Critical Role of Transcription Factor Cyclic AMP Response Element Modulator in β1-Adrenoceptor–Mediated Cardiac Dysfunction
Background— Chronic stimulation of the β1-adrenoceptor (β1AR) plays a crucial role in the pathogenesis of heart failure; however, underlying mechanisms remain to be elucidated. The regulation by transcription factors cAMP response element-binding protein (CREB) and cyclic AMP response element modulator (CREM) represents a fundamental mechanism of cyclic AMP–dependent gene control possibly implicated in β1AR-mediated cardiac deterioration.
Methods and Results— We studied the role of CREM in β1AR-mediated cardiac effects, comparing transgenic mice with heart-directed expression of β1AR in the absence and presence of functional CREM. CREM inactivation protected from cardiomyocyte hypertrophy, fibrosis, and left ventricular dysfunction in β1AR-overexpressing mice. Transcriptome and proteome analysis revealed a set of predicted CREB/CREM target genes including the cardiac ryanodine receptor, tropomyosin 1α, and cardiac α-actin as altered on the mRNA or protein level along with the improved phenotype in CREM-deficient β1AR-transgenic hearts.
Conclusions— The results imply the regulation of genes by CREM as an important mechanism of β1AR-induced cardiac damage in mice.
Received April 15, 2008; accepted October 22, 2008.
Human heart failure develops as a consequence of various cardiac diseases and is among the most frequent causes of mortality and morbidity in Western countries.1 Chronic stimulation of the β1-adrenoceptor (β1AR) by elevated plasma catecholamines and subsequent activation of the cyclic AMP (cAMP)–dependent signal transduction pathway play a crucial role in the progression of human heart failure. Accordingly, blockade of β1AR decreases mortality in human heart failure.2–4 Conversely, stimulation of the β1AR-mediated signaling pathway evokes cardiac hypertrophy, fibrosis, and dysfunction as well as increased mortality in various animal models including transgenic mice with heart-directed expression of the β1AR,5,6 the stimulatory GTP-binding protein Gsα,7,8 or protein kinase A.9 By contrast, the cardiac-directed expression of adenylyl cyclase type VI evokes protective effects in transgenic mice,10 which may be explained by the activation of signaling pathways other than cAMP.11 Alterations in the expression of various cardiac proteins such as junctin, a transmembrane protein in the sarcoendoplasmic reticulum,12,13 the uncoupling protein 2 (Ucp2), and the transcription factors Fhl114 or Nab115 were linked to the phenotype of β1AR-transgenic mice. However, until now, little was known about the underlying mechanisms bridging the gap from chronic β1AR stimulation or β1AR overexpression to expressional changes and successive cardiac dysfunction in the failing heart.
Clinical Perspective p 88
The regulation by transcription factors of the cAMP response element-binding protein (CREB)/cAMP response element modulator (CREM) family represents an important mechanism of cAMP-dependent gene control and may contribute to the altered cardiac gene regulation after chronic β1AR stimulation and in human heart failure.16,17 Different mouse models with gain or loss of CREM function suggested CREM as a key regulator of cardiac function implicated in multiple signaling pathways,18–20 and the inducible CREM isoform ICER (inducible cAMP early repressor) was identified as a regulator of apoptosis in cardiac myocytes.20a In the present study, we tested whether CREM is a critical mediator of β1AR-mediated detrimental cardiac effects. To test this hypothesis, we studied transgenic mice with heart-directed expression of the β1AR (β1ARTG) in the absence and presence of a functional Crem gene in comparison with wild-type (WT) and CREM-deficient mice serving as controls.
β1ARTG mice were described previously,6 and we thank Dr Günther Schütz, DKFZ Heidelberg, Germany, for providing Crem−/− mice.21 We crossed heterozygous β1ARTG mice (FVB/N strain) with Crem+/− mice (129Sv:C57Bl/6 genetic background). The resulting β1ARTG/Crem+/− intercrosses were then mated to Crem+/− mice (129Sv:C57Bl/6). All experiments were performed with WT mice, Crem-normal β1AR-transgenic mice (β1ARTG), non–β1AR-transgenic Crem-deficient mice (Crem−/−), and Crem-deficient β1AR-transgenic mice (β1ARTG/Crem−/−) aged 35 to 40 weeks (unless otherwise indicated) from the resultant F2–F6 generations in accordance with approved protocols of local animal welfare authorities. Experiments with F2 mice were in part repeated with mice of the F6 generation and showed comparable results. Thus, differences between groups were not due to inhomogeneities of the genetic background.
Four-micrometer paraffin sections of ventricular tissue were stained according to Masson-Goldner and with hematoxylin-eosin and analyzed by a pathologist in a blinded manner. Diameters of at least 100 transversely cut cells were determined per individual heart, and mean values of the diameters from 3 to 6 animals per group were compared.
Left Ventricular Catheterization
Mice were anesthetized, and left ventricular (LV) function was assessed as described previously.18 The β1AR agonist dobutamine was administered via the cannulated left jugular vein.
Isolation of Cardiomyocytes
Primary cardiac myocytes from adult mouse hearts were isolated as described previously.22–24 Isolated myocytes were homogenized by pipetting up and down in a lysis buffer containing 7 mol/L urea, 2 mol/L thiourea, 10 mmol/L Tris base, 5 mmol/L Mg acetate, and 4% CHAPS and titrated to pH 8.0. The lysate was centrifuged at 12 000g and 4°C for 1 hour, and the supernatant was stored at −80°C for further use in 2-dimensional (2D) differential gel electrophoresis (DIGE) and immunoblot experiments.
SDS-PAGE and Quantitative Immunoblotting
Preparation of ventricular homogenates, electrophoresis on polyacrylamide gels, transfer onto nitrocellulose membranes, and immunologic detection with the use of the ECL detection system (Amersham Biosciences, GE Healthcare, Piscataway, NJ) or 125I-labeled protein A followed previous descriptions.17–19,23 Annexin A4 was detected on Western blots with the use of a rabbit polyclonal anti-annexin A4 antibody (Abcam, Cambridge, UK; 1:2000). We thank Dr L.R. Jones, Indianapolis, Ind, for providing anti-Ryr2, anti-erca2a, anti-triadin, and anti-junctin antibodies.
Radioligand Binding Assay
Cardiac ventricular membranes were prepared and binding assays were performed with the use of the nonselective βAR ligand (+/−)[125I]cyanopindolol (ICYP; Amersham; 150 pmol/L) as described previously.18,19
Quantitative Real-Time Polymerase Chain Reaction
Total ventricular RNA was extracted with the use of the RNA Midiprep Kit (Qiagen, Hilden, Germany), and corresponding cDNAs were generated with the RevertAid First Strand cDNA Synthesis Kit (MBI Fermentas, St Leon-Rot, Germany). The quantitative real-time polymerase chain reaction (PCR) was performed with the LightCycler system following the manufacturer’s specifications with the use of LightCycler Fast Start DNA Master SYBR Green I Mix (Roche Diagnostics, Penzberg, Germany) and the following primers: Gapdh, 5′-ATGGCCTTCCGTGTTCCTAC-3′ (forward), 5′-GGCCCCTCCTGTTATTATGG-3′ (reverse); Fhl1, 5′-AGAAGTACGTGCAGAAGGATGG-3′ (forward), 5′-ACAGGATCTTGCCATCCTTGG-3′ (reverse); Ucp2, 5′-GCATTGGCCTCTACGACTCTG-3′ (forward), 5′-TTGGGAGAGGTCCCTTTCCA-3′ (reverse); Nab1, 5′-GGTGGTGATGATGTCCAGCA-3′ (forward), 5′-TGTTGGCGATCCCTCTGGTA-3′ (reverse); Thbs1, 5′-TCCCCAAAAGAGACAAAGG-3′ (forward), 5′-CACGTTCCTAGGAAAAAGG-3′ (reverse); Ryr2, 5′-GCCGGACATGAAGTGTGATG-3′ (forward), 5′-AGGTAGTTGGCCAGGTTGTG-3′ (reverse). Levels of particular mRNAs were determined with the help of LightCycler software version 3.5 with appropriate calibration curves obtained with different amounts of control cDNAs. Results are mean values of duplicate runs for each sample normalized to WT median value except values for Thbs1 mRNA, which were normalized to the median of the β1ARTG group because the transcript was not detected in WT hearts.
Continuous data are summarized as mean±SEM with statistical evaluation with the use of 1-way ANOVA and Student-Newman-Keuls post hoc testing for the analysis of multiple groups or with the use of the Student t test for the comparison of 2 groups. The results from the catheterization experiment were analyzed by 2-way ANOVA with repeated measures followed by Student-Newman-Keuls post hoc testing performed between groups at each dose of dobutamine applied. PCR data are shown as box plot graphs indicating percentiles. Here, statistical analysis was performed with the use of the nonparametric Kruskal-Wallis test and Dunnett T3 multiple comparison post hoc test not requiring equal variances. Post hoc tests were only performed if omnibus tests (ANOVA or Kruskal-Wallis procedure) revealed significant differences between groups. Additional methods are described in detail in the expanded Methods in the online-only Data Supplement.
The authors had full access to and take full responsibility for the integrity of the data. All authors have read and agree to the manuscript as written.
To assess a potential role of CREM in the long-term detrimental effects of β1AR overexpression, 4 mouse lines were generated and studied: WT mice, Crem-normal β1AR-transgenic mice (β1ARTG), Crem-deficient non–β1AR-transgenic mice (Crem−/−), and Crem-deficient β1AR-transgenic mice (β1ARTG/Crem−/−). No differences were found in mean body weight or in heart weight indexed to body weight between groups (not shown), in agreement with previous data from β1ARTG and Crem−/− mice.6,18 Microscopic evaluation revealed increased cardiomyocyte diameters in β1ARTG cardiac ventricles that were normalized in the absence of CREM in β1ARTG/Crem−/− mice (Figure 1A). Staining according to Masson-Goldner displayed widespread areas of focal intercellular fibrosis and cell disarray in β1ARTG compared with WT mice (Figure 1B). Inactivation of CREM in β1ARTG/Crem−/− mice resulted in a considerable reduction in the degree of fibrosis and in normalization of cellular morphology comparable to WT mice.
Invasive Hemodynamic Assessment of Cardiac LV Function
In vivo assessment of cardiac function was performed by LV catheterization both under basal conditions and after stimulation with increasing doses of the β1AR agonist dobutamine. No differences were found in basal or dobutamine-stimulated heart rates within groups (Figure 2 and Table I in the online-only Data Supplement). Hence, the interpretation of data is not complicated by changes in intrinsic heart rate. In the absence of dobutamine, β1ARTG, Crem−/−, and β1ARTG/Crem−/− mice showed reduced maximal LV pressure (LVP) and reduced maximal velocities of contraction (dP/dtmax) and relaxation (dP/dtmin) compared with WT controls, in agreement with previous data from β1ARTG and Crem−/− mice.6,18 The maximal effects of dobutamine on LVP, dP/dtmax, and dP/dtmin were not decreased in Crem−/− compared with WT mice (Figure 2). β1ARTG mice displayed significantly reduced effects of dobutamine on these parameters (Figure 2), reflecting the development of heart failure in this model. As a main result of this study, inactivation of CREM in β1ARTG/Crem−/− mice normalized impaired dobutamine-stimulated LVP, dP/dtmax, and dP/dtmin in β1ARTG to values not different from those in Crem−/− mice (Figure 2). Taken together, impaired LV response to dobutamine in β1ARTG mice was rescued by ablation of the Crem gene in β1ARTG/Crem−/− mice.
Cardiac Gene Expression of Regulatory Proteins
Several genes were described previously to be differentially expressed on the protein or the mRNA level in β1ARTG versus WT mice, including junctin,13 Fhl1,14 Ucp2,14 and Nab1.15 Therefore, we tested whether the improved phenotype of β1ARTG/Crem−/− mice is associated with a normalized expression of one of these genes compared with β1ARTG mice. In agreement with previous results, protein levels of junctin were reduced to 47% in β1ARTG versus WT mice (Table 1); however, differences between β1ARTG/Crem−/− versus β1ARTG groups were not statistically significant. Similarly, Fhl1 mRNA levels were increased in β1ARTG hearts, regardless of the presence or absence of CREM. Nonparametric ANOVA (Kruskal-Wallis test) indicated significant differences in Ucp2 or Nab1 mRNA levels between groups, and Ucp2 or Nab1 mRNA levels tended to be increased in β1ARTG versus WT hearts, in agreement with published results14,15; however, this effect did not reach significance in the post hoc test. Nevertheless, no statistically significant differences were found in Ucp2 or Nab1 mRNA levels that could explain the partial rescue in β1ARTG/Crem−/− versus β1ARTG groups (Figure 3A). Levels of other cardiac proteins, namely, the calcium-regulating proteins calsequestrin, triadin, phospholamban, and Serca2a, the Ca2+-ATPase of the sarcoplasmic reticulum, as well as protein levels of protein kinase A, Ca2+/calmodulin-dependent protein kinase II, and the CREM-related transcription factor CREB, were not statistically different between groups (Table 1). We performed radioligand binding studies to test whether the improved phenotype of β1ARTG/Crem−/− mice is due to a reduced density of βAR compared with β1ARTG mice. Specific binding of cardiac ventricular membranes to ICYP was not different between β1ARTG/Crem−/− and β1ARTG groups (150 pmol/L ICYP; n=3; in fmol/mg protein: WT, 99±19 [95% confidence interval, 40 to 227]; β1ARTG, 1683±336 [95% confidence interval, 670 to 3890]*; β1ARTG/Crem−/−, 1957±716 [95% confidence interval, 334 to 8650]*; 1-way ANOVA P=0.0004, *P<0.05 versus WT).
Oligonucleotide microarray analysis was performed to obtain the expressional profile associated with the improved phenotype in β1ARTG/Crem−/− mice. Twelve candidate genes met our inclusion criteria for statistically significant changes in expression (Table 2). All of these candidate genes were upregulated in β1ARTG/Crem−/− versus β1ARTG, and no gene was downregulated in β1ARTG/Crem−/− versus β1ARTG mice, meeting the statistical inclusion criteria. Five candidate genes are predicted target genes of CREM-related transcription factor CREB (Table 2).25 Selected candidate genes were validated by real-time PCR to address possible statistical limitations of the microarray analysis. One upregulated gene, thrombospondin-1 (Thbs1), is implied in cardioprotection by limiting fibrosis in infarcted mouse heart,26 and a functionally active cAMP response element was described in the mouse Thbs1 gene promoter,25,27 suggesting Thbs1 as a direct cardioprotective target gene of CREM. As revealed by real-time PCR, Thbs1 mRNA was increased 75-fold in β1ARTG/Crem−/− hearts compared with the levels in β1ARTG hearts (Figure 3A). Thrombospondin-1 protein was immunologically detected in sections from both β1ARTG/Crem−/− and β1ARTG hearts (Figure I in the online-only Data Supplement) and showed a prevailing interstitial localization in β1ARTG/Crem−/− myocardium, whereas it was localized mainly in sites of focal fibrosis in β1ARTG hearts. The upregulation of another gene with potential cardioprotective properties, Sfrp1 (secreted frizzled related protein 1), was also confirmed by real-time PCR (n=8 to 10; data not shown); however, changes in Sfrp1 mRNA in β1ARTG/Crem−/− versus β1ARTG hearts were not accompanied by significant changes on the protein level in cardiac ventricular homogenates (n=8 to 10; data not shown). Because Sfrp1 improves cardiac hypertrophy via glycogen synthase kinase 3β (GSK-3β),28,29 we further studied both expression and phosphorylation of GSK-3β and did not observe any difference between groups (n=7; data not shown). We further analyzed mRNA levels of the cardiac ryanodine receptor (Ryr2) because microarray analysis suggested a possible (insignificant) upregulation of this functionally important cardiac gene in β1ARTG/Crem−/− hearts and because predicted nonclassic CREs were described in the Ryr2 gene promoter.25 Ryr2 mRNA was increased 3-fold in β1ARTG/Crem−/− versus β1ARTG hearts (Figure 3A), and Ryr2 protein was upregulated 1.65-fold in the β1ARTG/Crem−/− compared with the β1ARTG group (Figure 3B). Moreover, we tested whether protein levels of Ryr2 are altered in β1ARTG versus WT cardiac ventricular homogenates. However, no difference was found between the groups (Ryr2 signals standardized to calsequestrin signals; WT, n=5, 100±27%; β1ARTG, n=5, 78±18%).
For a deeper screening for expressional changes between β1ARTG/Crem−/− versus β1ARTG, ventricular homogenates underwent 2D-DIGE (Table II and Figure II in the online-only Data Supplement) and subsequent mass spectrometry. The experiment provided evidence for protein changes with folds between 1.50 and 2.18. Thirteen proteins were indicated as upregulated in β1ARTG/Crem−/− versus β1ARTG and 1 downregulated (Table II in the online-only Data Supplement). The 14 proteins could be assigned confidently by matrix-assisted laser desorption/ionization time-of-flight (MALDI-TOF) peptide mapping, and they were verified by sequence analysis on the basis of post–source decay multiplex analysis or manual nanoelectrospray ion trap fragmentation. One of these proteins, annexin A4 (Anxa4), a putative target gene of CREB,25 was upregulated 1.7-fold in ventricular homogenates from β1ARTG/Crem−/− versus β1ARTG hearts (Figure 3B). Because Anxa4 is expressed in both cardiomyocytes and noncardiomyocytes, we tested whether Anxa4 is upregulated in primary myocytes isolated from β1ARTG/Crem−/− versus β1ARTG hearts; however, no difference between groups was found (Figure 3C). Therefore, it is conceivable that Anxa4 is upregulated in β1ARTG/Crem−/− versus β1ARTG noncardiomyocytes because of the global CREM inactivation not confined to myocytes.
To specifically screen for expressional changes in cardiac myocytes and to control the confounding influence of developing heart failure due to the β1AR transgene, we isolated primary cardiac ventricular myocytes from both “young” (aged 11.2±1.5 weeks, before the secondary development of heart failure6) and “old” (aged 38.3±1.6 weeks, with heart failure phenotype6) β1ARTG and β1ARTG/Crem−/− mice and repeated the 2D-DIGE with 3 replicates in each of the 4 groups (Table 3). Sixty-one and 76 spots were differentially expressed in old β1ARTG/Crem−/− versus old β1ARTG cardiac myocytes and in young β1ARTG/Crem−/− versus young β1ARTG cardiomyocytes, respectively. Within these sets of 61 and 76 spots, 24 spots were identical, and all of these 24 spots were regulated in the same direction in old β1ARTG/Crem−/− versus β1ARTG and in young β1ARTG/Crem−/−versus β1ARTG cardiac myocytes. Hence, these proteins are differentially expressed as a consequence of CREM inactivation in β1ARTG/Crem−/− versus β1ARTG cardiac myocytes, independently of age and the confounding influence of the heart failure phenotype. We have analyzed 13 of these 24 spots as well as 16 additional spots regulated solely in old or in young β1ARTG/Crem−/− versus β1ARTG cardiac myocytes by MALDI-TOF peptide mapping and/or mass spectrometry using manual nano–mass spectrometry/mass spectrometry or liquid chromatography/mass spectrometry/mass spectrometry. Of these overall 29 spots, 4 spots did not produce analyzable spectra. The spots numbered 2531, 3318, and 3342 were assigned to proliferin-related protein precursor and α-2,8-sialyltransferase 8E by MALDI-TOF peptide mapping, respectively. The remaining 22 spots were confidently assigned by MALDI-TOF, and 21 of these 25 spots were verified subsequently by quadrupole TOF sequence analysis, as documented in Table V in the online-only Data Supplement. Among the verified spots, 4 and 6 spots were identified as variants, modifications, or both of tropomyosin 1α and cardiac actin, respectively, which are both predicted target genes of CREB and which are both regulated in old and young β1ARTG/Crem−/− versus β1ARTG cardiac myocytes. Additional proteins regulated in both old and young β1ARTG/Crem−/−versus β1ARTG cardiac myocytes are troponin I, ATP synthase-γ, malate dehydrogenase, and NADH dehydrogenase [ubiquinone] 1 beta subcomplex subunit 10.
Our data show that (1) CREM inactivation protects β1ARTG mice from cardiomyocyte hypertrophy, fibrosis, and impaired LV response to βAR stimulation; (2) the partially rescued phenotype in β1ARTG/Crem−/− mice is accompanied by the altered expression of cardiac proteins, including the cardiac ryanodine receptor as well as the contractile proteins tropomyosin 1α and cardiac α-actin; and (3) this rescue is not due to a normalized expression of junctin, Fhl1, Ucp2, or Nab1, genes previously described to be differentially expressed in β1ARTG versus WT mice. These results indicate that CREM is required for the development of detrimental β1AR-mediated cardiac effects and imply the regulation of cardiac genes via CREM as a relevant mechanism of cardiac dysfunction after chronic β1AR stimulation.
The data presented here clearly demonstrate that Crem−/− mice are more resistant to harmful effects evoked by transgenic overexpression of the β1AR. This suggests CREM as a fundamental link between chronic β1AR stimulation and consecutive effects, namely, cardiac hypertrophy, fibrosis, and dysfunction. Importantly, the partial rescue in β1ARTG/Crem−/− mice was not caused by a reduced expression of transgenic β1AR as a consequence of CREM inactivation because radioligand binding assays revealed equal βAR densities in β1ARTG/Crem−/− versus β1ARTG mice. As a main result of this study, inactivation of CREM fully restored impaired dobutamine-stimulated LV performance in β1ARTG mice. This indicates that CREM is critical for β1AR-mediated deterioration of cardiac function and for decreased cardiac responsiveness to βAR stimulation, an important feature of the failing heart. As shown previously,18 the lack of CREM itself leads to reduced LV function under basal conditions. Thus, inactivation of CREM in β1ARTG/Crem−/− mice restored impaired dobutamine-stimulated LV function to the level of Crem−/− mice, but the reduced basal LV function in Crem−/− hearts was not improved by the concomitant expression of the β1AR. Furthermore, the restoration of cardiomyocyte hypertrophy and the reduction in cardiac fibrosis by inactivation of CREM in β1ARTG mice suggest that CREM represents a critical positive regulator of β1AR-mediated hypertrophy in the heart and that CREM is involved in pathways regulating cardiac fibrosis. This is in agreement with previous reports from related genetic mouse models. Transgenic mice with heart-directed expression of CREM-IbΔC-X, a human cardiac CREM isoform generating transcriptional repressors,30 developed ventricular cardiomyocyte hypertrophy,19 whereas 2 independent Crem−/− mouse models displayed normal cardiac morphology under basal conditions.18,20
Consistent results from transgenic mice with heart-directed expression of CREM-IbΔC-X19 and from Crem−/− mice18 suggested that CREM regulates LV function via an upregulation of Serca2a. In contrast, reduced basal LV function in Crem−/− mice was not combined with a downregulation of Serca2a in the present study. This discrepancy may either be due to differences in age (15 to 25 weeks in the previous study, 35 to 40 weeks in the present study) or due to different genetic backgrounds of mice. Examination of additional Crem−/− and WT mice aged 16 to 20 weeks and with the original genetic background (129Sv:C57Bl/6) confirmed our previous result of a downregulation of Serca2a (n=7 to 8; G.L., S.P.R., W.S., F.U.M., unpublished data, 2004). Moreover, LV dysfunction was associated with unchanged Serca2a protein levels in another Crem−/− mouse model.20 Hence, Crem−/− hearts concordantly display LV dysfunction in the absence of β1AR stimulation, but the contribution of a downregulation of Serca2a is not fully understood.
Junctin, Ucp2, Fhl1, and Nab1 genes were previously reported to be differentially expressed in β1ARTG versus WT hearts, but their expression was not different in β1ARTG/Crem−/− versus β1ARTG mice. Thus, these genes are not relevant for the improved phenotype of β1ARTG/Crem−/−, and it is not likely that CREM contributes to their altered expression in β1ARTG hearts. Because their expression was not conormalized along with reduced fibrosis, restored hypertrophy, and dobutamine-stimulated LV function in β1ARTG/Crem−/− hearts, they are either not relevant for the phenotype of β1ARTG mice or they are functionally dominated by other genes differentially expressed as a consequence of CREM inactivation. Transcriptome and proteome analysis filtered out candidate genes differentially expressed in CREM-deficient versus CREM-normal β1ARTG cardiac ventricular homogenates. The majority of these genes were upregulated in β1ARTG/Crem−/− compared with β1ARTG hearts, and among the candidate genes, no predicted CREB/CREM target gene was downregulated in β1ARTG/Crem−/− versus β1ARTG hearts. This suggests that cardiac CREM isoforms, not discriminating between cardiomyocytes and noncardiomyocytes, predominantly act as transcriptional repressors. The precise cardiac role of most of these candidate genes is not known except the cardiac ryanodine receptor (Ryr2) representing a central regulator of calcium homeostasis in cardiac myocytes31 and thrombospondin-1 and procollagens associated with the extracellular matrix. Myocardial Ryr2 is expressed exclusively in cardiomyocytes32 and mediates the calcium release from the sarcoplasmic reticulum in cardiomyocytes. It therefore plays a crucial role in cardiac excitation-contraction coupling.33 Thus, the upregulation of Ryr2 may at least in part explain the increased dobutamine-stimulated LV function in β1ARTG/Crem−/− mice.34 Thrombospondin-1 (Thbs1) is expressed in cardiomyocytes from cardiac allografts35 and is strongly upregulated after cardiac injury, whereas it is not detectable in normal adult myocardium,36 in agreement with the observation that Thbs1 mRNA was not detectable in WT hearts. It was suggested that Thbs1 limits cardiac fibrosis and preserves cardiac matrix integrity after pathological events.26,36 Hence, an upregulation of Thbs1 in cardiac myocytes may in principle contribute to beneficial effects of CREM inactivation and to reduced cardiac fibrosis in β1ARTG/Crem−/− hearts. However, the altered predominant localization of Thbs1 in β1ARTG/Crem−/− versus β1ARTG cardiac ventricular sections was not combined with increased Thbs1 protein levels in ventricular homogenates (n=3 to 4; data not shown). It is conceivable that a local upregulation of Thbs1 was not detectable on immunoblots because of an enrichment of Thbs1 in areas of local fibrosis that were reduced in β1ARTG/Crem−/− versus β1ARTG hearts, but this discrepancy raises the question of whether Thbs1 is of relevance for the improved phenotype in β1ARTG/Crem−/− mice. Annexin A4 (Anxa4), a predicted target gene of CREB,25 was increased on the protein level in ventricular homogenates from β1ARTG/Crem−/− versus β1ARTG hearts, but it was not altered in isolated cardiomyocytes. This suggests that Anxa4 is upregulated in noncardiomyocytes as a consequence of the global inactivation of CREM and that the expression of Anxa4 in cardiomyocytes is not regulated by CREM. Anxa4 is reportedly expressed in cardiac myocytes and upregulated in human failing hearts.37 In fibroblasts and epithelial cells, it has an inhibitory effect on a Ca2+-dependent chlorine current by inhibiting calcium/calmodulin-dependent protein kinase II–ion channel interaction and preventing phosphorylation of the channel.38 The relevance of the latter findings for the partial rescue in β1ARTG/Crem−/− versus β1ARTG hearts is not clear.
The proteome analysis in isolated cardiomyocytes from old and young β1ARTG/Crem−/− and β1ARTG mice has overcome important limitations of the 2D-DIGE experiment in ventricular homogenates from old mice, in particular the lack of specificity in regard to cardiac myocytes and the lack of control of the confounding influence of age and the phenotype on the gene regulation. Hence, the proteins regulated to be parallel in old and young cardiomyocytes are altered in the specific cell type affected by the β1AR transgene and reflect a β1AR:CREM interaction independent of the confounding influence of the heart failure phenotype. Among the regulated proteins, there are 2 predicted target genes of CREB and CREM, tropomyosin 1α and cardiac α-actin, which are both upregulated as a result of CREM inactivation in 3 of 4 and 3 of 6 spots (including the spots regulated solely in old or young cardiomyocytes), respectively. This is in agreement with the hypothesis that cardiac CREM isoforms mainly act as repressors. However, the 2D-DIGE analysis in cardiomyocytes also revealed a number of proteins downregulated due to CREM inactivation in β1ARTG/Crem−/− versus β1ARTG groups. It remains to be elucidated whether these spots represent splice variants and/or posttranslational modifications and whether activating CREM isoforms, eg, CREMτ, play a role in cardiomyocytes. Interestingly, tropomyosin 1α, cardiac α-actin, and troponin I are all concordantly regulated in old and young β1ARTG/Crem−/− versus β1ARTG cardiomyocytes and are all part of the contractile apparatus of the myocyte as well as myosin light chain 1 (also referred to as slow-twitch muscle B/ventricular isoform MLC1SB), which was only regulated in old cardiomyocytes. Therefore, changes in distinct cardiac contractile proteins are associated with the partial rescue of β1ARTG mice by CREM inactivation and may well contribute to the improved LV function of β1ARTG/Crem−/− mice. Other proteins regulated in both old and young β1ARTG/Crem−/− versus β1ARTG myocytes are ATP synthase-γ, malate dehydrogenase, and NADH dehydrogenase [ubiquinone] 1 beta subcomplex subunit 10. All of these proteins are mitochondrial enzymes involved in the energy supply of cells, here cardiomyocytes. It is conceivable that expressional changes of these proteins contribute to the improved phenotype of β1ARTG/Crem−/− mice; however, little is known about the cardiac role of these proteins in general and in the context of heart failure.
Taken together, the morphological, functional, and expressional findings clearly emphasize the importance of CREM as a central regulator of gene expression in the pathogenesis of heart failure. The partial rescue and concomitant changes of Ryr2, Anxa4, and other genes encoding functionally important contractile proteins emphasize their physiological role in the heart and suggest these genes as key proteins implicated in the pathogenesis of heart failure. The present results should therefore initiate further investigations to study the functional role of target genes of CREM and to develop strategies for inhibition of CREM as a novel therapeutic option to treat heart failure.
The authors wish to thank Anna-Maria Mehlich, Vera Samoilova, Christine Schröder, and Andrea Walter for their excellent technical assistance.
Sources of Funding
This work was supported by the Deutsche Forschungsgemeinschaft (DFG Mu1376/10-1/2/3) and the Interdiziplinäres Zentrum für Klinische Forschung/Interdisciplinary Centre for Clinical Reseach Münster (DLR/BMBF/IZKF Mu01/021/2004).
Colucci WS, Braunwald E. Pathophysiology of heart failure. In: Braunwald E, ed. Heart Disease. Philadelphia, Pa: WB Saunders; 1997.
Bisognano JD, Weinberger HD, Bohlmeyer TJ, Pende A, Raynolds MV, Sastravaha A, Roden R, Asano K, Blaxall BC, Wu SC, Communal C, Singh K, Colucci W, Bristow MR, Port DJ. Myocardial-directed overexpression of the human beta(1)-adrenergic receptor in transgenic mice. J Mol Cell Cardiol. 2000; 32: 817–830.
Engelhardt S, Hein L, Wiesmann F, Lohse MJ. Progressive hypertrophy and heart failure in beta1-adrenergic receptor transgenic mice. Proc Natl Acad Sci U S A. 1999; 96: 7059–7064.
Geng YJ, Ishikawa Y, Vatner DE, Wagner TE, Bishop SP, Vatner SF, Homcy CJ. Apoptosis of cardiac myocytes in Gsalpha transgenic mice. Circ Res. 1999; 84: 34–42.
Antos CL, Frey N, Marx SO, Reiken S, Gaburjakova M, Richardson JA, Marks AR, Olson EN. Dilated cardiomyopathy and sudden death resulting from constitutive activation of protein kinase A. Circ Res. 2001; 89: 997–1004.
Roth DM, Gao MH, Lai NC, Drumm J, Dalton N, Zhou JY, Zhu J, Entrikin D, Hammond HK. Cardiac-directed adenylyl cyclase expression improves heart function in murine cardiomyopathy. Circulation. 1999; 99: 3099–3102.
Gao MH, Tang T, Guo T, Sun SQ, Feramisco JR, Hammond HK. Adenylyl cyclase type VI gene transfer reduces phospholamban expression in cardiac myocytes via activating transcription factor 3. J Biol Chem. 2004; 279: 38797–38802.
Engelhardt S, Boknik P, Keller U, Neumann J, Lohse MJ, Hein L. Early impairment of calcium handling and altered expression of junctin in hearts of mice overexpressing the beta1-adrenergic receptor. FASEB J. 2001; 15: 2718–2720.
Gaussin V, Tomlinson JE, Depre C, Engelhardt S, Antos CL, Takagi G, Hein L, Topper JN, Liggett SB, Olson EN, Lohse MJ, Vatner SF, Vatner DE. Common genomic response in different mouse models of beta-adrenergic-induced cardiomyopathy. Circulation. 2003; 108: 2926–2933.
Matus M, Lewin G, Stümpel F, Buchwalow IB, Schneider MD, Schütz G, Schmitz W, Müller FU. Cardiomyocyte-specific inactivation of transcription factor CREB in mice. FASEB J. 2007; 21: 1884–1892.
Müller FU, Lewin G, Matus M, Neumann J, Riemann B, Wistuba J, Schütz G, Schmitz W. Impaired cardiac contraction and relaxation and decreased expression of sarcoplasmic Ca2+-ATPase in mice lacking the CREM gene. FASEB J. 2003; 17: 103–105.
Müller FU, Lewin G, Baba HA, Boknik P, Fabritz L, Kirchhefer U, Kirchhof P, Loser K, Matus M, Neumann J, Riemann B, Schmitz W. Heart-directed expression of a human cardiac isoform of cAMP-response element modulator in transgenic mice. J Biol Chem. 2005; 280: 6906–6914.
Isoda T, Paolocci N, Haghighi K, Wang C, Wang Y, Georgakopoulos D, Servillo G, Della Fazia MA, Kranias EG, Depaoli-Roach AA, Sassone-Corsi P, Kass DA. Novel regulation of cardiac force-frequency relation by CREM (cAMP response element modulator). FASEB J. 2003; 17: 144–151.
Yan C, Miller CL, Abe J. Regulation of phosphodiesterase 3 and inducible cAMP early repressor in the heart. Circ Res. 2007; 100: 489–501.
Grote-Wessels S, Baba HA, Boknik P, El Armouche A, Fabritz L, Gillmann HJ, Kucerova D, Matus M, Müller FU, Neumann J, Schmitz M, Stumpel F, Theilmeier G, Wohlschlaeger J, Schmitz W, Kirchhefer U. Inhibition of protein phosphatase 1 by inhibitor-2 exacerbates progression of cardiac failure in a model with pressure overload. Cardiovasc Res. 2008.
Kirchhefer U, Klimas J, Baba HA, Buchwalow IB, Fabritz L, Hüls M, Matus M, Müller FU, Schmitz W, Neumann J. Triadin is a critical determinant of cellular Ca cycling and contractility in the heart. Am J Physiol. 2007; 293: H3165–H3174.
Kirchhefer U, Neumann J, Baba HA, Begrow F, Kobayashi YM, Reinke U, Schmitz W, Jones LR. Cardiac hypertrophy and impaired relaxation in transgenic mice overexpressing triadin 1. J Biol Chem. 2001; 276: 4142–4149.
Zhang X, Odom DT, Koo SH, Conkright MD, Canettieri G, Best J, Chen H, Jenner R, Herbolsheimer E, Jacobsen E, Kadam S, Ecker JR, Emerson B, Hogenesch JB, Unterman T, Young RA, Montminy M. Genome-wide analysis of cAMP-response element binding protein occupancy, phosphorylation, and target gene activation in human tissues. Proc Natl Acad Sci U S A. 2005; 102: 4459–4464.
Frangogiannis NG, Ren G, Dewald O, Zymek P, Haudek S, Koerting A, Winkelmann K, Michael LH, Lawler J, Entman ML. Critical role of endogenous thrombospondin-1 in preventing expansion of healing myocardial infarcts. Circulation. 2005; 111: 2935–2942.
Salnikow K, Wang S, Costa M. Induction of activating transcription factor 1 by nickel and its role as a negative regulator of thrombospondin I gene expression. Cancer Res. 1997; 57: 5060–5066.
Barandon L, Couffinhal T, Ezan J, Dufourcq P, Costet P, Alzieu P, Leroux L, Moreau C, Dare D, Duplaa C. Reduction of infarct size and prevention of cardiac rupture in transgenic mice overexpressing FrzA. Circulation. 2003; 108: 2282–2289.
Hardt SE, Sadoshima J. Glycogen synthase kinase-3beta: a novel regulator of cardiac hypertrophy and development. Circ Res. 2002; 90: 1055–1063.
Müller FU, Boknik P, Knapp J, Neumann J, Vahlensieck U, Oetjen E, Scheld HH, Schmitz W. Identification and expression of a novel isoform of cAMP response element modulator in the human heart. FASEB J. 1998; 12: 1191–1199.
Sun XH, Protasi F, Takahashi M, Takeshima H, Ferguson DG, Franzini-Armstrong C. Molecular architecture of membranes involved in excitation-contraction coupling of cardiac muscle. J Cell Biol. 1995; 129: 659–671.
Zhao XM, Hu Y, Miller GG, Mitchell RN, Libby P. Association of thrombospondin-1 and cardiac allograft vasculopathy in human cardiac allografts. Circulation. 2001; 103: 525–531.
Schellings MW, Pinto YM, Heymans S. Matricellular proteins in the heart: possible role during stress and remodeling. Cardiovasc Res. 2004; 64: 24–31.
Matteo RG, Moravec CS. Immunolocalization of annexins IV, V and VI in the failing and non-failing human heart. Cardiovasc Res. 2000; 45: 961–970.
Chan HC, Kaetzel MA, Gotter AL, Dedman JR, Nelson DJ. Annexin IV inhibits calmodulin-dependent protein kinase II-activated chloride conductance: a novel mechanism for ion channel regulation. J Biol Chem. 1994; 269: 32464–32468.
Heart failure, the common end stage of various cardiac diseases, is defined as the inability of the heart to cover the body’s blood supply and is associated with a poor prognosis. Supported by clinical studies and studies on various animal models, chronic stimulation of the β1-adrenoceptor (β1AR) by elevated plasma catecholamines is regarded as a central pathogenetic factor of human heart failure, leading to the progression of the disease. Transcription factors of the cyclic AMP response element binding protein/cyclic AMP response element modulator (CREM) family mediate a cyclic AMP–dependent control of gene expression and may therefore contribute to β1AR-mediated detrimental cardiac effects. In the present study, we show that the inactivation of CREM protects from cardiomyocyte hypertrophy, fibrosis, and left ventricular dysfunction in mice with heart-directed expression of the β1AR, along with the altered expression of various myocardial proteins including the cardiac ryanodine receptor, tropomyosin 1α, and cardiac α-actin. These findings support the notion that the regulation of genes by CREM represents an important mechanism of β1AR-induced cardiac damage and suggest the inhibition of CREM as a promising therapeutic option for treating heart failure.
↵*The first 2 authors contributed equally to this work.
The online-only Data Supplement is available with this article at http://circ.ahajournals.org/cgi/content/full/CIRCULATIONAHA.108.786533/DC1.