Three major MAP kinase signaling cascades, ERK, p38, and JNK, play significant roles in the development of cardiac hypertrophy and heart failure in response to external stress and neural/hormonal stimuli. To study the specific function of each MAP kinase branch in adult heart, we have generated three transgenic mouse models with cardiac-specific and temporally regulated expression of activated mutants of Ras, MAP kinase kinase (MKK)3, and MKK7, which are selective upstream activators for ERK, p38, and JNK, respectively. Gene expression profiles in transgenic adult hearts were determined using cDNA microarrays at both early (4–7 days) and late (2–4 wk) time points following transgene induction. From this study, we revealed common changes in gene expression among the three models, particularly involving extracellular matrix remodeling. However, distinct expression patterns characteristic for each pathway were also identified in cell signaling, growth, and physiology. In addition, genes with dynamic expression differences between early vs. late stages illustrated primary vs. secondary changes on MAP kinase activation in adult hearts. These results provide an overview to both short-term and long-term effects of MAP kinase activation in heart and support some common as well as unique roles for each MAP kinase cascade in the development of heart failure.
- gene regulation
- transgenic mice
the onset of heart failure is a chronic process, usually involving compensatory hypertrophy at an early stage to adapt to internal and external stressors (10, 11). However, under continuous stress stimulation (either mechanical or neurohormonal), this adaptive remodeling will lead to decompensation and overt heart failure. The underlying molecular mechanisms for both adaptation and decompensation are still largely unknown and have been the focus of intensive investigations (1, 3, 19, 23, 25, 27, 29, 39). From recent studies, several intracellular signaling pathways, including mitogen-activated protein kinases (MAPKs), are identified to be potentially important players in the process of cardiac hypertrophy and the development of heart failure (1, 23). However, their specific contributions to cardiac hypertrophy and heart failure remain unclear and in many cases controversial (3, 4, 19, 20, 25, 27, 29, 37, 38, 40). MAPKs consist of highly conserved cascades of upstream MAP kinase kinase kinases (MEKKs), MAP kinase kinases (MKKs), and MAP kinases. The major branches of MAPKs include extracellular signal-regulated kinase (ERK1/2), ERK-5, c-Jun NH2-terminal kinase (JNK), and stress-activated p38 MAPKs. Activation of ERKs is often associated with tyrosine kinase receptor-mediated signal transduction, involving Ras/Raf/MEK/ERK cascade, whereas JNK and p38 are most often activated by stress-related signals, involving specific upstream kinases such as MKK4/7 for JNK and MKK3/6 for p38.
All branches of MAPKs have been shown to be induced in various conditions of heart failure, but the specific contribution of each MAPK pathway remains unclear. The relative specificity of the upstream MAPK kinase for each branch of the MAPKs has been utilized to investigate the role of individual MAPK pathways in cardiac hypertrophy and heart failure both in vitro and in vivo (8, 18–20, 25, 27, 29, 34, 37, 38). Indeed, previous work from our and other laboratories has demonstrated unique features of cardiac pathology in response to chronic activation of individual MAPK pathways. Specifically, ERK activation by Ras is shown to induce cardiac hypertrophy with cellular defects in intracellular calcium regulation (12, 13, 40). In addition to the prototypic ERK1/2 pathway, a number of other downstream pathways, including phosphatidylinositol 3-kinase (PI3K)/AKT, can also be involved in the Ras-mediated signaling that contributes to cardiac hypertrophy (23). In contrast, targeted activation of both p38 and JNK leads to restrictive cardiomyopathy with no hypertrophy (17–20, 25, 27, 29); in addition, p38 activation results in contractile dysfunction associated with inflammatory induction (16, 19, 20), while JNK activation impairs cell-cell communication associated with transforming growth factor (TGF)-β induction (17, 25, 26). Therefore, activation of each MAPK pathway appears to contribute to specific features of cardiac pathology.
Our current study aims to reveal the underlying molecular mechanism for the specific effects of MAPK activation in heart by analyzing the gene expression profiles in transgenic hearts where each MAPK cascade is selectively activated. However, chronic activation of one signaling pathway often triggers secondary feedback mechanisms that may alter the primary gene induction over time, thus masking the direct molecular signaling events. Therefore, we employed a recently developed inducible gene switch strategy to achieve targeted expression of transgenes in adult transgenic heart in a temporally controlled manner (28, 33). Specifically, constitutively activated HRas-v12, MKK7D, and MKK3bE proteins are expressed in adult hearts as selective activators of ERK, JNK, and p38 MAPK pathways, respectively. Affymetrix MOE 430 2.0 gene chips containing 45,000 probe sets to detect >34,000 well-characterized mouse genes and other transcripts (expressed sequence tags; ESTs) were used to determine the global gene expression profile in cardiac tissue at early and late time points of the transgene induction. The expression pattern was analyzed on the basis of Gene Ontology, established biological activities and signaling pathways, and the changes of a subset of genes were confirmed by quantitative real-time RT-PCR. These results revealed both common and unique features of gene expression changes in each transgenic model and a dynamic gene expression pattern over the time course of MAPK activation. The data support an important role for MAPK pathways in the onset of heart failure and point out potential mechanisms in MAPK-mediated pathological changes in heart. Our study demonstrates the critical importance of combining specific genetic manipulation with genomic approaches to dissect mechanisms of complex diseases like heart failure and provides an important source of information for both basic understanding of the disease and development of new therapies.
MATERIALS AND METHODS
The mouse strains for all of the transgenic lines are C57black background and were established after nearly 2 yr of crossbreeding between the floxed alleles and the Mer-Cre-Mer (MCM) strain. α-Myosin heavy chain (MHC)-floxed-HRas-v12/MKK3bE/MKK7D transgenic mice were established as reported previously (19, 20, 37, 38, 40). They were bred with MCM mice (from Dr. J. Molkentin, Cincinnati Children's Hospital) (33) to generate double-transgenic animals harboring both floxed transgenes and the MCM transgene. At 12 wk of age, the double-transgenic mice and nontransgenic littermate controls were treated via intraperitoneal (ip) injection of tamoxifen at a dosage of 20 mg/kg body wt once a day for 3 consecutive days as reported. The hearts were harvested at an early (4–7 days post-first tamoxifen injection) and a late (2–4 wk) time point. Left ventricles were dissected and rapidly frozen in liquid nitrogen and stored at −80°C before protein and RNA analysis. The animal protocol used in these experiments was approved by the UCLA Institutional Animal Care and Use Committee (no. 2003-105).
Protein was extracted from frozen left ventricular tissue, and Western blots were performed using the following antibodies: anti-phospho-ERK (Thr202/Tyr204), anti-phospho-p38 (Thr180/Tyr182), anti-phospho-JNK (Thr183/Tyr185), anti-MKK3 (Cell Signaling), Ras (Upstate Biotechnology), and anti-MKK7 (Santa Cruz Biotechnology) at a dilution of 1:1,000. Horseradish peroxidase (HRP)-conjugated secondary antibody was used at a dilution of 1:10,000.
RNA purification and microarray preparation.
The left ventricles of the hearts were crushed, and total RNA was isolated using TRIzol (Invitrogen) according to manufacturer's protocol. In most cases, triplicate samples were used for each transgenic line at each time point; however, only duplicates were available for the HRas-v12 early and the MKK3bE early and late samples. The samples were then purified using RNeasy Mini Kit (Qiagen). Reverse transcription of the mRNA, probe labeling, and hybridization were performed using the standard Affymetrix protocols and conducted at the UCLA DNA Microarray Core. The DNA hybridization utilized the Affymetrix MOE 430 2.0 gene chip containing 45,000 probe sets that analyze the expression level of >39,000 transcripts and variants from >34,000 well-characterized mouse genes (Affymetrix). Chip Scanner 3000 with GeneChip Operating Software (GCOS) v1.1.1 (Affymetrix) was used in scanning.
Real-time RT-PCR analysis.
Total RNA was isolated as described above, and reverse transcription was performed using SuperScript First-Strand Synthesis kit for RT-PCR (Invitrogen) according to the manufacturer's protocol. Quantitative real-time RT-PCR was performed, using 27.5 μl of iQ SYBR Green Supermix (BioRad) containing 250 nM specific primers (Supplemental Table S1; available at the Physiological Genomics web site)1 and cDNA transcribed from 5 μg of total RNA using iCycler thermocycler (BioRad) for 40 cycles of reaction, with 45 s each of 95°C, 60°C, and 72°C, and expression levels were analyzed using Optical System software v3.1 (BioRad) and normalized against GAPDH. Representative quantification graphs were plotted using the relative florescence unit (RFU) vs. PCR amplification cycle.
Affymetrix Microarray Suite (MAS) 5.0 gene expression values were normalized with the median invariant method (15). To filter out significant genes, two approaches were used. The primary filtering method used the standard two-sample t-test in pairwise comparisons involving animal types (control, JNK, RAS, p38) and different time points (early vs. late). Unless otherwise stated, the selection criteria are as follows: 1) |the fold change| >2, 2) |difference| >500, and 3) P value < 0.05. Pairwise comparisons were performed and the gene lists generated. For each comparison, a two-way (genes and samples) clustering was performed. Each clustering resulted in two output files, a cluster information file and a cluster image file. Through querying Expression Analysis Systematic Explorer (EASE) (http://david.niaid.nih.gov/david/ease.htm) (14), a gene category report for each of the selected gene lists was also provided. Pathway analysis was performed using the GenMAPP software (http://www.genmapp.org) and the Kyoto Encyclopedia of Genes and Genomes (KEGG), and functional classification was performed using Gene Ontology (GO) terms. The empirical data can be obtained from the Gene Expression Omnibus (GEO) database with accession number GSE3530.
RESULTS AND DISCUSSION
Specific activation of MAPK pathway in adult transgenic heart.
The temporally regulated transgene expression in adult heart was achieved using an inducible gene switch approach as schematically illustrated in Fig. 1A. After 3 days of tamoxifen administration in double-transgenic animals harboring both floxed transgenic constructs and Mer-Cre-Mer, the expression of the target genes, HRas-v12, MKK7D, and MKK3bE, was confirmed by immunoblot of protein samples prepared from the ventricular tissues or RT-PCR on isolated mRNA, respectively (Fig. 1, B and C). The expression of Ras, MKK3b, and MKK7 mutants was associated with selective activation of ERK, p38, and JNK MAPKs, respectively, as demonstrated in the immunoblots using phosphospecific antibodies (Fig. 1D). These data indicate that selective activation of ERK, p38, and JNK pathways is achieved in the adult hearts following targeted and inducible expression of their upstream activators.
Quality control and validation of microarray data.
RNA samples were prepared from left ventricles of age-, gender-, and treatment-matched double-transgenic and wildtype controls, as described in materials and methods. The quality assessments of RNA samples, amplification, biotin incorporation, hybridization, and staining were performed according to established protocols (9). The empirical data obtained were deposited in the GEO database with the accession number GSE3530. Subsets of genes identified to be significantly changed by microarray analysis were further confirmed by quantitative real-time RT-PCR (qRT-PCR). As shown in Fig. 2, after normalization against expression levels of GAPDH, the fold changes determined by qRT-PCR were plotted against the corresponding fold changes obtained from microarray results. For all the 29 genes/samples tested, the direction (positive/negative) was consistent between microarray data and qRT-PCR results (R2 value of 0.73), demonstrating the high fidelity of the microarray analysis to detect gene expression changes. The qRT-PCR results were validated using representative sample graphs showing differences in expression of a particular gene compared with GAPDH expression as controls (Fig. 2B).
Global gene expression profiles.
Affymetrix MAS 5.0 gene expression values were normalized with the median invariant method (15). To filter out significant changes, we used a standard two-group t-test in pairwise comparisons involving animal types (control, JNK, RAS, p38) and different time points (early, late) using the following selection criteria: |the fold change| >2, |difference| >500, and P < 0.05. For each comparison, two-way (genes and samples) clustering resulted in a cluster tree as illustrated in Supplemental Fig. S1. It is clear that different animal groups are delineated according to their genotypes, suggesting that a unique gene expression profile is identified in heart with selected activation of each MAPK pathways. Within each animal model, data sets from early vs. late time points are also clearly segregated, suggesting a dynamic change of global gene expression profile over time after MAPK activation. To provide a global view of gene expression changes among the three transgenic models, we used a more stringent filtering criteria (P ≤ 0.01, difference ≥500, fold change ≥2, and a <90% confidence bound of fold change) to identify genes that are changed in transgenic vs. control hearts. The false discovery rate by random permutation (100) was shown to be 10% for HRas-v12, 11.2% for MKK3bE, and 5.8% for MKK7D transgenic hearts. This resulted in a total of 185 genes in HRas-v12 hearts, 251 genes in MKK3bE hearts, and 257 genes in MKK7D hearts that were identified to be significantly changed compared with the controls (Fig. 3). Among the 693 genes altered in all three transgenic hearts, 36 genes, or merely 5.2%, are shared by all models. However, 94 genes (13.6%) are shared by both MKK3bE and MKK7D hearts, while 57 (8.2%) and 47 (6.8%) genes are shared by HRas-v12 with MKK3bE and MKK7D, respectively. The full list of genes is presented in Supplemental Table S2. These data suggest again that the three MAPK pathways have both common and unique downstream effects in the heart; however, p38 and JNK pathways appear to have more overlapping target genes relative to the Ras/ERK pathway. This global view of the gene expression profile is correlated well with phenotypic differences, as HRas-v12 transgenic hearts develop hypertrophic cardiomyopathy, while both MKK3bE and MKK7D transgenic hearts have a restrictive cardiomyopathy phenotype. Therefore, these differences and common features in the global gene expression pattern underscore the potential mechanisms underlying hypertrophic vs. restrictive cardiomyopathy phenotypes.
To further dissect the functional aspect of gene expression changes for each transgenic model, GO analysis was performed among the genes with significant changes in expression levels (|fold change| >2, |difference| >100, P value < 0.05) from each transgenic model at either early or late time points (Fig. 4). Among major categories of genes (at least 3% represented among total no. of genes) with significant hits (>5 categories), genes involved in cell communication and signal transduction are shared by all three models. However, Ras activation led to significant changes in genes involved in cell proliferation, immune response, redox regulation, and stress responses; in contrast, p38 MAPK activation resulted in changes in genes for morphogenesis, organogenesis, growth factor activation, cytoskeletal interaction, and cytokine activation. Lastly, JNK activation showed unique changes in genes for angiogenesis, potassium regulation, and extracellular matrix. These changes revealed potential mechanisms underlying specific features of cardiac pathology and once again demonstrate common as well as specific features of the gene expression profile in hearts with MAPK activation.
Cellular signaling pathway.
To investigate the functional significance of gene expression changes in the transgenic heart, we used the GenMAPP program (5) to depict signaling pathways that are known to be involved in the progression of heart failure (Fig. 5). It is immediately apparent that there are common features as well as significant differences in the components and dynamic pattern of signaling gene expression among the three transgenic models. To our surprise, the HRas-v12 transgenic heart showed relatively fewer changes of signaling genes at the early time point; in contrast, a significant number of signaling genes are up- or downregulated in JNK- or p38-activated hearts at both early and late time points. There appears to be a general trend toward an increase in the TGF-β signaling pathway among all three transgenic models. However, TGF-β induction in MKK3bE hearts is limited at the receptor level, while HRas-v12 and MKK7D hearts show induction of receptor genes as well as TGF-β cofactors and downstream signal molecules. It is reported that TGF-β signaling is involved in the transition from compensated hypertrophy to heart failure and myocardial fibrosis (30); therefore, our data suggest that different MAPKs may contribute to the overall progression of heart failure and myocardial remodeling by augmenting TGF-β signaling involving different components of the TGF-β cascade (Supplemental Fig. S2). The difference in temporal pattern of induction also suggests that MAPK-mediated TGF-β activity involves both direct (JNK and p38) and indirect (Ras) mechanisms. These common and unique impacts on TGF-β signaling may play a role in the different manifestation of interstitial remodeling in MAPK-activated hearts, i.e., collagen induction in HRas-v12 and MKK3bE hearts and fibronectin induction in MKK7D hearts. Similarly, differences and common features in the gene profiles can also be seen in PKC, PKD, and MAPK pathways. In the case of the MAPK pathway, HRas-v12 and MKK3bE mice show a decrease in apoptotic components involved in Ras and p38 pathways, while MKK7D shows induction of such genes, in good agreement with the earlier observation of apoptosis in JNK-activated hearts, which was absent in Ras- and p38-activated hearts (Ref. 1 and B. Petrich, A. Ota, and Y. Wang, unpublished results). Finally, nuclear receptors including peroxisome proliferator-activated receptor (PPAR) and retinoid X receptor (RXR), which are closely correlated with energy metabolism and gene expression (31), are downregulated consistently among all transgenic animals (Fig. 5). This result suggests a significant role of MAPK pathways in the regulation of energy metabolism in hearts, an observation supported by other GenMAPP analyses described below. Overall, the data suggest variable regulations on signaling pathways by specific MAPK pathways that may contribute to pathological remodeling in the transgenic hearts.
Using GenMAPP analysis, we also uncovered a striking impact of MAPK activation on the metabolic pathways. In HRas-v12 transgenic hearts, most of the genes in the electron transport chain, fatty acid metabolism, glycolysis/gluconeogenesis, and Krebs (tricarboxylic acid; TCA) cycle are significantly decreased at the early time point but are modestly improved at the later time point (Fig. 6). The remarkable suppression of metabolic genes includes some of the key regulators for energy metabolism, such as PPARα, very-long-chain-specific acyl-CoA dehydrogenase (VLCAD), muscle carnitine palmitoyl transferase (mCPT)1, uncoupling protein (UCP), and glucose transporter type-4 (GLUT4) (Fig. 7). Decreases in VLCAD and other genes in the fatty acid metabolic pathway (Supplemental Fig. S3) suggest that fatty acids, which are the preferred metabolic substrate in the heart, are not being metabolized, leading to a potential increase in free fatty acids. Previous studies have shown that, during the process of heart failure, free fatty acids modulate the transcription of PPARα, GLUT4, and UCP (21, 22, 24, 31, 35, 36). Early reports suggest that free fatty acids mediate loss of GLUT4 and induction of UCP2 and UCP3 that can lead to reduced glucose uptake and lower ATP synthesis. These findings predict that constitutively activated Ras may yield to an energy-deficient myocardium, which might contribute to the compromised cardiac function and the transition to the decompensated state. Similar patterns of reduction in metabolic gene expression are detected in MKK7D and MKK3bE transgenic hearts, even at the early time point, and become further downregulated at the later time point (Fig. 6 and Supplemental Fig. S4), suggesting that energy deficiency is a common feature of heart failure involving MAPK activation.
Myocardial remodeling and inflammatory response.
In the HRas-v12 transgenic hearts, there is a significant induction of extracellular matrix proteins such as collagen, fibronectin, thrombospondin, and laminin at the early time point that is also sustained at the later time point. Tissue inhibitor of metalloproteinases (TIMP) is significantly increased with a corresponding decrease in matrix metalloproteinases (MMPs) (Fig. 8). However, there seems to be no induction of the inflammatory pathway in response to Ras activation (Fig. 8 and Supplemental Fig. S5). The gene expression changes for the JNK-activated hearts mostly show a delayed induction in genes involved in both remodeling and inflammatory responses. The same follows for the p38-activated heart (Fig. 8). TGF-β and tumor necrosis factor (TNF)α have both been implicated in the development of cardiac hypertrophy and are major contributors to tissue fibrosis in various tissues including heart (30, 37, 38). Cyclooxygenase (COX)-2 activity also has been implicated recently in the development of heart failure (7), and both JNK and p38 activation have been shown to play an integral part in the regulation of TNFα, IL-1β, and COX-2 expression in the heart (6, 16). These findings indicate that extracellular matrix remodeling is a common feature of MAPK-mediated cardiomyopathy, as predicted from previous reports (19, 20, 25, 40). However, different mechanisms may underlie these changes mediated by specific MAPK pathways. The Ras-ERK pathway may directly regulate the expression of extracellular matrix proteins; in contrast, JNK and p38 pathway-induced remodeling may involve inflammatory gene induction.
Ion channel and calcium regulation.
Electrophysiology and calcium homeostasis are important aspects of excitation and contraction properties of the cardiac myocytes and are often misregulated in failing hearts, leading to arrhythmia and contractile dysfunction (2). Analysis of MKK7D and MKK3bE transgenic models shows no significant changes in the expression of ion channels that may compromise contractility or excitation property. However, several significant changes are observed in the HRas-v12 model (Table 1). Specifically, at the late time point, there is no significant decrease in the L-type calcium channel that is critical to calcium influx. However, a significant increase in sodium channels may affect action potential profile during depolarization. In addition, decreases in potassium channels imply that K+ is not being efficiently extruded from the cardiac myocytes and thereby prolongs the action potential duration. Furthermore, decreases in sarcoplasmic reticulum (SR) Ca2+-ATPase (SERCA) expression may contribute to reduced SR calcium uptake, loss of SR calcium content, and reduced contractility. The observed increase in sodium calcium exchanger (NCX) expression may function to compensate for the loss of SR calcium uptake during diastole but can also exacerbate the loss of intracellular calcium store and systolic function and contribute to action potential prolongation (32). Our genomic analysis clearly demonstrates a specific role of Ras-mediated MAPK pathways in the regulation of ion channel and calcium homeostasis in cardiomyocytes. Indeed, SR calcium defects, contractile dysfunction, and arrhythmia have been observed in Ras-activated ventricular myocytes (12, 13, 40) but not in MKK7D or MKK3bE transgenic animals, and cardiac arrhythmia is significantly induced only in HRas-v12 transgenic animals, not in MKK7D or MKK3bE transgenic animals (unpublished observations). Therefore, this microarray analysis provides a molecular fingerprint of genes that might contribute to the contractile dysfunction and arrhythmia in Ras-activated hypertrophic hearts.
In summary, the activation of multiple signaling pathways during the development of heart failure makes the task of dissecting the roles of specific pathways extremely difficult. Furthermore, elucidating the primary vs. secondary effects of a particular pathway can be equally challenging. Our microarray analysis of genetic models with targeted activation of Ras, p38, and JNK pathways at defined time points in adult hearts provides some new insights into the contribution of each of the MAPK pathways to the development of heart failure. Analyses of global gene expression in Ras-, p38-, and JNK-activated hearts reveal distinct profiles for each of the pathways. The changes in the expression profiles between the early and late time points within each MAPK pathway further signify the differences of the primary vs. secondary effect of MAPK activation during the progression to heart failure. Although MAPK activation leads to distinct gene profiles, some commonality among the pathways is also identified, with the p38 and JNK pathways being more closely associated with each other than the Ras pathway. These similarities and differences may outline the shared features as well as the disparity in the development of restrictive cardiomyopathy, as seen in MKK3bE and MKK7D mice, and hypertrophic cardiomyopathy, as shown in HRas-v12 mice. The distinct gene expression changes in cellular signaling, metabolism, remodeling, inflammatory response, and calcium regulation in the HRas-v12, MKK3bE, and MKK7D mice highlight some of the specific contributions of each pathway to different aspects of pathological changes in heart failure. Therefore, these results facilitate the dissection of the underlying molecular mechanisms of the disease and help to identify new venues for the development of effective heart failure therapies.
This work is supported by National Heart, Lung, and Blood Institute (NHLBI) Grants HL-62311, HL-70079, HL-080111, and HL-70709 to Y. Wang. Microarray experiments and analysis were supported through the UCLA NHLBI Shared Microarray Facility (R01 HL-072367). Y. Wang is an Established Investigator of the American Heart Association.
↵1 The Supplemental Material for this article (Supplemental Tables S1 and S2 and Supplemental Figs. S1–S5) is available online at http://physiolgenomics.physiology.org/cgi/content/full/00224.2005/DC1.
Article published online before print. See web site for date of publication (http://physiolgenomics.physiology.org).
Address for reprint requests and other correspondence: Y. Wang, Depts. of Anesthesiology and Medicine, Division of Molecular Medicine and Molecular Biology Institute, David Geffen School of Medicine at UCLA, Los Angeles, CA 90095 (e-mail:).
- Copyright © 2006 the American Physiological Society