Gab family proteins are essential for postnatal maintenance of cardiac function via neuregulin-1/ErbB signaling

Grb2-associated binder (Gab) family of scaffolding adaptor proteins coordinate signaling cascades downstream of growth factor and cytokine receptors. In the heart, among EGF family members, neuregulin-1β (NRG-1β, a paracrine factor produced from endothelium) induced remarkable tyrosine phosphorylation of Gab1 and Gab2 via erythroblastic leukemia viral oncogene (ErbB) receptors. We examined the role of Gab family proteins in NRG-1β/ErbB-mediated signal in the heart by creating cardiomyocyte-specific Gab1/Gab2 double knockout mice (DKO mice). Although DKO mice were viable, they exhibited marked ventricular dilatation and reduced contractility with aging. DKO mice showed high mortality after birth because of heart failure. In addition, we noticed remarkable endocardial fibroelastosis and increase of abnormally dilated vessels in the ventricles of DKO mice. NRG-1β induced activation of both ERK and AKT in the hearts of control mice but not in those of DKO mice. Using DNA microarray analysis, we found that stimulation with NRG-1β upregulated expression of an endothelium-stabilizing factor, angiopoietin 1, in the hearts of control mice but not in those of DKO mice, which accounted for the pathological abnormalities in the DKO hearts. Taken together, our observations indicated that in the NRG-1β/ErbB signaling, Gab1 and Gab2 of the myocardium are essential for both maintenance of myocardial function and stabilization of cardiac capillary and endocardial endothelium in the postnatal heart.

Dilated cardiomyopathy (DCM) is an common cause of heart failure. Epidemiological studies suggest that 25%–30% of DCM is inherited. Among the mutations associated with DCM in humans and mice, several involve genes encoding cytoskeletal proteins and sarcomere-related proteins (1); however, mutations in these known genes account for only a minor proportion of the heritable cardiomyopathies in humans. Cardiac function is maintained by cytokine- and growth factor–triggered intracellular signaling. Genetically modified mice, in which intracellular signaling molecules are either activated or perturbed, also exhibit cardiac dysfunction, suggesting that coordination of signal transduction systems is critical for the preservation of cardiac function (2).

The Grb2-associated binder (Gab) family proteins, which serve as scaffolding adaptor proteins, crucially intervene between receptors and intracellular signaling molecules to coordinate the signaling cascades of cytokines, growth factors, antigens, and numerous other molecules (3–5). Multiple phosphorylated tyrosine residues of Gab proteins become docking sites for Src homology-2 domain-containing molecules. Docking of Gab to tyrosine phosphatase SHP2 and the p85 regulatory subunit of PI3K leads to the activation of ERK and AKT, respectively (4, 5). Three Gab family members, Gab1, Gab2, and Gab3, have been identified in mammals and are structurally similar (4, 5). Conventional Gab1 knockout (Gab1KO) mice display embryonic lethality with impaired development of heart, placenta, skin, and muscle (6, 7). Gab2KO mice do not show any obvious developmental defects but display impaired allergic responses and osteoclast defects (8–11). Gab3KO mice exhibit no obvious phenotype (12).

We previously demonstrated the importance of Gab1-ERK5 signaling in cardiomyocyte hypertrophy through the leukemia inhibitory factor–gp130–dependent (LIF-gp130–dependent) signaling pathway (13). Gab family proteins are also involved in EGF family–erythroblastic leukemia viral oncogene (EGF family–ErbB) receptor family signaling (6, 14, 15). EGF family–ErbB receptor signaling plays crucial roles in heart development and preservation of adult cardiac function (16, 17). Among the EGF family members, neuregulin-1 (NRG-1) (18) and heparin-binding EGF-like growth factor (HB-EGF) (19) are particularly important agonists for ErbB receptors on cardiomyocytes. NRG-1 serves as a paracrine factor that is shed from the endothelium and activates the ErbB4 homodimer or ErbB2/ErbB4 (also known as HER2/HER4) heterodimer on cardiomyocytes (16, 17, 20, 21). NRG-1–, ErbB2-, and ErbB4-deficient mice display embryonic lethality and similar defects in ventricular trabeculation (22–24). HB-EGF–deficient mice also display abnormal valvular development and cardiac dysfunction (25, 26).

The importance of ErbB signaling in the adult heart was first revealed by the unforeseen adverse effects of trastuzumab (Herceptin), a monoclonal Ab against ErbB2 used in the treatment of breast cancer. Trastuzumab induces heart failure when combined with anthracycline treatment (17, 27, 28). In addition to this clinical evidence, cardiomyocyte-specific ErbB2- and ErbB4-deficient mice both exhibit DCM (29–31). However, the precise intracellular signaling responsible for ErbB-regulated cardiac function is still unclear.

In the present study, we used myocardium-specific deletion of Gab family proteins in the mice to demonstrate that Gab1 and Gab2 in the myocardium are essential for transmitting the signal from NRG-1β/ErbB to directly maintain myocardial function and to subsequently stabilize capillary and endocardial endothelium in the postnatal heart.

Gab1 and Gab2 are engaged in coordination of NRG-1β/ErbB signaling pathway in the myocardium. We aimed at exploring the function of Gab family proteins in the heart. Thus, we first examined the expression of Gab family transcripts by RT-PCR and detected the mRNA of Gab1 and Gab2, but not that of Gab3, in the murine heart (Supplemental Figure 1; supplemental material available online with this article; doi:10.1172/JCI30651DS1). To elucidate how Gab1 and Gab2 are involved in the intracellular signaling in the heart, mice were injected with various cytokines and growth factors. Among these agonists, ErbB receptor-activating agonists, including NRG-1β, HB-EGF, and EGF, induced strong tyrosine phosphorylation of Gab1 and Gab2 and the subsequent association of Gab1 and Gab2 with SHP2 and p85 (Figure 1, A and B). We identified 2 Gab1 isoforms, high–molecular weight (high-MW) Gab1 (120–130 kDa) and low-MW Gab1 (100 kDa). Notably, the high-MW Gab1 underwent tyrosine phosphorylation upon stimulation exclusively with NRG-1β, while low-MW Gab1 was phosphorylated by NRG-1β, HB-EGF, and EGF (Figure 1A). We confirmed that the high-MW Gab1 is a cardiac-specific isoform using molecular mass spectrometric analysis, which showed that the high-MW band that was recognized by anti-Gab1 Ab in Western blot analysis indeed contained the partial amino acid sequence of Gab1 (Supplemental Figure 2, A–C). Activation of both ERK and AKT was found only when stimulated with NRG-1β, HB-EGF, and EGF (Figure 1C), although activation of AKT was most strongly induced by IGF-1.

Figure 1

Gab1 and Gab2 are engaged in coordination of NRG-1β/ErbB signaling pathway in the myocardium. Tyrosine phosphorylation of Gab1 (A) and Gab2 (B) and their association with SHP2 and p85 were analyzed by IP of the heart lysates. Mouse heart lysates were prepared at 5 minutes after injection with the cytokines and growth factors listed at top. Heart lysates were subjected to IP with anti-Gab1 (A) or anti-Gab2 (B) serum, followed by IB analysis using the Ab indicated at the left. (C) Activation levels of ERK and AKT were assessed by phospho-specific Ab. Tyrosine phosphorylation of Gab1 (D) Gab2 (E) and their association with SHP2 and p85 was examined by IP of cell lysates from neonatal rat cardiomyocytes (CM) or noncardiomyocytes (non-CM) stimulated with either NRG-1β (50 ng/ml) or HB-EGF (50 ng/ml) for 5 minutes. IP complexes were subjected to IB using the Ab indicated at the left. (F) NRG-1β– and HB-EGF–dependent activation of ERK and AKT was examined in CM and non-CM as in C. Tyrosine phosphorylation of Gab1 (G) and Gab2 (H) and their association with SHP2 and p85 in the mouse hearts were analyzed after injection with 5 μg of NRG-1β as in A and B, respectively. Heart lysates were prepared at the indicated time after injection. Gab1 and Gab2 underwent tyrosine phosphorylation and associated with SHP2 and p85 in a time-dependent manner upon NRG-1β stimulation. (I) Activation of ERK and AKT were assessed as in C. Arrows denote 2 isoforms of Gab1. Representative blots of 3 experiments are shown. PY99, antibody recognizing phospho-tyrosine.

We examined whether the difference in Gab1 phosphorylation was due to the diversity of the cell types. To distinguish the signaling processes in cardiomyocytes from those in noncardiomyocytes, including fibroblasts, endothelial cells, and vascular smooth muscle cells in the heart, we analyzed the action of NRG-1β and HB-EGF in neonatal rat cardiomyocytes and noncardiomyocytes that had been isolated using the Percoll gradient method (32). NRG-1β induced tyrosine phosphorylation of Gab1 and Gab2, the subsequent association of Gab1 and Gab2 with SHP2 and p85, and the activation of ERK and AKT in cardiomyocytes but not in noncardiomyocytes (Figure 1, D–F). In clear contrast, HB-EGF induced those changes more strongly in noncardiomyocytes than in cardiomyocytes (Figure 1, D–F). It should be noted that tyrosine phosphorylation of the high-MW Gab1 in cardiomyocytes was induced after stimulation with NRG-1β but not with HB-EGF (Figure 1D). These findings suggest that NRG-1β acts as a highly selective agonist for cardiomyocytes, in agreement with previous reports (33).

Therefore, we focused on the NRG-1β–dependent signaling pathway through Gab1 and Gab2 in the murine hearts. Gab1 and Gab2 underwent tyrosine phosphorylation and associated with SHP2 and p85 after injection with NRG-1β in a time-dependent manner (Figure 1, G and H). In addition, both ERK and AKT were also activated by NRG-1β in a time-dependent manner (Figure 1I). We also checked the activation of ErbB family receptors of murine hearts stimulated with NRG-1β. NRG-1β induced tyrosine-phosphorylation of ErbB2 and ErbB4 but not that of ErbB1 (EGFR) or ErbB3 in accordance with a previous report in which cardiomyocytes were used in vitro (Supplemental Figure 3, A–D) (21). Furthermore, Gab1 associated with ErbB4 in a phosphorylation-dependent manner after injection with NRG-1β (Supplemental Figure 3E). These data suggest the engagement of Gab family proteins in the coordination of NRG-1β/ErbB signaling pathway.

Generation of cardiomyocyte-specific Gab1 conditional knockout mice. To elucidate the function of Gab family proteins in myocardium, we first generated cardiomyocyte-specific Gab1 conditional knockout (Gab1CKO) mice using the Cre-loxP system. Using homologous recombination in embryonic stem cells, we created a Gab1^flox allele by introducing 2 loxP sites into introns flanking exon 2, which encodes part of the pleckstrin homology domain (Figure 2A). The protein expression of Gab1 in all tissues of mice homozygous for the Gab1-loxP–targeted allele (Gab1^flox/flox mice) was almost the same level as in wild-type mice (data not shown). To cause recombination of the floxed allele exclusively in cardiomyocyte lineage, Gab1^flox/flox mice were crossed with transgenic mice expressing α-myosin heavy chain promoter–driven Cre recombinase (α-MHC–Cre mice) (34, 35) (Figure 2A). We confirmed the Cre-mediated recombination during embryogenesis (E10.5 and E14.5) by crossing α-MHC–Cre mice with enhanced GFP reporter mice (Supplemental Figure 4A). The Gab1CKO (Gab1^flox/flox α-MHC-Cre(+)) mice were born normally at the expected Mendelian frequency, whereas Gab1KO mice were embryonically lethal (6). In addition, the Gab1CKO mice displayed normal appearance and normal cardiac morphology at birth (Supplemental Figure 5A).

Figure 2

Generation of DKO mice. (A) Schematic illustration of genomic structure of the Gab1 wild-type, Gab1 flox, and Gab1-deleted alleles and a targeting vector. loxP sequences are indicated by black triangles. Restriction enzyme sites for EcoRI and HindIII are indicated as R and H, respectively. Fragments detected by the probe (short bold line) used for Southern blot analysis after digestion of genomic DNA with EcoRI and HindIII are indicated as solid lines measuring 4.3 kb, 3.8 kb, and 5.7 kb. HSV-TK, herpes simplex virus–thymidine kinase. (B) Southern blot analysis demonstrated recombination of the Gab1^flox allele in the heart, but not in the kidney, of Gab1^flox/flox mice, which possessed the α-MHC–Cre allele. (C) Following IP, expression of Gab1 and Gab2 was examined by IB using anti-Gab1 (top row) and anti-Gab2 (middle row) serums. SHP2 was examined as a loading control (bottom row). Note that 2 isoforms of Gab1 were detected at the different MW exclusively in the heart (arrows) and that the high-MW Gab1 isoform in the heart was completely depleted in Gab1CKO and DKO. The low-MW Gab1 was also reduced by 80% in the heart of Gab1CKO and DKO mice compared with control and Gab2KO mice.

We observed the expected genetic recombination at the Gab1 locus in the ventricles of Gab1CKO mouse hearts but not in other tissues (Figure 2B). In order to estimate the expression of Gab1 protein, immunoblot analyses were performed using the extracts from heart, liver, and kidney (Figure 2C). As described above, 2 isoforms of Gab1 proteins were detected in hearts, while low-MW Gab1 was commonly detected, suggesting that the high-MW Gab1 is a cardiac-specific isoform. Moreover, high-MW Gab1 protein was deleted in Gab1CKO hearts, suggesting that high-MW Gab1 is a product of the same Gab1 gene that has low MW. In addition, we used Percoll gradient centrifugation to analyze the expression of Gab1 in cardiomyocytes and noncardiomyocytes isolated from neonatal rat hearts (32) and detected the high-MW isoform of Gab1 exclusively in cardiomyocytes (Supplemental Figure 2D).

In Gab1CKO mice, the high-MW Gab1 was completely deleted and the low-MW Gab1 was reduced to about 20% of control (Gab1^flox/flox) littermates. The residual low-MW Gab1 protein might be attributed to the noncardiomyocytes present in the heart. These data indicated the successful depletion of Gab1 in the cardiomyocytes (Figure 2C), because α-MHC promoter functions exclusively in the myocardium. In 3-day-old Gab1CKO mouse hearts, we detected an extent of Gab1 protein depletion similar to that of 3- or 10-week-old mice (Supplemental Figure 4B).

Generation of cardiomyocyte-specific Gab1/Gab2 double knockout mice. In murine hearts, mRNAs of Gab1 and Gab2 were detected by RT-PCR (Supplemental Figure 1). Gab2 can rescue the loss of Gab1 for activation of ERK in the EGF signaling pathway (36). We thus assumed that Gab2 might compensate for the deletion of Gab1 in the cardiomyocytes of Gab1CKO mice.

To completely deplete Gab family proteins in cardiomyocytes, Gab1CKO mice were crossed with Gab2KO mice. We created Gab1^flox/floxGab2^–/–α-MHC–Cre(+) mice by crossing Gab1^+/floxGab2^–/–α-MHC–Cre(+) mice with Gab1^flox/floxGab2^–/–α-MHC–Cre(–) mice in the final breeding. The offspring of these crossings were recovered at expected Mendelian ratios as follows: Gab1^+/floxGab2^–/–α-MHC–Cre(–) (n = 44; 24.6%); Gab1^+/floxGab2^–/–α-MHC–Cre(+) (n = 46; 25.7%); Gab1^flox/floxGab2^–/–α-MHC–Cre(–) (n = 39; 21.8%); Gab1^flox/floxGab2^–/–α-MHC–Cre(+) (n = 50; 27.9%). Thereafter, we analyzed the following 4 groups of mice: Gab1^flox/floxGab2^+/+α-MHC–Cre(–) (control); Gab1^flox/floxGab2^+/+α-MHC–Cre(+) (Gab1CKO); Gab1^flox/floxGab2^–/–α-MHC–Cre(–) (Gab2KO); and Gab1^flox/floxGab2^–/–α-MHC–Cre(+) (DKO). Both Gab2KO and DKO mice displayed normal appearance and normal cardiac morphology at birth (Supplemental Figure 5A). Gab2 protein was completely depleted in the Gab2KO and DKO mice, indicating the successful depletion of Gab1 and Gab2 in the cardiomyocytes of DKO mice (Figure 2C).

DKO mice display dilated cardiomyopathic features accompanied by endocardial fibroelastosis. We performed gross morphological examination of the hearts of the 4 groups at 10 weeks of age because we did not find any morphological abnormalities in the hearts of Gab1CKO, Gab2KO, or DKO mice at birth (Supplemental Figure 5A). Although there was no morphological difference among Gab1CKO, Gab2KO, and control mice (Figure 3A), DKO mice exhibited significantly higher heart weight–to–body weight ratios than the other 3 groups without significant differences in body weight (Figure 3A and Figure 4B). Histological examination also demonstrated both left and right ventricular enlargement in DKO mice similar to DCM (Figure 3B).

Figure 3

DKO mice display dilated cardiomyopathic features accompanied by EFE. (A) Representative images of whole hearts from 4 groups at 10 weeks of age. (B) Transverse sections of the hearts were stained using the elastica van Gieson method. DKO hearts showed marked biventricular dilation and slight wall thinning compared with the other 3 groups of hearts. (C and E) Higher magnification of elastica van Gieson–stained section of DKO heart shows the focal accumulation of elastic fibers (black) in the endocardium (arrows in B and C). (D) Masson’s trichrome–stained section of DKO heart shows focal accumulation of collagen (blue) in the endocardium (arrow in D). (E and F) Boxed regions of C and D, respectively, are enlarged. Scale bars: 1 mm (A and B); 20 μm (C–F).

Figure 4

DKO mice exhibit dilated cardiomyopathic features. (A) Representative examples of M-mode echocardiographic images of LV from each group of mice at 10 weeks of age. LVEDD, LV end-diastolic dimension; LVESD, LV end-systolic dimension. (B) Heart weight/body weight (HW/BW) ratio of control mice (n = 9), Gab1CKO mice (n = 6), Gab2KO mice (n = 6), and DKO mice (n = 10) at 10 weeks of age. (C) LVEDD, (D) fractional shortening (%FS), and (E) interventricular septal thickness (IVST) of control mice (n = 8), Gab1CKO mice (n = 8), Gab2KO mice (n = 7), and DKO mice (n = 14) at 10 weeks of age. There were no significant differences in BW or heart rate among the 4 groups. (F) The maximum first derivative of LV pressure (LV dP/dt_max) and (G) the minimum first derivative of LV pressure (LV dP/dt_min) were obtained by catheterization of LV from right carotid artery in control mice (n = 7), Gab1CKO mice (n = 6), Gab2KO mice (n = 7), and DKO mice (n = 7) at 12 weeks of age. *P < 0.01 compared with all other genotypes.

A significant accumulation of elastic fibers and collagen was observed exclusively in the endocardium of DKO mice (Figure 3, B–F), while fibrotic replacement was not found in the interstitial spaces of the ventricles of DKO mice (Supplemental Figure 6, A and B). There was no significant increase in the number of apoptotic myocardial cells in the hearts of DKO mice compared with those of control mice (Supplemental Figure 7, A and B). The endocardial deposition of elastic fibers and collagen was not found in the neonates of DKO, but was found to some extent in all of the DKO mice after 3 weeks (Supplemental Figure 5A and data not shown). These endocardium-specific changes were coincident with the pathological features of endocardial fibroelastosis (EFE), the genetic causality of which has not been fully elucidated to date (37, 38). We further examined the vasculature in the heart by immunostaining with anti-vWF Ab. Intriguingly, we found abnormally dilated vessels positively stained with anti-vWF Ab exclusively in the LV of DKO mice but not in those of control, Gab1CKO, or Gab2KO mice (Figure 5A). These dilated vessels in DKO mice exhibited the impairment in recruitment of α-SMA–positive VSMCs (Figure 5, B and C). These findings indicate that the maintenance system for both endocardial and vascular endothelium might be disturbed in the DKO mouse hearts. Furthermore, EFE and increased abnormal vessels in the hearts of DKO mice were indirectly ascribed to the lack of Gab1 and Gab2 in the myocardium because there was no abnormality in the other 3 groups.

Figure 5

DKO mice display vascular abnormalities in the ventricles. (A) Heart sections from 4 groups of mice at 6 weeks of age were immunostained with anti-vWF Ab. vWF-positive, abnormally dilated vessels were observed in the left ventricles of DKO mice (arrows) but not in those of control, Gab1CKO, or Gab2KO mice. Representative photographs are shown. (B and C) Heart sections from control (B) and DKO (C) mice at 6 weeks of age were immunostained with anti-vWF and anti–α-SMA Abs. The abnormally dilated vessels in DKO mice were not surrounded by α-SMA–positive VSMCs in most cases (C, top panels), although vessels of normal diameter near the epicardium in DKO mice were surrounded by α-SMA–positive VSMCs (C, bottom panels) as observed in control mice (B). Representative images are shown. Scale bars: 200 μm (A); 20 μm (B and C).

We assessed in vivo cardiac function by echocardiography and cardiac catheterization. Echocardiography revealed a significant increase in LV end-diastolic dimension (Figure 4, A and C), decreased fractional shortening (Figure 4, A and D), and decreased interventricular septal wall thickness (Figure 4E) in 10-week-old DKO mice compared with age-matched mice of the other 3 groups. Although we did not find a significant changes of LV end-diastolic dimension or fractional shortening between the DKO and control mice at 3 weeks of age, we did observe these changes after 6 weeks of age (Supplemental Figure 8, A and B). Consistent with the echocardiographic findings, cardiac catheterization at 12 weeks of age revealed a marked reduction of the maximum first derivative of LV pressure exclusively in DKO (Figure 4F), demonstrating a reduction in myocardial contractility of the DKO hearts. The accompanying reduction of the minimum first derivative of LV pressure in the DKO mouse hearts indicated the impairment of LV relaxation (Figure 4G). There were no significant differences in heart rate or LV peak pressure among the 4 groups (data not shown). This relaxation failure was supported by the electron microscopic findings. We noticed that sarcomere length was reduced in the DKO mouse hearts, which indicated the hypercontraction phenotype (39), although we could detect slight changes in the mitochondria of DKO mouse hearts (Supplemental Figure 7, C and D). In agreement with the reduced contractility and relaxation reflecting heart failure, the fetal cardiac gene program was reactivated, as evidenced by the significant increase in both atrial natriuretic peptide (ANP) and skeletal α-actin (α-SKA) mRNAs in DKO mice (Figure 6, A–C).

Figure 6

DKO mice die of heart failure. (A) Northern blot analyses of the hearts from control, Gab1CKO, Gab2KO, and DKO mice (n = 3 for each group) at 12–14 weeks of age showed the increased expression of mRNAs for ANP and α-SKA in DKO mice. GAPDH mRNA was also measured for sample loading control. (B and C) The relative levels of ANP and α-SKA mRNA (normalized to GAPDH mRNA levels) were quantified from 3 mouse hearts in each group. (*P < 0.01 compared with all other groups.) (D) Kaplan-Meier curves showing survival rate in control mice (n = 30), Gab1CKO mice (n = 30), Gab2KO mice (n = 30), and DKO mice (n = 66) mice by 500 days. The number of dead DKO mice was 48 (72.7%); P < 0.001 for DKO versus control, Gab1CKO, and Gab2KO mice by log-rank test.

Approximately 70% of the DKO mice died, presumably of heart failure accompanied by pleural effusion, between 3 and 72 weeks of age (Figure 6D). We observed remarkably dilated ventricles in DKO mice that had died of heart failure (Supplemental Figure 5B, right panel). The other 3 groups of mice lived normally during the observation period of 500 days (Figure 6D). In agreement with this survival analysis, we did not observe any enlargement of the hearts of Gab1CKO and Gab2KO mice at 300 and 500 days of age. (Supplemental Figure 5B and data not shown). These data indicate that depletion of both Gab1 and Gab2 in the myocardium result in DCM-like phenotype accompanied by EFE.

Gab1 and Gab2 are required for NRG-1β/ErbB signaling in the heart. To determine requirements of Gab1 and Gab2 in NRG-1β–triggered signaling in the myocardium, we examined the activation of ERK and AKT after injection of NRG-1β. NRG-1β–induced activation of ERK and AKT was completely abrogated in DKO mice but not in the other 3 groups (Figure 7, A–C), suggesting a compensatory function of Gab1 and Gab2 in the heart. Consistently, tyrosine phosphorylation of Gab1 and subsequent association with SHP2 and p85 were observed in control and Gab2KO mice but not in Gab1CKO or DKO mice (Figure 7D). Tyrosine phosphorylation of Gab2 and subsequent association with SHP2 and p85 were conversely observed in control or Gab1CKO mice but not in Gab2KO or DKO mice (Figure 7E). Tyrosine phosphorylation of ErbB2 and ErbB4 was comparable among the 4 groups (Figure 7F). IGF-1– and HB-EGF–dependent activation of ERK and AKT were not affected in the hearts of DKO mice (Supplemental Figure 9, A and B). These data indicate that Gab1 and Gab2 are required exclusively for NRG-1β/ErbB signal-dependent activation of ERK and AKT in the heart.

Figure 7

Gab1 and Gab2 are required for NRG-1β–dependent ERK and AKT activation in the heart. (A) NRG-1β–induced activation of ERK and AKT in the hearts from the indicated mice was assessed using phospho-specific Abs. Activation of ERK and AKT was exclusively attenuated in DKO hearts compared with the other 3 groups. Representative blots of 4 experiments are shown. (B) Phosphorylation of ERK was quantified against total ERK (n = 4). (C) Phosphorylation of AKT was quantified against total AKT (n = 4). *P < 0.05, **P < 0.01 for the indicated groups. Tyrosine phosphorylation of Gab1 (D) and Gab2 (E) and their association with SHP2 and p85 in hearts from the 4 groups of mice after injection with NRG-1β was examined as in Figure 1, A and B. Arrows in D denote the 2 isoforms of Gab1. (F) Tyrosine phosphorylation of ErbB2 (upper panels) and ErbB4 (lower panels) in hearts from the 4 groups were assessed at 5 minutes after NRG-1β injection. Tyrosine phosphorylation of ErbB receptors in the murine hearts upon NRG-1β stimulation was examined by IP with anti-ErbB2 or anti-ErbB4 Ab, followed by IB with the Abs indicated at the left.

Angiopoietin 1 upregulation induced by NRG-1β is impaired in Gab1/Gab2-deficient myocardium. Because we observed no cardiac abnormalities in Gab2KO mice, we determined that the primary cause of EFE and abnormal vessels in DKO mouse hearts was not the lack of Gab2 in endothelial cells. To identify the potential signal defect that caused EFE and malformed vessels downstream of the NRG-1β/ErbB–Gab1/Gab2 signaling pathway in the myocardium, we used microarrays to carry out a global survey of mRNA in control and DKO mice treated with or without NRG-1β for 8 hours. We found several transcripts that were upregulated by stimulation with NRG-1β in the hearts of control mice but not in those of DKO mice (Figure 8A). Among these transcripts presented in the cluster diagram, we considered thrombospondin 1 (TSP1) and angiopoietin 1 (Ang1) to be potential paracrine factors from myocardium and Eph receptor A4 (EphA4) to be important for the intercellular communication between cardiomyocytes and surrounding cells.

Figure 8

Gab1 and Gab2 are required for the NRG-1β–induced Ang1 upregulation and endothelial stabilization in the heart. (A) RNAs from the ventricles of control and DKO mice (n = 3 per group) were prepared 8 hours after injection with NRG-1β or vehicle. This preparation was performed twice. We used pooled RNAs from 3 mice and performed Affymetrix DNA microarrays independently 2 times (indicated as 1st and 2nd). Cluster analysis was performed of upregulated (red) and downregulated (green) genes in NRG-1β–treated control and DKO mice. Color intensity is relative to the median (black). (B) Northern blot analysis demonstrated the upregulation of Ang1 mRNA in the hearts of control mice but not in those of DKO mice following injection with NRG-1β (n = 3 per group). GAPDH mRNA was checked for gel loading. (C) Quantitative analysis of Ang1 mRNA (normalized to GAPDH mRNA) (n = 3 per group; **P < 0.01 between the indicated groups). (D and E) Ang1 expression was upregulated in cardiomyocytes but not in noncardiomyocytes after stimulation with NRG-1β. (F) Quantitative analysis of Ang1 mRNA normalized to GAPDH mRNA (n = 4; **P < 0.01). (G) Cryosections of the hearts from control and DKO mice were immunostained with anti-CD31 Ab. Representative results are shown. Scale bars: 50 μm. (H) Capillary densities were counted from the number of capillaries per square millimeter. **P < 0.01.

To address the pathogenesis of endocardial and vascular abnormalities observed in DKO mouse hearts, we focused on Ang1 because it has an important role in maturation of both vascular endothelium and endocardial endothelium in vivo (40–42). We confirmed by northern blot analysis that NRG-1β upregulated Ang1 mRNA in the hearts of control mice, but not DKO mice (Figure 8, B and C). NRG-1β consistently induced significant upregulation of Ang1 mRNA in cultured cardiomyocytes but not in noncardiomyocytes (Figure 8, D–F). In association with defective expression of Ang1, CD31-positive capillary density was significantly decreased in the LV of DKO mice compared with control (Figure 8, G and H). Taken together, these findings suggest that the lack of NRG-1β–induced upregulation of Ang1 might be one of the possible causes for pathogenesis of EFE and abnormal vasculatures in DKO mouse hearts.

To our knowledge, the present study is the first to reveal the essential roles of Gab family proteins for NRG-1β/ErbB signaling pathway in the heart. Gab1 and Gab2 were markedly tyrosine phosphorylated in the myocardium after stimulation with NRG-1β among various growth factors and cytokines. Tyrosine-phosphorylated Gab1 and Gab2 subsequently associated with SHP2 and p85, resulting in strong activation of both ERK and AKT in the myocardium. NRG-1β–dependent activation of ERK and AKT was almost completely abrogated in the DKO mouse hearts. In agreement with NRG-1β–dependent downstream signaling defects, DKO mice displayed DCM-like phenotypes and EFE with aging. Interestingly, DKO mouse hearts also displayed abnormally dilated vessels with the loss of VSMCs. To address the mechanism for the abnormality in endocardial/vascular endothelium in DKO mouse hearts, we performed DNA microarray analysis and found several vasculature-regulating gene transcripts, such as Ang1, upregulated by NRG-1β in control, but not in DKO, mouse hearts. Thus, Gab family proteins mediate NRG-1β–dependent stabilization of endocardial/vascular endothelium through the paracrine system from cardiomyocytes in the heart.

Gab1 and Gab2 are specifically required for coordination of NRG-1β/ErbB–dependent signaling pathway in the myocardium. NRG-1β shed from endothelial cells activates ErbB2/ErbB4 heterodimer or ErbB4 homodimer on the cardiomyocytes (16, 17, 21). Consistent with this notion, we found that NRG-1β induced prominent tyrosine phosphorylation of Gab1 and Gab2 in cardiomyocytes but not in noncardiomyocytes. In addition, the cardiomyocyte-specific, high-MW isoform of Gab1 was tyrosine phosphorylated after stimulation with NRG-1β but not with other agonists including HB-EGF and EGF. It has been reported that HB-EGF–deficient mice develop heart failure (25, 26). Given that HB-EGF induced a much stronger tyrosine phosphorylation of Gab1 and Gab2 in noncardiomyocytes than in cardiomyocytes in our study and that valvular structures are developed from noncardiomyocytes (19), the heart failure observed in HB-EGF–deficient mice might have resulted from abnormal signaling in the development of the valvular apparatus. Therefore, the cardiac phenotypes observed in DKO mice were mainly ascribable to the defects of the NRG-1β/ErbB signaling pathway in the myocardium. Consistent with this, similar DCM-like phenotypes are found in cardiac-specific ErbB2- and ErbB4-deficient mice (29–31).

NRG-1β activates both ERK and PI3K/AKT pathways in cardiomyocytes in vitro, both of which have been implicated in modulation of cell survival and protein synthesis (21, 43). NRG-1β actually induced strong activation of ERK and AKT in the hearts of control, but not DKO, mice. This finding provides what we believe to be the first in vivo evidence that Gab1 and Gab2 are required for transmission of the NRG-1β/ErbB signal to downstream signaling pathways, ERK and AKT. DKO mice progressively developed DCM phenotypes, demonstrating clearly that Gab1 and Gab2 were essential for maintenance of myocardial function through transmission of NRG-1β/ErbB signaling pathway (Figure 9).

Figure 9

Schematic illustration of the roles of Gab family proteins in the myocardium. NRG-1β shed from the endocardial or capillary endothelium in the heart activates ErbB receptors on the myocardium, resulting in tyrosine phosphorylation of Gab family proteins and subsequent activation of ERK and AKT. NRG-1β/ErbB–Gab1/Gab2 signaling in the myocardium is directly required for postnatal maintenance of myocardial function. In addition, NRG-1β/ErbB–Gab1/Gab2 signaling indirectly contributes to postnatal stabilization of capillary or endocardial endothelium, possibly through Ang1 upregulation (dotted line).

DKO mice also exhibited abnormal deposition of elastic fibers and collagen specifically in the endocardium, reminiscent of the pathological features observed in primary EFE. Clinically, primary EFE is found mainly in infants, children, and adolescents and is frequently accompanied by contractile deterioration similar to DCM. Although there have been some reports suggesting the heritable causality of primary EFE (37, 38), the precise pathogenetic mechanisms have not been elucidated to date. These DKO mice may provide the first mouse model of EFE. Further genetic analysis of cardiac-specific isoform of Gab1 will certainly contribute to our understanding of the pathogenesis of EFE.

DKO mouse hearts also displayed abnormal vasculatures as well as EFE. Microarray analysis enabled us to identify several transcripts that were upregulated by NRG-1β in the control hearts but not in DKO hearts. Among these transcripts selected in the cluster analysis, TSP1, EphA4, and Ang1 have been reported to be involved in the intercellular-dependent vascular regulation (40, 44, 45). Intriguingly, NRG-1β/ErbB2/ErbB4 signaling, Ang1/Tie2 signaling, VEGF/VEGFR2 signaling, and serotonin-mediated (5-HT_2B–mediated) signaling are required for the proper maturation of endocardium (16, 17, 40, 46, 47). Moreover, Ang1- or Tie2-deficient mice exhibit embryonic lethality accompanied by abnormally dilated vessels as well as defects in the endocardium (40, 42, 48). Furthermore, we demonstrated for the first time that postnatal cardiomyocytes are important Ang1-producing cells, whereas Ang1 has been believed to be mainly secreted from vascular mural cells such as pericytes and VSMCs (40, 41). Thus, we could propose that the defective expression of Ang1 might be involved in the pathogenesis of EFE and abnormal vessels in DKO hearts, though we cannot exclude the possibility that other vasculature-regulating genes, such as TSP1 and EphA4, play important roles in endocardial maintenance. Cardiac-specific gene ablation of Ang1 would be helpful to understand its importance in cardiomyocyte-endothelial cell interactions.

So far, it has been well established that NRG-1 functions as a cytoprotective growth factor in cardiomyocytes (17, 21, 43). Here, our findings propose a novel function of NRG-1; NRG-1 regulates vascular homeostasis through the paracrine expression of endothelium stabilization factors, such as Ang1, via Gab family proteins. Importantly, accumulating evidence has revealed that normal endothelial function is required for the maintenance of myocardial function (16). Collectively, Gab1 and Gab2 in the myocardium are essential for both maintenance of myocardial function and stabilization of capillary or endocardial endothelium through transmission of NRG-1β/ErbB signaling (Figure 9).

View Supplemental data

This work was supported in part by grants from the Ministry of Education, Culture, Sports, Science, and Technology of Japan (to Y. Nakaoka and N. Mochizuki); the Ministry of Health, Labour, and Welfare of Japan (to N. Mochizuki); the Program for the Promotion of Fundamental Studies in Health Sciences of the National Institute of Biomedical Innovation (to T. Minami, T. Kodama, and N. Mochizuki); the Japan Heart Foundation (Novartis Grant for Research Award on Molecular and Cellular Cardiology, to Y. Nakaoka); the Mitsubishi Pharma Research Foundation (to Y. Nakaoka); and the Uehara Memorial Foundation (to Y. Nakaoka). We thank Y. Matsuura, M. Yoshida, M. Miyabayashi, M. Maeoka, Y. Ohba, M. Sato, K. Sako, and K. Shioya for their technical assistance; S. Higashiyama, O. Nakagawa, and O. Ohara for helpful comments; K. Komamura for advice on echocardiography; and W.W. Hall and K. Yamauchi-Takihara for critical reading of the manuscript.

Address correspondence to: Naoki Mochizuki, Department of Structural Analysis, National Cardiovascular Center Research Institute, 5-7-1 Fujishirodai, Suita, Osaka 565-8565, Japan. Phone: 81-6-6833-5012, ext. 2508; Fax: 81-6-6835-5461; E-mail: nmochizu@ri.ncvc.go.jp. Or to: Yoshikazu Nakaoka: Department of Cardiovascular Medicine, Osaka University Graduate School of Medicine, 2-2, Yamada-oka, Suita, Osaka 565-0871, Japan. Phone: 81-6-6879-3835; Fax: 81-6-6879-3839; E-mail: ynakaoka@imed3.med.osaka-u.ac.jp.

Nonstandard abbreviations used: Ang1, angiopoietin 1; ANP, atrial natriuretic peptide; DCM, dilated cardiomyopathy; DKO, cardiomyocyte-specific Gab1/Gab2 double knockout; EFE, endocardial fibroelastosis; EphA4, Eph receptor A4; ErbB, erythroblastic leukemia viral oncogene; Gab, Grb2-associated binder; Gab1CKO, cardiomyocyte-specific Gab1 conditional knockout; Gab1KO, conventional Gab1 knockout; HB-EGF, heparin-binding EGF-like growth factor; LIF, leukemia inhibitory factor; α-MHC, α-myosin heavy chain; NRG-1, neuregulin-1; α-SKA, skeletal α-actin; TSP1, thrombospondin 1.

Conflict of interest: The authors have declared that no conflict of interest exists.

Reference information: J. Clin. Invest.117:1771–1781 (2007). doi:10.1172/JCI30651

Y. Nakaoka’s present address is: Department of Cardiovascular Medicine, Osaka University Graduate School of Medicine, Osaka, Japan.

Hisao Hirota is deceased.

1. Seidman, J.G., Seidman, C. 2001. The genetic basis for cardiomyopathy: from mutation identification to mechanistic paradigms. Cell. 104:557-567.
   View this article via: CrossRef PubMed Google Scholar
2. Chien, K.R. 1999. Stress pathways and heart failure. Cell. 98:555-558.
   View this article via: CrossRef PubMed Google Scholar
3. Schlessinger, J., Lemmon, M.A. 2003. SH2 and PTB domains in tyrosine kinase signaling. Sci. STKE. 2003:RE12.
   View this article via: PubMed Google Scholar
4. Gu, H., Neel, B.G. 2003. The “Gab” in signal transduction. Trends Cell Biol. 13:122-130.
   View this article via: CrossRef PubMed Google Scholar
5. Nishida, K., Hirano, T. 2003. The role of Gab family scaffolding adapter proteins in the signal transduction of cytokine and growth factor receptors. Cancer Sci. 94:1029-1033.
   View this article via: CrossRef PubMed Google Scholar
6. Itoh, M., et al. 2000. Role of Gab1 in heart, placenta, and skin development and growth factor- and cytokine-induced extracellular signal-regulated kinase mitogen-activated protein kinase activation. Mol. Cell. Biol. 20:3695-3704.
   View this article via: CrossRef PubMed Google Scholar
7. Sachs, M., et al. 2000. Essential role of Gab1 for signaling by the c-Met receptor in vivo. J. Cell Biol. 150:1375-1384.
   View this article via: CrossRef PubMed Google Scholar
8. Gu, H., et al. 2001. Essential role for Gab2 in the allergic response. Nature. 412:186-190.
   View this article via: CrossRef PubMed Google Scholar
9. Nishida, K., et al. 2002. Requirement of Gab2 for mast cell development and KitL/c-Kit signaling. Blood. 99:1866-1869.
   View this article via: CrossRef PubMed Google Scholar
10. Wada, T., et al. 2005. The molecular scaffold Gab2 is a crucial component of RANK signaling and osteoclastogenesis. Nat. Med. 11:394-399.
    View this article via: CrossRef PubMed Google Scholar
11. Nishida, K., et al. 2005. FcepsilonRI-mediated mast cell degranulation requires calcium-independent microtubule-dependent translocation of granules to the plasma membrane. J. Cell Biol. 170:115-126.
    View this article via: CrossRef PubMed Google Scholar
12. Seiffert, M., et al. 2003. Gab3-deficient mice exhibit normal development and hematopoiesis and are immunocompetent. Mol. Cell. Biol. 23:2415-2424.
    View this article via: CrossRef PubMed Google Scholar
13. Nakaoka, Y., et al. 2003. Activation of gp130 transduces hypertrophic signal through interaction of scaffolding/docking protein Gab1 with tyrosine phosphatase SHP2 in cardiomyocytes. Circ. Res. 93:221-229.
    View this article via: CrossRef PubMed Google Scholar
14. Bentires-Alj, M., et al. 2006. A role for the scaffolding adapter GAB2 in breast cancer. Nat. Med. 12:114-121.
    View this article via: CrossRef PubMed Google Scholar
15. Yamasaki, S., et al. 2003. Gab1 is required for EGF receptor signaling and the transformation by activated ErbB2. Oncogene. 22:1546-1556.
    View this article via: CrossRef PubMed Google Scholar
16. Brutsaert, D.L. 2003. Cardiac endothelial-myocardial signaling: its role in cardiac growth, contractile performance, and rhythmicity. Physiol. Rev. 83:59-115.
    View this article via: PubMed Google Scholar
17. Garratt, A.N., Ozcelik, C., Birchmeier, C. 2003. ErbB2 pathways in heart and neural diseases. Trends Cardiovasc. Med. 13:80-86.
    View this article via: CrossRef PubMed Google Scholar
18. Falls, D.L. 2003. Neuregulins: functions, forms, and signaling strategies. Exp. Cell Res. 284:14-30.
    View this article via: CrossRef PubMed Google Scholar
19. Iwamoto, R., Mekada, E. 2006. ErbB and HB-EGF signaling in heart development and function. Cell Struct. Funct. 31:1-14.
    View this article via: CrossRef PubMed Google Scholar
20. Lemmens, K., Segers, V.F., Demolder, M., De Keulenaer, G.W. 2006. Role of neuregulin-1/ErbB2 signaling in endothelium-cardiomyocyte cross-talk. J. Biol. Chem. 281:19469-19477.
    View this article via: CrossRef PubMed Google Scholar
21. Zhao, Y.Y., et al. 1998. Neuregulins promote survival and growth of cardiac myocytes. Persistence of ErbB2 and ErbB4 expression in neonatal and adult ventricular myocytes. J. Biol. Chem. 273:10261-10269.
    View this article via: CrossRef PubMed Google Scholar
22. Gassmann, M., et al. 1995. Aberrant neural and cardiac development in mice lacking the ErbB4 neuregulin receptor. Nature. 378:390-394.
    View this article via: CrossRef PubMed Google Scholar
23. Lee, K.F., et al. 1995. Requirement for neuregulin receptor erbB2 in neural and cardiac development. Nature. 378:394-398.
    View this article via: CrossRef PubMed Google Scholar
24. Meyer, D., Birchmeier, C. 1995. Multiple essential functions of neuregulin in development. Nature. 378:386-390.
    View this article via: CrossRef PubMed Google Scholar
25. Iwamoto, R., et al. 2003. Heparin-binding EGF-like growth factor and ErbB signaling is essential for heart function. Proc. Natl. Acad. Sci. U. S. A. 100:3221-3226.
    View this article via: CrossRef PubMed Google Scholar
26. Jackson, L.F., et al. 2003. Defective valvulogenesis in HB-EGF and TACE-null mice is associated with aberrant BMP signaling. EMBO J. 22:2704-2716.
    View this article via: CrossRef PubMed Google Scholar
27. Slamon, D.J., et al. 2001. Use of chemotherapy plus a monoclonal antibody against HER2 for metastatic breast cancer that overexpresses HER2. N. Engl. J. Med. 344:783-792.
    View this article via: CrossRef PubMed Google Scholar
28. Suter, T.M., Cook-Bruns, N., Barton, C. 2004. Cardiotoxicity associated with trastuzumab (Herceptin) therapy in the treatment of metastatic breast cancer. Breast. 13:173-183.
    View this article via: CrossRef PubMed Google Scholar
29. Crone, S.A., et al. 2002. ErbB2 is essential in the prevention of dilated cardiomyopathy. Nat. Med. 8:459-465.
    View this article via: CrossRef PubMed Google Scholar
30. Garcia-Rivello, H., et al. 2005. Dilated cardiomyopathy in Erb-b4-deficient ventricular muscle. Am. J. Physiol. Heart Circ. Physiol. 289:H1153-H1160.
    View this article via: CrossRef PubMed Google Scholar
31. Ozcelik, C., et al. 2002. Conditional mutation of the ErbB2 (HER2) receptor in cardiomyocytes leads to dilated cardiomyopathy. Proc. Natl. Acad. Sci. U. S. A. 99:8880-8885.
    View this article via: CrossRef PubMed Google Scholar
32. Nakagawa, O., et al. 1995. Rapid transcriptional activation and early mRNA turnover of brain natriuretic peptide in cardiocyte hypertrophy. Evidence for brain natriuretic peptide as an “emergency” cardiac hormone against ventricular overload. J. Clin. Invest. 96:1280-1287.
    View this article via: JCI PubMed Google Scholar
33. Cote, G.M., Miller, T.A., Lebrasseur, N.K., Kuramochi, Y., Sawyer, D.B. 2005. Neuregulin-1alpha and beta isoform expression in cardiac microvascular endothelial cells and function in cardiac myocytes in vitro. Exp. Cell Res. 311:135-146.
    View this article via: CrossRef PubMed Google Scholar
34. Agah, R., et al. 1997. Gene recombination in postmitotic cells. Targeted expression of Cre recombinase provokes cardiac-restricted, site-specific rearrangement in adult ventricular muscle in vivo. J. Clin. Invest. 100:169-179.
    View this article via: JCI PubMed Google Scholar
35. Yamaguchi, O., et al. 2004. Cardiac-specific disruption of the c-raf-1 gene induces cardiac dysfunction and apoptosis. J. Clin. Invest. 114:937-943.
    View this article via: JCI PubMed Google Scholar
36. Meng, S., Chen, Z., Munoz-Antonia, T., Wu, J. 2005. Participation of both Gab1 and Gab2 in the activation of the ERK/MAPK pathway by epidermal growth factor. Biochem. J. 391:143-151.
    View this article via: CrossRef PubMed Google Scholar
37. Moss, A.J., and Adams, F.H. 1995. Moss’ heart disease in infants, children, and adolescents. Williams & Wilkins. Baltimore, Maryland, USA. 1,085 pp.
38. Westwood, M., Harris, R., Burn, J.L., Barson, A.J. 1975. Heredity in primary endocardial fibroelastosis. Br. Heart J. 37:1077-1084.
    View this article via: CrossRef PubMed Google Scholar
39. Xu, X., et al. 2005. ASF/SF2-regulated CaMKIIdelta alternative splicing temporally reprograms excitation-contraction coupling in cardiac muscle. Cell. 120:59-72.
    View this article via: CrossRef PubMed Google Scholar
40. Brindle, N.P., Saharinen, P., Alitalo, K. 2006. Signaling and functions of angiopoietin-1 in vascular protection. Circ. Res. 98:1014-1023.
    View this article via: CrossRef PubMed Google Scholar
41. Davis, S., et al. 1996. Isolation of angiopoietin-1, a ligand for the TIE2 receptor, by secretion-trap expression cloning. Cell. 87:1161-1169.
    View this article via: CrossRef PubMed Google Scholar
42. Suri, C., et al. 1996. Requisite role of angiopoietin-1, a ligand for the TIE2 receptor, during embryonic angiogenesis. Cell. 87:1171-1180.
    View this article via: CrossRef PubMed Google Scholar
43. Baliga, R.R., et al. 1999. NRG-1-induced cardiomyocyte hypertrophy. Role of PI-3-kinase, p70(S6K), and MEK-MAPK-RSK. Am. J. Physiol. 277:H2026-H2037.
    View this article via: PubMed Google Scholar
44. Lawler, J. 2000. The functions of thrombospondin-1 and-2. Curr. Opin. Cell Biol. 12:634-640.
    View this article via: CrossRef PubMed Google Scholar
45. Ogita, H., et al. 2003. EphA4-mediated Rho activation via Vsm-RhoGEF expressed specifically in vascular smooth muscle cells. Circ. Res. 93:23-31.
    View this article via: CrossRef PubMed Google Scholar
46. Liao, W., et al. 1997. The zebrafish gene cloche acts upstream of a flk-1 homologue to regulate endo­thelial cell differentiation. Development. 124:381-389.
    View this article via: PubMed Google Scholar
47. Nebigil, C.G., et al. 2000. Serotonin 2B receptor is required for heart development. Proc. Natl. Acad. Sci. U. S. A. 97:9508-9513.
    View this article via: CrossRef PubMed Google Scholar
48. Sato, T.N., et al. 1995. Distinct roles of the receptor tyrosine kinases Tie-1 and Tie-2 in blood vessel formation. Nature. 376:70-74.
    View this article via: CrossRef PubMed Google Scholar
49. Nishida, K., et al. 1999. Gab-family adapter proteins act downstream of cytokine and growth factor receptors and T- and B-cell antigen receptors. Blood. 93:1809-1816.
    View this article via: PubMed Google Scholar
50. Osugi, T., et al. 2002. Cardiac-specific activation of signal transducer and activator of transcription 3 promotes vascular formation in the heart. J. Biol. Chem. 277:6676-6681.
    View this article via: CrossRef PubMed Google Scholar
51. Taniguchi, M., et al. 1998. Efficient production of Cre-mediated site-directed recombinants through the utilization of the puromycin resistance gene, pac: a transient gene-integration marker for ES cells. Nucleic Acids Res. 26:679-680.
    View this article via: CrossRef PubMed Google Scholar
52. Georgakopoulos, D., et al. 1998. In vivo murine left ventricular pressure-volume relations by miniaturized conductance micromanometry. Am. J. Physiol. 274:H1416-H1422.
    View this article via: PubMed Google Scholar
53. Nishio, R., Sasayama, S., Matsumori, A. 2002. Left ventricular pressure-volume relationship in a murine model of congestive heart failure due to acute viral myocarditis. J. Am. Coll. Cardiol. 40:1506-1514.
    View this article via: CrossRef PubMed Google Scholar
54. Minami, T., Miura, M., Aird, W.C., Kodama, T. 2006. Thrombin-induced autoinhibitory factor, Down syndrome critical region-1, attenuates NFAT-dependent vascular cell adhesion molecule-1 expression and inflammation in the endothelium. J. Biol. Chem. 281:20503-20520.
    View this article via: PubMed CrossRef Google Scholar