Activation of gp130 Transduces Hypertrophic Signal Through Interaction of Scaffolding/Docking Protein Gab1 With Tyrosine Phosphatase SHP2 in Cardiomyocytes

Construction of Recombinant Adenovirus

According to a previous study,^9,12 the most C-terminal 2 tyrosine residues (Tyr-627 and Tyr-659) of Gab1 are required for binding to SHP2. Substitution of these tyrosine residues by phenylalanine renders the molecule incapable of binding to SHP2. The wild-type and mutated human Gab1 cDNAs are designated as Gab1^WT and Gab1^F627/659, respectively. A dominant-negative form of murine ERK5 (ERK5^AEF) was created by mutating dual phosphorylation site (Thr-219 and Tyr-221 with alanine and phenylalanine, respectively).²⁹ The adenovirus vectors expressing Gab1^WT (AdGab1^WT), Gab1^F627/659 (AdGab1^F627/659), and ERK5^AEF (AdERK5^AEF) were generated according to a protocol described elsewhere.³⁰

Cell Culture and Protocol for Adenovirus Infection

Primary cultures of neonatal rat cardiomyocytes were prepared from ventricles of 1-day-old Wistar rats (Kiwa Jikken Dobutsu) as described previously.³¹ At 16 hours after plating, cardiomyocytes were infected with adenovirus diluted in Medium-199 with 2% FBS at a multiplicity of infection (moi) of 20 and incubated for 8 hours. In the dual infection of adenovirus vectors, cardiomyocytes were infected with each virus at an moi of 10. After removal of viral suspension, cardiomyocytes were serum starved for 16 hours and stimulated with reagents. Infection efficiency, determined by lacZ gene expression in cultured cardiomyocytes, is consistently >90% with this method. Adenovirus vector expressing β-galactosidase (Adβ-gal) was used as a control.

Immunoprecipitation and Immunoblotting

The methods of immunoprecipitation were essentially as described previously.⁵ After stimulation, cells were immediately lysed in lysis buffer (20 mmol/L Tris-HCl [pH 7.4], 150 mmol/L NaCl, 1% Nonidet P-40, 500 μmol/L sodium vanadate, 1 mmol/L dithiothreitol, 5 μg/mL aprotinin, 5 μg/mL leupeptin, and 1 mmol/L phenylmethylsulfonyl fluoride). The cleared lysates were incubated with 2 μL of anti-Gab1 serum or 4 μL of anti-SHP2 antibody and 20 μL of protein A-Sepharose for 12 hours at 4°C. Collected immune complexes were eluted with 20 μL of 2×Laemmli’s SDS loading buffer, separated on a 4% to 20% gradient polyacrylamide gel (Dai-ichi Kagaku), electrotransferred to a polyvinylidene difluoride membrane (Immobilon-P; Millipore), and processed for immunoblotting analysis essentially as described previously.⁵ The ECL system was used for detection.

Northern Blot Analysis

Northern blot analysis was performed as previously described.²¹ Total cellular RNA (8 μg) was loaded in each lane and size fractionated by 1% formaldehyde-agarose gel electrophoresis. The probes for BNP, SKA, and GAPDH were kindly donated by Dr K.R. Chien (University of California, San Diego, Calif).

Immunofluorescence

For immunofluorescence, cardiomyocytes were grown on glass coverslips coated with gelatin. Cells were incubated in the presence or absence of LIF 1×10³ U/mL for 24 hours. Cells were fixed with 2% formaldehyde and permeabilized with 0.1% Triton X-100. Cells were incubated with monoclonal anti-sarcomeric α-actinin antibody, followed by incubation with fluorescein-conjugated goat anti-mouse secondary antibody. Cardiomyocytes stained against sarcomeric α-actinin were viewed by fluorescence microscopy. Cell size was estimated by measuring the area over which individual sarcomeric α-actinin–positive cells had attached (planimetry), and cell length and cell width were determined as described previously.²⁰

Statistics

Statistical analysis was performed with Student’s t test. Values of P<0.05 were considered significant.

An expanded Materials and Methods section can be found in the online data supplement available at http://www.circresaha.org.

Gab1 and SHP2 Are Tyrosine Phosphorylated by LIF, and Phosphorylated Gab1 Interacts With SHP2 and p85 in Cardiomyocytes

To examine which ligand induces tyrosine phosphorylation of Gab1 in cardiomyocytes, cells were incubated with LIF, norepinephrine (NE), ET-1, and Ang II for 5 minutes. Among these, LIF exclusively induced tyrosine phosphorylation of Gab1 (Figure 1A). Furthermore, SHP2, the major binding partner of Gab1, was also tyrosine phosphorylated only by LIF (Figure 1B). Therefore, we focused on the gp130-dependent signaling pathway through Gab1 in cardiomyocytes.

• Download figure
• Open in new tab
• Download powerpoint

Figure 1. LIF induces tyrosine phosphorylation of Gab1 and SHP2 in cardiomyocytes. A, Serum-deprived neonatal rat cardiomyocytes were treated with 1×10³ U/mL LIF, 2 μg/mL NE, 100 nmol/L ET-1, and 100 nmol/L Ang II for 5 minutes. Cell lysates were immunoprecipitated with anti-Gab1 serum followed by immunoblotting with anti-phosphotyrosine antibody (PY99) (top). Blots were reprobed with anti-Gab1 antibody (bottom). B, Cell lysates were immunoprecipitated with anti-SHP2 antibody followed by immunoblotting with PY99 (top). Blots were reprobed with anti-SHP2 antibody (bottom).

Gab1 was tyrosine phosphorylated by LIF in a time-dependent and dose-dependent manner (Figures 2A and 2B). SHP2 was also tyrosine phosphorylated by LIF in a time-dependent and dose-dependent manner (Figures 2C and 2D). To elucidate the association of Gab1 with other SH2-containing molecules, cell lysates from cardiomyocytes treated with LIF were immunoprecipitated with anti-Gab1 serum and subjected to immunoblotting with anti-SHP2 and anti-p85 antibodies. SHP2 and p85 were coprecipitated with Gab1 in response to LIF (Figure 2E). Gab1 was also coprecipitated with SHP2 by the immunoprecipitation with anti-SHP2 antibody (Figure 2F). These results demonstrated that LIF induced tyrosine phosphorylation of Gab1, leading to association of Gab1 with SHP2 and p85 in cardiomyocytes.

• Download figure
• Open in new tab
• Download powerpoint

Figure 2. Tyrosine phosphorylation of Gab1 and SHP2 and complex formation of Gab1 with SHP2 and p85 after stimulation with LIF in cardiomyocytes. A and C, Cardiomyocytes were stimulated with 1×10³ U/mL LIF for the indicated periods of time. Cell lysates were immunoprecipitated with anti-Gab1 serum (A) or anti-SHP2 antibody (C) followed by immunoblotting with PY99. B and D, Cardiomyocytes were stimulated with the indicated concentrations of LIF for 5 minutes. Cell lysates were treated as above. E and F, Cardiomyocytes were stimulated with 1×10³ U/mL LIF for 5 minutes. Tyrosine phosphorylation was detected by immunoprecipitation with anti-Gab1 serum (E) or anti-SHP2 antibody (F) followed by immunoblotting with PY99 (top). Blots were reprobed with anti-SHP2 (E and F), Gab1 (E and F), and p85 antibodies (E) (bottom).

Tyrosine Phosphorylation of Gab1 and Association of Gab1 With SHP2 in AdGab1^WT or AdGab1^F627/659-Treated Cardiomyocytes

We constructed recombinant adenovirus vectors expressing Gab1^WT and Gab1^F627/659. Figure 3A shows a schematic representation of these adenovirus vectors, which are named AdGab1^WT and AdGab1^F627/659. We examined tyrosine phosphorylation of Gab1 in Adβ-gal–treated, AdGab1^WT–treated, or AdGab1^F627/659-treated cardiomyocytes. As shown in Figure 3B, tyrosine phosphorylation of Gab1 and the amount of coprecipitated SHP2 with Gab1 were increased in AdGab1^WT-treated cardiomyocytes after LIF stimulation, compared with Adβ-gal–treated cardiomyocytes. In AdGab1^F627/659-treated cardiomyocytes, tyrosine phosphorylation of Gab1 was observed in the same manner as in AdGab1^WT-treated cardiomyocytes, but SHP2 was not coprecipitated with Gab1. As shown in Figure 3C, Gab1 was not coprecipitated with SHP2 in AdGab1^F627/659-treated cardiomyocytes. These results indicate that Gab1^WT and Gab1^F627/659, which were overexpressed through adenovirus-mediated gene transfer, could function effectively in cardiomyocytes.

• Download figure
• Open in new tab
• Download powerpoint

Figure 3. Generation of adenovirus vectors expressing Gab1^WT and Gab1^F627/659. A, Structures of adenovirus vectors expressing wild-type Gab1 (AdGab1^WT) and mutated Gab1, which could not bind SHP2 (AdGab1^F627/659). Substitution of Tyr-627 and Tyr-659 each by phenylalanine renders the molecule incapable of binding to SHP2 on tyrosine phosphorylation. The transgene is driven under the control of the CMV promoter. Adenovirus vector expressing Adβ-gal was used as a control. MBD indicates c-Met binding domain. B and C, Tyrosine phosphorylation of Gab1 or SHP2 in cardiomyocytes infected with adenovirus vectors and association of Gab1 with SHP2. Cardiomyocytes were infected at an moi of 20 with AdGab1^WT, AdGab1^F627/659, or Adβ-gal. Cardiomyocytes were treated with 1×10³ U/mL LIF for 5 minutes. Tyrosine phosphorylation was detected by immunoprecipitation with anti-Gab1 serum (B) or anti-SHP2 antibody (C) followed by immunoblotting with PY99 (top). Blots were reprobed with anti-Gab1 and SHP2 antibodies.

Elongation of Cardiomyocytes Induced by LIF Is Enhanced in AdGab1^WT-Treated Cardiomyocytes but Suppressed in AdGab1^F627/659-Treated Cardiomyocytes

To elucidate the biological roles of Gab1 in gp130-signaling pathway in cardiomyocytes, we examined the morphological effects of Gab1^WT and Gab1^F627/659 on cardiomyocyte hypertrophy in response to LIF. As shown in Figures 4A and 4D, LIF induced longitudinal elongation in Adβ-gal–treated cardiomyocytes. In AdGab1^WT-treated cardiomyocytes, elongation of cardiomyocytes induced by LIF was enhanced compared with Adβ-gal–treated cardiomyocytes (Figures 4B and 4E). On the contrary, in cardiomyocytes expressing Gab1^F627/659, this morphological change was significantly inhibited (Figures 4C and 4F).

• Download figure
• Open in new tab
• Download powerpoint

Figure 4. Gab1 mediates the longitudinal elongation of cardiomyocytes in response to LIF. Cardiomyocytes were infected at an moi of 20 with Adβ-gal (A and D), AdGab1^WT (B and E), or AdGab1^F627/659 (C and F) for 8 hours. Cells were serum deprived for 16 hours and treated without (A through C) or with 1×10³ U/mL LIF (D through F) for an additional 24 hours before fixation and immunostaining with anti-sarcomeric α-actinin antibody. Representative data are shown. G through J, Cardiomyocytes stained for sarcomeric α-actinin were viewed by fluorescence microscopy. A total of 150 sarcomeric α-actinin–positive cells were examined for each measurement. Cell size was examined by measuring the area to which individual sarcomeric α-actinin–positive cells attached (planimetry). Cell surface areas, length, and width were determined using MacScope software as a relative ratio to those of control Adβ-gal–treated cardiomyocytes without LIF stimulation. For each cardiomyocyte measured, cell length to width ratio was also calculated (J). Values are mean±SD; *P<0.001 vs LIF(−); †P<0.001 vs Adβ-gal; #P<0.001 vs AdGab1^WT. Experiments were repeated 3 times with similar results.

We quantified the cell surface area of cardiomyocytes infected with these adenovirus vectors. As shown in Figure 4G, LIF increased cell surface area by 70% in Adβ-gal–treated or AdGab1^WT-treated cardiomyocytes. However, in AdGab1^F627/659-treated cardiomyocytes, the increase in cell surface area induced by LIF was almost abolished. To characterize the hypertrophic phenotype of these cardiomyocytes, we measured cell length and cell width according to a previously reported method.²⁰ As shown in Figure 4H, cell length was significantly increased in AdGab1^WT-treated cardiomyocytes, compared with that in Adβ-gal–treated cardiomyocytes. In contrast, increase in cell length was significantly suppressed in AdGab1^F627/659-treated cardiomyocytes. Cell width was significantly decreased in response to LIF in AdGab1^WT-treated cardiomyocytes but not altered in Adβ-gal–treated or AdGab1^F627/659-treated cardiomyocytes (Figure 4I). Compared with Adβ-gal–treated cardiomyocytes, the cell length to width ratio was significantly increased in AdGab1^WT-treated cardiomyocytes but suppressed significantly in AdGab1^F627/659-treated cardiomyocytes (Figure 4J). These findings indicate that elongation of cardiomyocytes in response to LIF is enhanced by overexpression of Gab1^WT but suppressed by that of Gab1^F627/659, suggesting that the interaction of Gab1 with SHP2 plays a crucial role in longitudinal elongation of cardiomyocytes in response to LIF.

Gab1 Regulates LIF-Induced Embryonic Gene Expression Through Interaction With SHP2

Reactivation of embryonic phenotype genes, such as atrial natriuretic factor (ANF), BNP, and SKA, is known to be associated with hypertrophic responses in cardiomyocytes. We examined how the interaction of Gab1 with SHP2 contributes to the induction of BNP and SKA mRNA expression after LIF stimulation. BNP mRNA was upregulated by LIF in AdGab1^WT-treated cardiomyocytes to the same extent as in Adβ-gal–treated cardiomyocytes. On the contrary, induction of BNP mRNA was abrogated in AdGab1^F627/659-treated cardiomyocytes (Figures 5A and 5B). In contrast, SKA mRNA expression was slightly increased in response to LIF in Adβ-gal–treated cardiomyocytes but was almost completely suppressed in AdGab1^WT-treated cardiomyocytes both in basal level and after LIF stimulation. In AdGab1^F627/659-treated cardiomyocytes, the expression of SKA mRNA was restored to the same extent as in Adβ-gal–treated cardiomyocytes (Figures 5A and 5C). These results indicate that the interaction of Gab1 with SHP2 is involved in the regulation of embryonic gene expression after stimulation with LIF.

• Download figure
• Open in new tab
• Download powerpoint

Figure 5. Gab1 regulates embryonic gene expression through interaction with SHP2 in cardiomyocytes. A, Cardiomyocytes were infected with Adβ-gal, AdGab1^WT, or AdGab1^F627/659 at an moi of 20 for 8 hours and serum deprived. At 24 hours after infection, cells were treated without or with 1×10³ U/mL LIF for an additional 24 hours. Total RNA was isolated and subjected to Northern blot analysis (8 μg/lane) using BNP and SKA cDNA probes. Equal loading and transfer conditions were confirmed by GAPDH hybridization. B and C, Relative intensity of the bands for BNP or SKA was assessed as the ratio to the intensity of GAPDH. The results were expressed as relative intensity over Adβ-gal–treated cells without LIF stimulation. Values are shown as mean±SD. *P<0.05 vs LIF(−); †P<0.05 vs Adβ-gal (n=3).

Interaction of Gab1 With SHP2 Plays a Crucial Role in Activation of ERK5 by LIF in Cardiomyocytes

To elucidate a potential mechanism in which Gab1-SHP2 interaction plays a role in gp130-mediated longitudinal elongation of cardiomyocytes, we examined the effects of Gab1^WT and Gab1^F627/659 on LIF-induced activation of MAP kinases (ERK5 and ERK1/2), AKT, and signal transducer and activator of transcription 3 (STAT3), which are known to mediate biological functions through gp130.^{21,28,31–33} These signaling molecules were rapidly activated by LIF in Adβ-gal–treated cardiomyocytes. Activation of ERK5 by LIF was augmented in AdGab1^WT-treated cardiomyocytes compared with Adβ-gal–treated cardiomyocytes. On the other hand, activation of ERK5 was reduced in AdGab1^F627/659- treated cardiomyocytes (Figure 6B). ERK1/2 was activated to the same extent in AdGab1^WT-treated cardiomyocytes as in Adβ-gal–treated cardiomyocytes. On the contrary, activation of ERK1/2 was reduced in AdGab1^F627/659-treated cardiomyocytes compared with Adβ-gal–treated cardiomyocytes (Figure 6C). Activation of AKT was enhanced in AdGab1^WT-treated or AdGab1^F627/659-treated cardiomyocytes compared with Adβ-gal–treated cardiomyocytes (Figure 6D). Activation of STAT3 was not altered in cardiomyocytes infected with Adβ-gal, AdGab1^WT, or AdGab1^F627/659. These results indicate that Gab1 plays a critical role in activation of ERK5 and ERK1/2 by LIF through interaction with SHP2 in cardiomyocytes. Based on a previous report,²⁸ the present finding suggests that the interaction of Gab1 with SHP2 might be involved in LIF-induced elongation of cardiomyocytes through activation of ERK5.

• Download figure
• Open in new tab
• Download powerpoint

Figure 6. Gab1 is involved in activation of ERK5 by LIF through interaction with SHP2 in cardiomyocytes. A, Cardiomyocytes were infected at an moi of 20 with Adβ-gal, AdGab1^WT, or AdGab1^F627/659 and serum deprived. At 24 hours after infection, cells were stimulated with 3×10² U/mL LIF for indicated periods of time. Total cell extracts were prepared and blotted with anti-phospho ERK5 (p-ERK5), anti-ERK5, anti-phospho ERK1/2 (p-ERK1/2), anti-ERK1/2, anti-phospho AKT (pAKT), anti-AKT, anti-phospho STAT3 (p-STAT3), and anti-STAT3 antibodies. Gab1 expression was confirmed with anti-Gab1 antibody. All data shown are one representative result from 3 independent experiments that had a similar result. B through D, Phosphorylation of ERK5, ERK1/2, and AKT was normalized to total ERK5, ERK1/2, and AKT using FluorChem-8000 (Alpha-Innotech).

Overexpression of the Dominant-Negative Form of ERK5 Abrogates the Effect of Gab1^WT on Longitudinal Elongation of Cardiomyocytes Induced by LIF

To determine whether ERK5 might participate in the LIF-activated signaling pathway that mediates longitudinal elongation of cardiomyocytes induced by LIF, we constructed recombinant adenovirus vector expressing dominant-negative form of ERK5 (AdERK5^AEF). As shown in Figure 7A, overexpression of ERK5^AEF almost abrogated LIF-induced longitudinal elongation of cardiomyocytes.

• Download figure
• Open in new tab
• Download powerpoint

Figure 7. Overexpression of ERK5^AEF abrogates the effect of Gab1^WT on longitudinal elongation of cardiomyocytes induced by LIF. A, Cardiomyocytes were infected with AdERK5^AEF. Cells were treated as described in Figure 4. Representative data are shown. B, Cardiomyocytes were dual infected with AdGab1^WT and Adβ-gal or with AdGab1^WT and AdERK5^AEF. Cells were treated as described in Figure 4. Representative data are shown. C through F, A total of 150 sarcomeric α-actinin–positive cells were examined for each measurement. Cell surface area, cell length, and cell width were examined as described in Figure 4. Values are mean±SD; *P<0.001 vs LIF(−); †P<0.001 vs Adβ-gal. Experiments were repeated 3 times with similar results.

To test whether overexpression of ERK5^AEF abrogates LIF-induced longitudinal elongation of cardiomyocytes overexpressing Gab1^WT, cardiomyocytes were dual infected with AdGab1^WT and Adβ-gal or with AdGab1^WT and AdERK5^AEF. LIF induced elongation of cardiomyocytes infected with AdGab1^WT and Adβ-gal. On the contrary, this morphological change was significantly inhibited in cardiomyocytes infected with AdGab1^WT and AdERK5^AEF (Figure 7B). We quantified the cell surface area, cell length, and cell width of these cardiomyocytes. As shown in Figures 7C and 7D, LIF increased cell surface area by 60% and cell length by 84% in cardiomyocytes infected with AdGab1^WT and Adβ-gal. However, in cardiomyocytes infected with AdGab1^WT and AdERK5^AEF, the increases in cell surface area and cell length induced by LIF were almost abrogated. Cell width was significantly decreased in response to LIF in cardiomyocytes infected with AdGab1^WT and Adβ-gal but not in cardiomyocytes infected with AdGab1^WT and AdERK5^AEF (Figure 7E). The cell length to width ratio was significantly increased by LIF in cardiomyocytes infected with AdGab1^WT and Adβ-gal but not in cardiomyocytes infected with AdGab1^WT and AdERK5^AEF (Figure 7F). Therefore, it seems that ERK5 might be an essential component of LIF-activated signaling pathway, leading to elongated morphology of cardiomyocytes through Gab1-SHP2 interaction.