The ubiquitin ligase MuRF1 protects against cardiac ischemia/reperfusion injury by its proteasome-dependent degradation of phospho-c-Jun

Abstract

Despite improvements in interventions of acute coronary syndromes, primary reperfusion therapies restoring blood flow to ischemic myocardium leads to the activation of signaling cascades that induce cardiomyocyte cell death. These signaling cascades, including the mitogen-activated protein kinase signaling pathways, activate cardiomyocyte death in response to both ischemia and reperfusion. We have previously identified muscle ring finger-1 (MuRF1) as a cardiac-specific protein that regulates cardiomyocyte mass through its ubiquitin ligase activity, acting to degrade sarcomeric proteins and inhibit transcription factors involved in cardiac hypertrophy signaling. To determine MuRF1's role in cardiac ischemia/reperfusion (I/R) injury, cardiomyocytes in culture and intact hearts were challenged with I/R injury in the presence and absence of MuRF1. We found that MuRF1 is cardioprotective, in part, by its ability to prevent cell death by inhibiting Jun N-terminal kinase (JNK) signaling. MuRF1 specifically targets JNK's proximal downstream target, activated phospho-c-Jun, for degradation by the proteasome, effectively inhibiting downstream signaling and the induction of cell death. MuRF1's inhibitory affects on JNK signaling through its ubiquitin proteasome-dependent degradation of activated c-Jun is the first description of a cardiac ubiquitin ligase inhibiting mitogen-activated protein kinase signaling. MuRF1's cardioprotection in I/R injury is attenuated in the presence of pharmacologic JNK inhibition in vivo, suggesting a prominent role of MuRF1's regulation of c-Jun in the intact heart.

Supplemental Figure 1

MuRF1−/− mice were challenged to a 30-minute left anterior descending (LAD) ischemia injury, followed by a 24-hour recovery in vivo. A: Determination of the areas of infarct as a fraction of the area at risk demonstrated that MuRF1−/− mice were not significantly protected against I/R injury in vivo. Conscious echocardiography was performed to determine the functional protection of cardiac MuRF1 to I/R injury in vivo. MuRF1−/− mice were not significantly protected functionally compared with wild-type mice determined by fractional shortening (bottom) and EF% (see Supplemental Table S3 at http://ajp.amjpathol.org; N = 5-7 per group as outlined in Supplemental Table S3 at http://ajp.amjpathol.org).

Figure 1

MuRF1 protects against I/R injury–induced cell death. A: Increasing MuRF1 in H9C2 cardiomyocytes using adenovirus protects against cell death after challenge with 1-hour simulated ischemia (SI) followed by a 24-hour reperfusion (R), determined by trypan blue exclusion. B: The number of apoptotic H9C2 cells observed by TUNEL assay after SI/R is significantly less with increased MuRF1 expression as shown in these representative fields. C: Quantitative analysis of TUNEL-positive cells was from three independent experiments (120 cells counted per experiment). D: Activated (cleaved) caspase-3 was increased in response to SI/R. Increasing MuRF1 blunted the amount of cleaved caspase-3 determined by Western immunoblot. E: Cardiomyocytes with decreased MuRF1 after 36 hours of siRNA transfection determined by immunoblot (below) have an enhanced cell death after SI/R (top) determined by DNA fragmentation analysis by flow cytometry. F: An increase in TUNEL-positive cells occurs after cardiomyocytes were infected with adenovirus siRNA-control or siRNA-MuRF1 at 50 multiplicity of infection for 24 hours, and then treated with normoxia or SI/R for 24 hours. Quantitative analysis was performed as described in C. G: Quantitative analysis of MuRF1, MuRF2, and MuRF3mRNA expression by RT-PCR in H9C2 cells after SI/R. Results are expressed as means ± SEM, *P < 0.01 versus Ad-GFP or siRNA-control.

Figure 2

MuRF1 protects against ischemia reperfusion injury by inhibiting c-Jun but not the ERK or p38 signaling pathways in a proteasome-dependent manner. A and B: MuRF1 protects against cell death by inhibiting JNK signaling. H9C2 cardiomyocytes were infected with 50 multiplicity of infection Ad-GFP or Ad-MuRF1 for 24 hours pretreated with U0126 (MAP kinase/Erk kinase inhibitor, inhibiting ERK1/2, 10 μmol/L), SB203580 (p38 inhibitor, 10 μmol/L), or SP600125 (JNK inhibitor, 10 μmol/L) before 30 minutes of simulated hypoxia and 24 hours reperfusion or normoxia. Cell death was determined by trypan blue exclusion (A) and by TUNEL staining (B) and demonstrated that SI/R cell death was mediated primarily by the JNK signaling pathway. C: H9C2 cardiomyocytes were transduced with 50 multiplicity of infection Ad-GFP or Ad-MuRF1, and the effects on MAPK signaling pathways were determined. By Western immunoblot, total and phospho-ERK1/2, total and phospho-p38, total and phospho-JNK1/2, total and phospho-c-Jun (Ser63 or Thr91), and Myc-MuRF1 were determined by Western blot with the indicated antibodies. Only phospho-c-Jun (Ser63 and Thr91) levels were significantly decreased when MuRF1 expression is increased, indicating that MuRF1 exerts its regulation of cell death by regulating JNK signaling. D: Cardiomyocytes with increasing MuRF1 challenged with SI/R have a blunted or decreased phospho-c-Jun response. E: Parallel studies using the proteasome inhibitor MG132 6 hours before harvest demonstrated that the blunted phospho-c-Jun response is dependent on proteasome degradation. F: Decreasing MuRF1 using siRNA conversely enhances phospho-c-Jun after SI/R challenge in cardiomyocytes. Results are expressed as means ± SEM, *P < 0.01 versus Ad-GFP. IP, immunoprecipitation; IB, immunoblot.

Figure 3

MuRF1 directly interacts with c-Jun to inhibit transcriptional activity. A: To assess the role of MuRF1 on phospho-c-Jun transcriptional activity, HEK293T cells were transfected with luciferase plasmids driven by the AP-1 promoter as indicated. The ubiquitin ligase activity of MuRF1, found in the RING motif, was necessary for the inhibition of AP-1 activity. B: To assess which c-Jun phosphorylation sites were necessary for MuRF1's inhibition of AP-1 activity, wild-type c-Jun and c-Jun mutants were co-transfected in HEK293T cells with the AP-1 luciferase reporter with and without MuRF1. Luciferase activity was determined after 24 hours. C: To assess how decreasing endogenous MuRF1 affects AP-1 activity, H9C2 cardiomyocytes were co-transfected with AP-1 luciferase, siRNA oligo controls, siRNA MuRF1 oligos, and luciferase activity was determined after 36 hours. D: To determine whether MuRF1 interacts directly with c-Jun to inhibit its activity, HEK293T cells were transfected with FLAG-cJun and pulled down by GST-MuRF1 or GST followed by immunoblot for c-Jun. E: This direct interaction was confirmed by specific immunoprecipitation of myc-MuRF1 after co-transfection of myc-MuRF1 and FLAG-c-Jun plasmids in HEK293T cells by immunoblot. F: The co-localization of MuRF1 and c-Jun in cardiomyocytes was identified with use of confocal microscopy and immunofluorescent staining of MuRF1 (anti-MuRF1/red), c-Jun (anti-c-Jun/green), and nuclei (DAPI/blue). *P < 0.01 versus wild-type c-Jun alone. **P < 0.01 versus siRNA control. n.s. indicates not significant.

Figure 4

MuRF1 binds c-Jun through its MFC domain found in all MuRF family proteins. A: The determination of the MuRF1 region that interacted with c-Jun was made by co-transfecting FLAG-c-Jun and myc-tagged MuRF1 deletion mutants and following interactions by immunoprecipitation of c-Jun. B: The region of c-Jun that is necessary to bind MuRF1 was determined using GST-MuRF1 fusion protein to pull down c-Jun deletion constructs. C: The schematic representation of how MuRF1 deletion constructs interacted with c-Jun shows the essential nature of the MFC region for MurF1 to bind c-Jun. D: The schematic representation of pull-down assays demonstrate that the 125-220–amino acid region of c-Jun is necessary for MuRF1 to interact with c-Jun.

Figure 5

The ubiquitin ligase activity of MuRF1 promotes mixed chain (K48 and K63) c-Jun ubiquitination in cells and in vitro. A: The RING finger domain has previously been shown to have ubiquitin ligase activity, which adds ubiquitin chains to substrates. To determine whether MuRF1's RING finger domain was necessary to ubiquitinate c-Jun, HEK293T cells were co-transfected with c-Jun, MuRF1 (or MuRF1 missing the RING finger domain), and His-tagged ubiquitin as indicated. Cell lysates were affinity purified for ubiquitin using nickel chromatography (which binds His). MuRF1, but not MuRF1 lacking the RING finger domain, ubiquitinates c-Jun in cells (top). B: In vitro studies were performed to determine the types of ubiquitin chains being added to c-Jun by MuRF1. In vitro ubiquitination reactions were performed with purified ubiquitin, E1, the E2-Ubc5c, purified c-Jun, GST-MuRF-1 wt or GST-MuRF1ΔRING or GST alone. Immunoblot analysis of ubiquitin demonstrated that MuRF1 enhanced the ubiquitination of c-Jun (left). The arrow indicates minimum ubiquitination by the MuRF1ΔRING construct.

Figure 6

MuRF1 preferentially binds, ubiquitinates, and degrades phosphorylated c-Jun. A: HEK293T cells were co-transfected with plasmids expressing MuRF1, c-Jun, and phosphorylation site c-Jun mutants. Interactions between MuRF1 and the different phosphorylation mutants were determined by immunoprecipitation. MuRF1 (anti-myc) immunoprecipitation bound wild-type c-Jun preferentially to c-Jun that was unable to be phosphorylated at S63 and 73A determined by immunoblot for c-Jun constructs (anti-FLAG). B: MuRF1's preferential binding of wild type c-Jun compared with S63, 73A mutants parallels MuRF1's ubiquitination of c-Jun. HEK293T cells co-transfected with plasmids expressing His-Ub, Myc-MuRF1, and Flag-c-Jun wild-type or c-Jun phosphorylation mutants as indicated. c-Jun was immunoprecipitated and ubiquitination was determined by immunoblot (top), despite equal expression of c-Jun and c-Jun mutants (bottom). c-Jun mutants lacking their S63, S73 phosphorylation sites are ubiquitinated less by MuRF1 in response to JNK stimulation. C: MuRF1 preferentially degrades c-Jun after its phosphorylation in response to cycloheximide. HEK293T cells were transfected with MuRF1, wild-type c-Jun, and a mutant c-Jun that was unable to be phosphorylated (c-Jun S63, 73A, T91, 93A). By immunoblot, only the wild-type c-Jun that can be phosphorylated is degraded by MuRF1.

Figure 7

Mice with increased cardiac MuRF1 exhibit attenuated I/R injury and have an enhanced degradation of cardiac c-Jun in response to I/R in vivo. A: Isolated working hearts from MuRF1 Tg+ and strain-matched wild-type mice were assayed at baseline and after global I/R injury on a Langendorff system. Hearts were allowed to equilibrate for at least 30 minutes, followed by 15 minutes of no-flow ischemia and 20 minutes reperfusion. The recovery of cardiac function (expressed as a % of baseline function) is the fraction of mean left ventricular developed pressure. MuRF1 Tg+ heart recovery trended to improve 20 minutes after perfusion was restarted. (Additional functional parameters are shown in Supplemental Table S1 at http://ajp.amjpathol.org. N = 5 to 6 per group outlined in Supplemental Table 1 at http://ajp.amjpathol.org). *P < 0.05. B: To determine the durability of this recovery, we challenged MuRF1 Tg+ mice to a 30-minute LAD ischemia injury, followed by a 24-hour recovery in vivo. Determination of the areas of infarct as a fraction of the area at risk demonstrated that MuRF1 Tg+ mice had significant protection against I/R injury in vivo (top). Conscious echocardiography was performed to determine the functional protection of cardiac MuRF1 to I/R injury in vivo. MuRF1 Tg+ mice had significantly less functional deficit compared with wild-type mice determined by fractional shortening (bottom) and EF% (see Supplemental Table S2 at http://ajp.amjpathol.org. N = 6 to 12 mice per group outlined in Supplemental Table S2 at http://ajp.amjpathol.org). Arrows indicate infarcted area in white. C: Cardiac c-Jun and phospho-cJun in MuRF1 Tg+ and wild-type hearts 24 hours after LAD I/R injury were determined by immunoblot. Significantly less c-Jun and phospho-c-Jun were found in four representative hearts consistent, with its role as a MuRF1 substrate in vivo. N = 3 to 4 hearts per group as indicated in blot. PhosphoSer63-c-jun/total c-jun ratio is the density of the phosphoSer63 c-jun (A.U.) divided by the density of total c-jun (A.U.) to determine the relative phosphorylation of c-jun site associated with activation. D: Cardiac cleaved caspase-3, phospho-ERK1/2/total ERK1/2, phospho-p38/total p38 at baseline and after 30 minutes ischemia and 24 hours reperfusion. E: Real-time PCR analysis of cardiac MuRF1, MuRF2, and MuRF3 mRNA in sham or I/R-challenged wild-type mice 24 hours after surgery. F: Histologic evaluation of cardiac infarction and cardiac function in MuRF1 Tg+ hearts pretreated with the JNK inhibitor SP600125 and challenged with I/R injury (further echocardiographic detail can be found in Supplemental Table S3 at http://ajp.amjpathol.org). *P < 0.05, ***P < 0.001.