atria and great vessels were removed, and the LV was separated from the RV. LV tissue was chopped into small pieces in wash buffer (1:1 DMEM-Ca²⁺-free Tyrode solution with 10% FCS) and agitated gently to facilitate cell dispersion. The cell suspension then was filtered through a 200-µm nylon mesh gauze, and the myocytes were separated from the nonmyocyte population by gravity sedimentation through a cushion of 6% BSA for 20 min. The myocyte pellet was then resuspended in ice-cold PBS and washed four times with PBS. This method has been shown previously (25, 26) to yield a myocyte-rich fraction containing >95% myocytes. Trypan blue exclusion showed >98% viability of myocytes, and immunostaining with antibody to cardiac troponin I showed that >98% of cells were positive for cardiac troponin I. Myocytes were used immediately for protein preparation.

Protein extraction and immunoblotting. Protein samples were prepared from both LV tissues and freshly isolated cardiomyocytes of AC and SH hearts. Samples prepared from LV tissue (200 mg/ml) were first homogenized with the use of an Ystral C10/25 homogenizer in an ice-cold extraction buffer containing 0.05 M Tris, 0.15 M NaCl, 2 mM EDTA, 2 mM phenylmethylsulfonyl fluoride (PMSF), 2 µg/ml leupeptin, and 2 µg/ml aprotinin, pH 7.4. Cells were lysed directly in ice-cold extraction buffer. After sonication of cells on ice, Triton X-100 was added to a final concentration of 0.5%. The lysate was extracted on ice for 15 min and then centrifuged at 12,000 g for 20 min. Protein concentrations in sample preparations were determined according to the method of Bradford (1) using BSA as a standard. Immunoblotting was performed exactly as described previously (2). Briefly, proteins (40 µg) were separated with 12% SDS-PAGE and transferred to nitrocellulose membranes. Membranes were blocked with 5% nonfat milk in PBS-0.2% Tween 20, followed by incubation with different antibodies diluted in PBS containing 1% nonfat milk for 1 h at room temperature. After membranes were washed with PBS-0.2% Tween 20, they were incubated with appropriate secondary antibodies conjugated to horseradish peroxidase for 1 h at room temperature. Bands were visualized by ECL and quantified using laser densitometry. Each protein sample was analyzed twice, and at least three different samples from AC and SH hearts were examined at each time point postoperation.

Immunoprecipitation and CDK activity assays. The following method is a modification of the method of Draetta et al. (5). Briefly, freshly isolated cardiomyocytes were lysed in ice-cold CDK lysis buffer containing 0.05 M Tris · HCl (pH 7.4), 0.25 M NaCl, 0.1% vol/vol Nonidet P-40, 0.005 M EDTA, 0.05 M NaF, 0.001 M Na₃VO₄, 0.001 M Na₄P₂O₇ · 10 H₂O, 0.01 M benzamidine, 50 µg/ml PMSF, 10 µg/ml N-tosyl-L-phenylalanine chloromethyl ketone, 10 µg/ml soybean trypsin inhibitor, 1 µg/ml aprotinin, and 1 µg/ml leupeptin. Lysates were sonicated and incubated on ice for 15 min and spun at 12,000 rpm (4°C) for 20 min. Myocyte protein samples (250 µg in a final volume of 750 µl) were immunoprecipitated with rabbit polyclonal antibodies to CDK2, CDK4, and CDK6 (1 µg/sample as determined by titration) coupled to protein A Sepharose beads. Immunocomplex-bound beads were washed four times with CDK lysis buffer before kinase activity assays. For the CDK2 kinase assay, beads were washed once more in wash buffer containing 0.05 M Tris · HCl (pH 7.4), 0.01 M MgCl₂, and 1 mM dithiothreitol (DTT), resuspended in 20 µl of wash buffer containing 125 µg/ml histone H1 substrate protein, and incubated for 5 min at 30°C. Five microliters of ATP buffer (10 mM ATP and 1 µCi/µl [-³²P]ATP in wash buffer) were then added to each tube, the tubes were incubated for 10 min at 30°C, and the reaction was terminated with the addition of 25 µl 2× Laemmli buffer. For CDK4 and CDK6 kinase activity assays, beads were washed once more in 1 ml of retinoblastoma (Rb) buffer containing 20 mM Na -glycerophosphate (pH 7.3), 15 mM MgCl₂, 5 µg/ml leupeptin, 1 mM benzamidine, 0.5 mM PMSF, 0.1 mM Na₃VO₄, and 1 mM DTT. Pelleted beads were resuspended in 30 µl Rb buffer containing 50 mM ATP, 10 µCi [-³²P]ATP, and 0.5 µg GST-pRb fusion protein per sample. The reaction was incubated at 30°C for 1 h and terminated with the addition of 30 µl 2× Laemmli buffer. Samples were boiled for 3 min, and proteins were separated with 12% SDS-PAGE. Gels were stained with Coomassie brilliant blue and dried before exposure to photographic film overnight at 70°C. For quantitative analysis of CDK complex activity, the resultant autoradiographs were scanned using laser densitometry. Six myocyte samples prepared from six different hearts were examined for each group at each time point postoperation.

Cell cycle analysis by FACS. Isolated cardiomyocytes (2 × 10⁶ cells per sample) were fixed in ice-cold 70% ethanol and stored at 4°C before analysis. Adult cardiomyocytes are rod-shaped cells that do not pass evenly through the filters of a FACS machine; furthermore, the majority of myocytes are binucleated or multinucleated, which leads to difficulties in measuring cell cycle profiles in terms of DNA content per cell. To overcome these problems, we isolated myocyte nuclei and performed cell cycle analysis on isolated myocyte nuclei. Fixed cells were pelleted and resuspended in 0.2 M HCl and 1 mg/ml pepsin and were incubated at 37°C for 12 min to digest sarcolemmal membranes and release intact nuclei. PBS was then added to stop the digestion. Nuclei were pelleted and washed twice in PBS. DNA staining solution (19) (1 ml) was then added (50 µg/ml Isoton, 200 µg/ml PI, 1 U/ml RNase) and mixed carefully. Nuclei were incubated at room temperature for a further 30 min. Labeled nuclei were analyzed for PI staining using a FACS Calibur analyzer (Becton-Dickinson).

Statistical analysis. Results are presented as means ± SE and were obtained from three different protein samples for immunoblotting and six different protein samples for in vitro kinase assays at each time point, and each sample was analyzed in duplicate. For FACS analysis, results were obtained from six different samples at each time point postoperation. Mean values were analyzed using one-way ANOVA, and if a significant difference was observed, then each AC value was compared with the SH value at the same time point using the Bonferonni t-test. Values for which P < 0.05 were considered significant.

	RESULTS

Top Abstract Introduction Materials & Methods Results Discussion References

Effect of AC on ventricular mass. Seven rats from either AC or SH groups were killed at 1, 3, 7, 14, 21, and 42 days after surgery. Body weight and ventricular weight were measured (Table 1). Relative to age-matched SH controls, LV weight in AC rats increased progressively with time (P < 0.01), whereas no significant difference in body weight or RV weight was observed between the two groups. The LV-to-body weight ratio (LV/body weight) in AC rats increased rapidly from day 1 to day 21 postoperation (from 12 to 50%) compared with LV/body weight in SH controls. After 21 days, LV mass in AC rats stabilized such that no further increase in LV/body weight was observed up to 6 wk postoperation.

Expression of G₁/S-phase cyclins and CDKs in LV during development of LVH. To determine the changes in protein expression of G₁/S phase cyclin and CDKs during LVH, we measured the protein levels of cyclins (A, D1, D2, D3, and E) and CDKs (CDK2, CDK4, CDK5, and CDK6) in LV tissues from AC and SH rats by immunoblotting. Equal protein loading in each experiment was confirmed by Coomassie blue staining of the gels and by probing a duplicate transfer membrane with a goat polyclonal antibody to cardiac troponin I. Each experiment was repeated twice to confirm the reproducibility of the results. Proteins extracted from the NIH/3T3 mouse fibroblast cell line after serum stimulation were used as a positive control for the expression of cyclins and CDK proteins. Cyclins A, D1, and E were undetectable by immunoblotting in LV tissues from both SH and AC hearts but were expressed in NIH/3T3 cells (3), whereas cyclins D2 and D3 and CDK2, CDK4, CDK5, and CDK6 were expressed differentially in LV tissues from both treatment groups. A representative experiment demonstrating these results is shown in Fig. 1. Thus the protein levels of cyclins D2 and D3 and of CDK4 and CDK6 detected in LV tissue obtained from SH rats showed no significant difference throughout the day 1-42 postoperative period, whereas expression of these proteins in LV tissues obtained from AC rats increased significantly by days 1 or 3 postoperation but returned to SH levels by day 21 postoperation. CDK2 protein was detected in LV tissues from both AC and SH rats, although levels in both groups increased by day 1 postoperation and then decreased progressively from day 3 postoperation. However, although the levels of CDK2 protein appeared higher in LV tissues from AC rats compared with levels expressed in SH rats, they were not significant. CDK5 protein levels showed a constant level of expression with no significant difference between SH and AC hearts. It was interesting to note that a significant increase in protein levels of CDK2 and CDK5 was observed in SH animals compared with levels in day 0 (without operation) controls. The precise reason for this upregulation in expression of these two CDKs following sham operation remains unknown at this time, although a number of factors could be responsible. One such possibility is that trauma experienced by SH and AC animals postoperation causes changes in neural and/or humoral factors that may induce subsequent changes in specific gene and/or protein expression in the heart.

View larger version (37K): [in this window] [in a new window]   Fig. 1.   Detection of cyclins and cyclin-dependent kinases (CDKs) during development of left ventricular hypertrophy (LVH) by immunoblotting. Equal amounts of protein samples (40 µg) obtained from LV tissues of rats with aortic constriction (AC) or sham operation (SH) at different time points postoperation were separated by 12% SDS-PAGE and transferred to nitrocellulose. Filters were probed with antibodies against different cyclins and CDKs as indicated. NIH, positive control (NIH/3T3 fibroblast cell line).

View larger version (27K): [in this window] [in a new window]   Fig. 2.   Changes in protein expression of cyclins D2 (A) and D3 (B) and of CDK4 (C) and CDK6 (D) postoperation. Three different protein samples from AC (filled bars) and SH (open bars) at each time point postoperation were examined twice by immunoblotting and scanned densitometrically. Results were normalized to expression of the same protein in normal day 0 samples (without operation) applied onto every gel and expressed as a ratio to normal control. Values are means ± SE. * P < 0.03 compared with day 0 and SH controls.

Changes in kinase activities of CDK2, CDK4, and CDK6 in myocytes during development of LVH. The observed increase in the protein expression of cyclins D2 and D3 and of CDK4 and CDK6 in LV tissues during the development of LVH does not necessarily imply that such complexes are active in cardiomyocytes. To determine whether differences in the activities of cyclin-CDK complexes occurred in AC cardiomyocytes postoperation, we isolated cardiomyocytes from LV tissues of AC and SH rats at days 0, 7, 14, and 28 postoperation. Protein lysates prepared from purified cardiomyocytes were used to measure the in vitro kinase activities of immunoprecipitated CDK2, CDK4, and CDK6. The specificity of each kinase assay was confirmed by preincubation of immunoprecipitating antibodies with corresponding peptides to CDK2, CDK4, and CDK6 before immunoprecipitation, which abolished completely the phosphorylation bands of histone H1 for CDK2 and the phosphorylation bands of GST-pRb for CDK4 and CDK6. In addition, negative controls, with protein lysate but no antibody, failed to show any phosphorylation bands of either histone H1 (CDK2) or the GST-pRb (CDK4 and CDK6) substrate proteins, thereby confirming the specificity of the kinase assays. CDK2, CDK4, and CDK6 immunoprecipitated from protein lysates of NIH/3T3 cells were used as positive controls in each experiment. Figure 3 shows a representative autoradiograph of histone H1 phosphorylation by CDK2 and GST-pRb fusion protein phosphorylation by CDK4 and CDK6. In cardiomyocytes isolated from normal rats at day 0, the kinase activities of CDK4 and CDK6 were very low in accordance with our previous findings in mature adult myocytes (3). Compared with kinase activities in day 0 controls, the kinase activities of CDK2, CDK4, and CDK6 were slightly, but not significantly, increased in cardiomyocytes isolated from SH hearts at days 7 and 14 postoperation. However, the kinase activities of CDK4 and CDK6 were upregulated significantly in cardiomyocytes isolated from AC hearts at day 7 compared with kinase activities in age-matched SH controls. The kinase activity for CDK6 was seen to return to SH control levels by day 14 postoperation, whereas CDK4 remained at an elevated level until day 14 after AC (Fig. 3) and returned to SH levels by day 28 postoperation (Fig. 4). To compare quantitatively the kinase activities of the specific cyclin-CDK complexes at different time points following AC or sham operation, the autoradiographs were scanned densitometrically and the results normalized to the density of the day 0 samples (expressed as a densitometry index of 1 in Fig. 4). The results from SH and AC myocytes were then expressed as ratios of the day 0 control value (Fig. 4). Compared with kinase activities in SH controls, kinase activities of CDK4 and CDK6 were upregulated significantly in cardiomyocytes isolated from AC hearts such that CDK4 activity was increased by 67.8% at day 7 and by 50.6% at day 14 postoperation, and CDK6 activity was increased by 72.3% at day 7 postoperation (P < 0.03). By day 28 postoperation, the levels of CDK4 and CDK6 kinase activities in cardiomyocytes isolated from AC hearts had returned to the levels observed in SH controls. No significant difference was observed for CDK2 kinase activity between myocytes isolated from SH and AC hearts up to 28 days postoperation.

View larger version (35K): [in this window] [in a new window]   Fig. 3.   Representative autoradiograph showing in vitro kinase activities of CDK2, CDK4, and CDK6 in AC and SH cardiomyocytes postoperation. Equal amounts of protein (250 µg) obtained from cardiomyocytes freshly isolated from LV tissues of AC and SH rats were immunoprecipitated with antibodies to CDK2, CDK4, and CDK6 as indicated. In vitro kinase activities of histone H1 phosphorylation by CDK2 and retinoblastoma protein (pRb) phosphorylation by CDK4 and CDK6 were examined as described in MATERIALS AND METHODS. NIH/3T3, positive control (NIH/3T3 fibroblast cell line); NEG, negative control (protein lysate, no antibody); d0, day 0 control.

View larger version (18K): [in this window] [in a new window]   Fig. 4.   Quantitative changes in kinase activities of CDK2 (A), CDK4 (B), and CDK6 (C) in AC and SH cardiomyocytes postoperation. Six different protein samples from myocytes freshly isolated from LV tissues of AC (filled bars) and SH rats (open bars) at days 7, 14, and 28 postoperation were examined for in vitro kinase activities of CDK2, CDK4, and CDK6 as described in MATERIALS AND METHODS. Autoradiographs were scanned densitometrically and results normalized to normal day 0 samples, expressed as a densitometry index value of 1. Results from SH and AC myocytes were expressed as ratios of normal control. Values are means ± SE. * P < 0.03 compared with day 0 and SH controls.

Changes in levels of cyclin D2/D3-CDK4/CDK6 complex formation in myocytes during development of LVH. Both CDK4 and CDK6 are D-type cyclin-associated kinases, and the formation of such complexes is an absolute requirement for kinase activity. Because we have observed that the protein levels of cyclins D2 and D3 and of CDK4 and CDK6 were upregulated transiently during the development of LVH (Fig. 2), we wanted to confirm whether there was concomitant increase in the formation of CDK4 or CDK6 complexes with cyclins D2 and/or D3 during myocyte hypertrophic growth. Accordingly, CDK4 and CDK6 proteins were immunoprecipitated directly from myocytes freshly isolated from hearts at days 0, 7, 14, and 28 after AC and separated by SDS-PAGE as described in MATERIALS AND METHODS. After the transfer of separated protein onto nitrocellulose, the filters were subsequently immunoblotted with antibodies to cyclins D2 or D3 to determine the levels of cyclin-CDK complexes during the development of LVH. In addition, cyclins D2 and D3 were immunoprecipitated from freshly isolated myocytes, and resultant immunoblots were probed with CDK4 and CDK6 to determine whether these proteins were coimmunoprecipitated in a complex with cyclin D2 or D3. The specificity of the various antibodies for their respective coimmunoprecipitated proteins in each experiment was confirmed by competitive inhibition experiments with the corresponding peptides used to raise the respective cyclin and CDK antibodies. In addition, the molecular weight of each coimmunoprecipitated protein band was compared with the same protein band identified in lysates obtained from proliferating NIH/3T3 cells, which was run on the same gel. Figure 5 shows representative immunoblots of proteins coimmunoprecipitated with cyclins D2 and D3, CDK4, and CDK6. CDK4 associates with both cyclin D2 and cyclin D3 at all time points measured. Compared with levels in day 0 samples, the levels of CDK4-cyclin D2 or CDK-cyclin D3 complexes were significantly increased at day 7 and then subsequently decreased with time, returning to day 0 levels by day 28 after AC. Interestingly, the CDK6-cyclin D2 and CDK6-cyclin D3 complexes formed showed different profiles during hypertrophy such that the levels of the CDK6-cyclin D2 complex increased from day 7 to day 14 and then returned to day 0 control levels by day 28 after AC. In contrast, the CDK6-cyclin D3 complex was already detected at a relatively high level in day 0 control samples, and this was shown to decrease progressively with time after day 7 after AC (Fig. 5). We noticed that levels of cyclin D3-CDK6 complexes at day 7 after AC were different depending on which protein was immunoprecipitated. If CDK6 was immunoprecipitated and detected with antibody to cyclin D3, the levels of cyclin D3 coimmunoprecipitating with CDK6 decreased significantly from day 7 after AC compared with levels in day 0 controls, whereas, if cyclin D3 was immunoprecipitated and detected with antibody to CDK6, the levels of CDK6 coimmunoprecipitating with cyclin D3 in samples at day 7 after AC showed an increase compared with levels in day 0 controls. A possible explanation for this discrepancy could be that the affinity and specificity of antibodies to cyclin D3 and CDK6 were different. Although the anti-cyclin D3 antibody supplied by Santa Cruz showed no cross-reaction with cyclin D2 by immunoblotting, we cannot exclude the possibility that, during immunoprecipitation, some intact cyclin D2, which was found at high levels by day 7 after AC, might immunoprecipitate with cyclin D3. Thus intact cyclin D2 and cyclin D3 complexes were formed with CDK4 and CDK6 during the hypertrophic growth of cardiomyocytes concomitant with the increase in the protein expression and activities of these molecules after the imposition of pressure overload.

View larger version (41K): [in this window] [in a new window]   Fig. 5.   Detection of cyclin D2/D3-CDK4/CDK6 complex formation during hypertrophic growth of cardiomyocytes. Equal amounts of protein (250 µg) obtained from cardiomyocytes freshly isolated from LV tissues of AC rats at days 0, 7, 14, and 28 were immunoprecipitated with antibodies to CDK4, CDK6, cyclin D2, and cyclin D3 as indicated on top of autoradiographs. Immunoprecipitated proteins (IP) were separated by 12% SDS-PAGE and transferred to nitrocellulose. Resultant filters were probed with antibodies to cyclins D2 and D3 and to CDK4 and CDK6 as indicated at left of autoradiographs.

Changes in cell cycle profile of cardiomyocytes during development of hypertrophy. Because cyclins D2 and D3 and CDK4 and CDK6 were significantly upregulated concomitant with the hypertrophic growth of cardiomyocytes after AC, we considered whether these changes in cell cycle regulatory molecule expression would affect the cell cycle profile of cardiomyocytes during the development of hypertrophy. Accordingly, we measured DNA content and cell cycle profile in myocyte nuclei by FACS analysis using a PI-staining technique (19, 20, 30). Figure 6 shows cell cycle profiles of adult cardiomyocyte nuclei obtained at day 7 after AC or sham operation. Compared with nuclei in day 0 or SH controls, a significant proportion of cardiomyocyte nuclei isolated from AC hearts was shown to progress from G₀/G₁ [DNA content 2n (cell ploidy)] into the G₂/M (DNA content 4n) phase of the cell cycle by day 7 postoperation. As shown in Fig. 6 and Table 2, 83.51 ± 2.37% of myocyte nuclei were found in G₀/G₁ with 13.97 ± 1.94% in G₂/M in day 0 control cells. By day 7 after AC, the percentage of myocyte nuclei in G₀/G₁ had decreased to 76.97 ± 0.74%, whereas the percentage of nuclei in G₂/M had increased to 22.18 ± 1.06% (P < 0.01). After day 14 of operation, the percentages of myocyte nuclei from AC or SH hearts distributed between the G₀/G₁ and G₂/M phases remained constant. In accordance with our protein and CDK activity data, changes in cell cycle profiles were only detected in myocyte nuclei prepared from the LV tissues of AC rats postoperation. Cell cycle profile of myocyte nuclei prepared from RV tissues showed no significant difference between AC and SH groups.

View larger version (13K): [in this window] [in a new window]   Fig. 6.   Cell cycle profiles of cardiomyocyte nuclei at day 0 (A) and at day 7 after sham operation (B) and AC (C). Cardiomyocyte nuclei were isolated enzymatically from LV tissues of AC and SH rats and labeled with propidium iodide as described in MATERIALS AND METHODS. DNA content in nuclei (arbitrary units) was determined using fluorescence-activated cell sorter (FACS) analysis. DNA profiles were obtained from FACS analyzer, and data were analyzed for number of nuclei in each phase (G1 and G2) of cell cycle using CellQuest software (Becton-Dickinson). Cell number is represented on a scale from 0 to 10,000 events, with G1 peak reading between 7,500 and 8,000 depending on treatment (see Table 2).

	DISCUSSION

Top Abstract Introduction Materials & Methods Results Discussion References

Recently, we have reported a transient downregulation of the CDKI molecules p21 and p27 during the early stages of pressure overload-induced LVH (12, 13). p21 and p27 are universal CDKIs that play an important role in arresting cells in the G₀/G₁ phase of the cell cycle (28, 31). As a result of our previous studies, we proposed that downregulation of p21 and p27 in cardiomyocytes during hypertrophic growth may lead to a concomitant upregulation in the expression and activity of certain cyclins and CDKs, thereby contributing to the hypertrophic response. Results from our present study are supportive of this hypothesis, because cyclins D2 and D3 and CDK4 and CDK6 were significantly upregulated at both the protein and activity levels in cardiomyocytes from day 7 to day 14 after AC, the period when LV/body weight was increasing rapidly (Table 1). It is possible that the significant increase in the kinase activities of CDK4 and CDK6 during the development of hypertrophy is due to the combination of the downregulation of the CDKIs p21 and p27 (12, 13) and the upregulation in the protein and activity levels of certain G₁ cyclin-CDK complexes. In support of our observations and hypothesis, Sadoshima et al. (23) have recently reported a transient upregulation in the protein expression of CDK4 and CDK6 during the hypertrophic response of cultured neonatal cardiomyocytes after angiotensin stimulation. It is well established that CDK4 and CDK6 bind only to the D-type cyclins and that the kinase activities of the resultant complexes are most closely linked to the control of G₁ phase progression and cell size control (24). With a series of coimmunoprecipitation experiments, we demonstrated that CDK4 associated with cyclins D2 and D3 and that the levels of these complexes increased significantly by day 7 after AC. We also showed an increase in the levels of complexes formed by cyclin D2 and CDK6 from day 7 to day 14 after AC, thereby suggesting that these G₁ phase cyclins and CDKs are involved in regulating the growth potential of cardiomyocytes during the development of LVH. We also observed a significant difference in the time course of the complexes formed by CDK6 with cyclin D2 or with cyclin D3 following AC such that CDK6-cyclin D2 complexes were upregulated at day 7 and day 14 after AC, which correlated with the adaptive growth of the myocytes, whereas CDK6-cyclin D3 complexes were present at high levels in day 0 hearts and were downregulated with time after AC. These results suggest that CDK6-cyclin D3 complexes might not be involved in the hypertrophic growth of cardiomyocytes during pressure overload-induced LVH.

Cyclin D1 is an important and major partner of both CDK4 and CDK6 in many cell types. However, we were unable to detect cyclin D1 in either whole LV tissues or in isolated cardiomyocytes of both SH and AC hearts. Cyclin D1 has been reported (10, 27) to be expressed differentially compared with cyclin D3 in response to serum stimulation and during terminal differentiation of skeletal muscle cells. These investigators reported that cyclin D1 mRNA was upregulated in response to serum stimulation and then downregulated during skeletal muscle differentiation. In contrast, cyclin D3 mRNA remained at constant levels in the presence or absence of serum stimulation. Interestingly, cultured neonatal cardiomyocytes also showed a different profile of cyclin D1 protein expression on stimulation with angiotensin II or serum (23). Thus, after 9 h of stimulation with angiotensin II, the expression of cyclin D1 protein in cultured neonatal cardiomyocytes was downregulated, whereas levels were upregulated in response to mitogenic stimulation with serum compared with control levels (23). In a previous study (3) relating to cell cycle regulatory molecule expression during normal rat cardiomyocyte development, we reported a differential protein expression between cyclin D1 and cyclins D2/D3. Thus, in rat cardiac muscle cells, cyclin D1 was detectable only at very low levels in fetal myocytes but was undetectable in neonatal and adult cells, whereas cyclins D2 and D3 were detectable at each stage of cardiomyocyte development, although levels decreased progressively with increasing age. Thus differential expression of cyclin D1 and cyclins D2/D3 in different cell types suggests different functions of these closely related D-type cyclins during muscle differentiation.

Cyclins E and A are known to associate with CDK2, and the resultant CDK2 kinase activity increases at the G₁/S boundary and in early S phase, respectively (6-8). During this study, we found that the protein levels of CDK2 were low in normal and SH cardiomyocytes, whereas CDK2 protein levels and histone H1 kinase activity were slightly, but not significantly, upregulated during the early stages following pressure overload. Because CDK2 activity was measurable in adult cardiomyocytes, it seemed likely that a regulatory subunit of the kinase complex, such as cyclin A or cyclin E, should be expressed in adult cardiomyocytes. However, cyclin A and cyclin E protein levels were undetectable in either LV tissues or in myocytes isolated from adult rat hearts. The cyclins are a family of proteins that in general demonstrate oscillating levels during the cell cycle (15). It has been shown that in human and rat cardiac tissues, cyclin A is downregulated to a very low level during terminal differentiation, and this has been postulated to be responsible for the permanent withdrawal of cardiomyocytes from the cell cycle (33). It is possible that, at the time points used in our study, the protein levels of cyclins A and E in cardiomyocytes examined are very low and cannot be detected by immunoblotting. However, because we were able to detect a low level of CDK2 kinase activity in cardiomyocytes, and because the kinase activity assay probably is more sensitive than immunoblotting, we propose that low levels of cyclin E and/or cyclin A probably are present within myocytes to enable CDK2 to become activated.

The DNA content of cardiomyocyte nuclei is known to alter significantly during cardiac development such that the percentage of nuclei displaying 2n DNA (G₁) and 4n DNA (G₂/M) varies between fetal and adult myocytes (22). In a recent study, we have shown that the number of S phase cells diminishes progressively from ~15% in fetal cells to <1% in mature adult myocytes (21). Concomitant with the loss of S phase cells, we have observed a significant increase in the percentage of myocytes arresting in G₂/M such that ~15% of adult myocyte nuclei are found in this phase of the cycle. Previous studies have shown that during hypertrophic growth an increase in DNA content (from 2n to 4n) occurs simultaneously with a significant increase in RNA and protein synthesis (22). However, in adult animals, even a maximum degree of ventricular hypertrophy is not accompanied by mitotic division of the myocyte nuclei (22). In the present study, we used FACS analysis to determine DNA content in myocyte nuclei labeled with PI and demonstrated that the percentage of G₂ phase myocyte nuclei (4n DNA content) increased from ~14% at day 0 to ~22% at day 7 after AC, with SH myocyte nuclei not differing significantly from day 0 control values. Thus the number of myocyte nuclei arresting in G₂ in hypertrophied hearts increases significantly by ~57%, indicating the cell cycle progression of myocyte nuclei from G₀/G₁ into G₂/M phase of the cell cycle during the development of LVH.

In summary, we have shown, for the first time, a significant upregulation of G₁ phase D-type cyclins (D2 and D3) and CDKs (CDK4 and CDK6) during the development of pressure overload-induced LVH in rats. We also have provided direct evidence for cell cycle progression of myocyte nuclei from G₀/G₁ into the G₂/M phase of the cell cycle during hypertrophic growth. Thus cell cycle-dependent molecules are involved in the development of cardiomyocyte hypertrophy such that the transient, but significant, downregulation in p21 and p27 CDKIs (12, 13) is accompanied by a concomitant upregulation in the expression and activities of cyclin D2/D3-CDK4/CDK6 complexes. Although our results implicate cell cycle regulatory molecules in cardiac hypertrophy, we are unable at the present time to determine whether changes in the expression and activities of these molecules are causal or occur as a consequence of hypertrophic growth. Future experiments designed to induce forced expression of these molecules in cardiomyocytes should offer an approach to determine the precise role of cell cycle regulatory molecules in cardiomyocyte hypertrophy and may lead to strategies for improving the prognosis of this disease.