Continuous Glycoprotein-130–Mediated Signal Transducer and Activator of Transcription-3 Activation Promotes Inflammation, Left Ventricular Rupture, and Adverse Outcome in Subacute Myocardial Infarction
Background— In patients with myocardial infarction, high serum levels of interleukin-6 cytokines predict a poor outcome. The common receptor of interleukin-6 cytokines, glycoprotein-130 (gp130), signals via janus kinase/signal transducer and activator of transcription (STAT), cytoplasmic protein tyrosine phosphatase/extracellular signal-regulated kinase, and phosphoinositide-3-kinase/Akt pathways, and the regulation of these pathways depends at least in part on the gp130 tyrosine-757 residue. By analyzing cardiomyocyte-specific gp130Y757F mutant mice, we investigated the effect of disturbed gp130 signaling after myocardial infarction.
Methods and Results— The cardiomyocyte-restricted α-myosin heavy chain-Cre-recombinase-loxP system was used to generate mice with gp130Y757F mutant cardiomyocytes (αMHC-Cretg/−;gp130fl/Y757F [Y757F]); all other cells carried at least 1 functional gp130 gene, ensuring normal gp130 signaling. Y757F mice displayed normal cardiac function and morphology at 3 months of age comparable to their nonmutant littermates. In response to myocardial infarction, Y757F mice displayed higher mortality associated with increased left ventricular rupture rate, sustained cardiac inflammation, and heart failure. These adverse effects were associated with prolonged and enhanced STAT3 activation and increased expression of interleukin-6 and of the complement-activating mannose-binding lectin C. Pharmacological inhibition of the complement system by cobra venom factor attenuated inflammation, prevented left ventricular rupture, and improved cardiac function in Y757F mice. Stronger effects were observed with a genetic reduction of STAT3 (STAT3flox/+) restricted to cardiomyocytes in Y757F mice, which prevented extensive upregulation of interleukin-6, complement activation, and sustained inflammation and lowered left ventricular rupture rate, heart failure, and mortality in subacute myocardial infarction.
Conclusion— Impaired downregulation of gp130-mediated STAT3 activation in subacute infarction promotes cardiac inflammation, adverse remodeling, and heart failure, suggesting a potential causative role of high interleukin-6 serum levels after myocardial infarction.
Received December 21, 2009; accepted May 3, 2010.
Distinct changes in the pattern of circulating interleukin-6 (IL-6) family cytokines and alteration in the myocardial expression and activation of their common receptor glycoprotein-130 (gp130) have been reported in patients with myocardial infarction (MI), cardiac hypertrophy, and chronic heart failure.1–4 High serum IL-6 levels in patients with MI or heart failure are consistently associated with reduced cardiac function and a poor outcome (ie, increased morbidity and mortality), an observation that contrasts with experimental data suggesting cardioprotective effects of gp130 receptor activation.5,6
Editorial see p 103
Clinical Perspective on p 155
The cardiac stress response via IL-6 cytokines involves a complex network of the gp130 homodimeric or heterodimeric receptor complex controlling 3 major downstream signaling pathways: cytoplasmic protein tyrosine phosphatase (SHP2)/extracellular signal-regulated kinase (ERK), phosphoinositide-3-kinase (PI3)/protein-kinase B (PKB, Akt), and janus kinase (JAK)/signal transducer and activator of transcription (STAT).1,7,8
On ligand binding, gp130 receptor activation initiates 2 events: the phosphorylation of a single membrane proximal tyrosine (Y) residue (Y757 in mice, Y759 in humans) necessary and sufficient for recruitment of SHP2 to the gp130 receptor, which subsequently activates ERK and PI3/Akt signaling cascades,1,9 and phosphorylation of several tyrosine residue at the membrane proximal C terminal end of the gp130 receptor, inducing docking of JAK and STAT proteins at the gp130 receptor, which initiates activation of JAK/STAT signaling.1,9
After activation, negative feedback mechanisms are important to control gp130 signaling. To achieve this control, gp130-STAT activation induces the expression of the suppressor of cytokine signaling-3 (SOCS3) protein, which binds to the Y757 and terminates activation of JAK/STAT signaling. By binding to the same Y757 site as SHP2, SOCS3 also reduces the formation of Y757/SHP2 complexes at the gp130 receptor and thereby limits ERK and Akt activation.1,9,10 Furthermore, SHP2, which initially activates ERK and Akt, later develops a tyrosine phosphatase activity that constrains activation of all gp130 downstream signaling pathways.1,9,10 Thus, the Y757 site appears to be specifically important to ensure a “physiological” type of gp130 signaling with timed phases of signaling activation and termination. Indeed, mutations at Y757 (ie, substitution of Y757 by a phenylalanine [F], gp130Y757F) prevent binding of SHP2 and SOCS3 and thereby disturb the gp130 downstream signaling network, resulting in prolonged activation of JAK/STAT signaling in the absence of gp130-mediated ERK and Akt activation.9 In the intestine, the gp130Y757F mutation is associated with chronic inflammation.9 We hypothesized that physiological activation of gp130 signaling might be protective after MI and that a disturbed balance of downstream signaling pathways resulting from a gp130Y757F point mutation may be detrimental.
Cell culture media were from Biochrome (Berlin, Germany). All other laboratory chemicals were from Sigma-Alderich (St Louis, Mo).
αMHC-Cretg/−; gp130fl/fl mice1 were crossed with mice harboring a systemic “knock-in” mutation in the cytoplasmic domain of the gp130 receptor replacing tyrosine-757 with a phenylalanine (gp130Y757F)11 and with STAT3flox mice.12 Sibling male mice (age, 12 to 16 weeks) were used in all studies. Sham and MI operations (permanent occlusion of the left anterior descending artery), echocardiography, and hemodynamic measurements (baseline and after 1 day, 3 days, 7 days, or 2 weeks as indicated) were performed as described.12,13 Mice suffering intraoperative death were excluded from the study. Cobra venom factor (CVF) was injected intraperitoneally (Aczon; 11 U · kg−1 · /d−1, 3 days to 2 weeks after MI). All animal studies were in compliance with the Guide for the Care and Use of Laboratory Animals as published by the US National Institutes of Health and approved by our local Institutional Review boards.
Left ventricular (LV) tissue was taken from patients undergoing heart transplantation as a result of end-stage heart failure caused by coronary artery disease (ischemic cardiomyopathy [ICM]; n=5) and from organ donors not suited for transplantation (n=3) (see the online-only Data Supplement).
Histology and Immunostaining
Infarct size measurements were performed on LV sections stained with hematoxylin and eosin as the ratio of scar length to LV circumference (see the online-only Data Supplement).12 In mice subjected to molecular analysis, MI size was determined by echocardiography and histology of a midventricular section (see the online-only Data Supplement). Histological, immunohistochemical, and morphological analyses were performed 2 weeks after sham operation or MI as described (see the online-only Data Supplement).12
Quantitative Reverse-Transcription Polymerase Chain Reaction, Western Blot, and Immunoprecipitation
RNA isolation, generation of complementary DNA (cDNA), quantitative reverse-transcription polymerase chain reaction (qRT-PCR), Western blot, and immunoprecipitation were performed as described (see the online-only Data Supplement).12 Specific antibodies (Cell Signaling Technology, Danvers, Mass) were used to detect the activation state of STAT3: P-STAT3 (phosphorylation [P] of tyrosine-705), ERK1/2; P-ERK1/2 (phosphorylation of tyrosine-202/204) and Akt, P-Akt (phosphorylation of serine-473). Information on all other antibodies and primer sequences are shown in the online-only Data Supplement.
Whole genome mouse microarray (Miltenyi Biotec) was performed on messenger RNA (mRNA) isolated from nonischemic LV tissue 2 weeks after MI on pools of n=5 mice (wild-type [WT] versus Y757F).
Cardiomyocytes from 3- to 4-month-old mice were isolated with 0.04% collagenase (Worthington) as described previously.11
Data are presented as mean±SD. Survival analyses were done by Kaplan-Meier plot followed by a log-rank test (Mantel- Cox). Differences between groups were analyzed by the Student t test or ANOVA followed by Bonferroni posthoc analyses as appropriate. A 2-tailed value of P<0.05 was considered statistically significant.
Baseline Characterization of Mice With Cardiomyocyte-Restricted gp130 Receptor Mutations
The genotypes of WT and mutant mice are illustrated in Figure 1A; the genetic crosses and genotyping are shown in Figure I of the online-only Data Supplement. Mice with a knockout for gp130 receptor signaling restricted to cardiomyocytes (αMHC-Cretg/−;gp130fl/fl; KO) have been described previously.1 Briefly, these mice carry 2 floxed gp130 alleles (gp130fl/f) and a transgene expressing the Cre-recombinase under the cardiomyocyte-specific α-myosin heavy chain (αMHC) promoter (αMHC-Cretg/−). With the induction of the MHC promoter around term, the floxed gp130 alleles become deleted, leading to a gp130 knockout only in postnatal cardiomyocytes. Non-Cre transgenic littermates carrying 2 fully functional gp130 receptor alleles in all cell types throughout life are called WT (gp130flox/flox; Figure 1A). Mice carrying the gp130 point mutation Y757F and a WT gp130 allele (gp130Y757F/fl: Y757F/+; Figure 1A) appear normal, indicating that a single WT gp130 allele is sufficient to ensure adequate gp130 signaling in all cells.11 In αMHC-Cretg/−;gp130Y757F/fl mice (Y757F, Figure 1A), the floxed gp130 allele is deleted in postnatal cardiomyocytes that now express only the mutant gp130Y757F protein, whereas all noncardiomyocyte cells express, in addition to the mutant protein, a WT gp130 protein. WT, KO, Y757F/+, and Y757F mice were born according to mendelian inheritance ratios, survived into adulthood, and displayed normal cardiac morphology and function at 3 months of age (Table I of the online-only Data Supplement).1
Post-MI Mortality Is Markedly Increased in Y757F Compared With KO and WT Mice
WT, KO, and Y757F sibling mice were exposed to MI. Post-MI mortality in a 2-week follow-up is markedly higher in Y757F mice (Figure 1B; n=44, 16 died; P<0.05 versus WT) compared with WT (n=44, 5 died). Surprisingly, KO mice with a complete deletion of gp130 signaling in cardiomyocytes survived better than Y757F mice (KO: n=39, 8 died; P=NS versus WT). The majority of Y757F mice (n=12) died of LV rupture 5 to 10 days after MI (Figure 1C), whereas only 1 WT and 1 KO mouse died of LV rupture. Post-MI survival rates of gp130Y757F/+ mice were comparable to those of WT mice (gp130fl/Y757F; 3 of 20 died; P=NS versus WT, no LV rupture), indicating that 1 functional gp130 allele in cardiomyocytes and in noncardiomyocytes is sufficient to ensure normal gp130 signaling after MI.
Y757F Mice Fail to Downregulate STAT3 Activation in Subacute MI
No significant differences were observed in the expression and activation of ERK1/2 or Akt between WT, KO, and Y757F mice at baseline (Figure I of the online-only Data Supplement). Baseline STAT3 protein levels were moderately lower in KO (−30±15%; P<0.05 versus WT) but not in Y757F mice (Figure I of the online-only Data Supplement) compared with WT mice, whereas STAT3 was similarly activated in all 3 genotypes (Figure I of the online-only Data Supplement).
At all post-MI time points, expression and activation of ERK1/2, Akt, and STAT1 were comparable in WT and Y757F mice (Figure 2A and B and data not shown). Both WT and Y757F mice displayed marked induction of STAT3 activation 1 day after MI, whereas it was moderately lower in KO mice (Figure 2A and 2B). In subacute infarction (3 days, 7 days, and 2 weeks after MI), activation of STAT3 decayed in WT mice and persisted at a significantly higher level in Y757F mice (Figure 2A and 2B). Furthermore, total STAT3 protein levels were higher in Y757F compared with WT mice 2 weeks after MI (Figure 2A and 2B).
Immunoprecipitation showed increased SOCS3 recruitment to the gp130 receptor in nonischemic LVs of WT mice 2 weeks after MI, whereas substantially less (−56%; P<0.05) SOCS3 protein was associated with the gp130 receptor complex in Y757F nonischemic LVs (Figure 2C). Because it is assumed that SOCS3 can bind to heterodimeric gp130/LIF receptor complexes as well, the remaining immunoprecipitated SOCS3 in Y757F may partly represent gp130 heterodimers in cardiomyocytes but may derive mainly from cardiac nonmyocyte cells expressing WT gp130.
Monoallelic STAT3 Ablation in Cardiomyocytes Improves Survival Rate in Y757F;STAT3low Compared With Y757F Mice After MI
To test the hypothesis that enhanced and prolonged STAT3 activation is responsible for adverse effects in infarcted Y757F mice, monoallelic STAT3 deletion in cardiomyocytes, but not in noncardiomyocytes, was obtained by replacing 1 WT STAT3 by a floxed STAT3 allele (STAT3fl; Figure I of the online-only Data Supplement). These αMHC-Cretg/−; gp130fl/Y757F;STAT3fl/+ mice (Y757F;STAT3low; Figure 1A) express reduced levels of functional STAT3 protein in cardiomyocytes, whereas all nonmyocytes express normal STAT3 levels, which in aggregate reduces cardiac STAT3 protein levels at baseline by ≈33% compared with WT and Y757F mice (Figure 3A and 3B). STAT3 protein levels in other organs (liver, skeletal muscle) are unchanged (Figure 3A and 3B). Two weeks after MI, Y757F;STAT3low mice displayed markedly improved survival rates compared with Y757F mice (mortality rate, 11% [n=18] versus 50% [n=12] in sibling Y757F mice; P<0.03). No LV ruptures were recorded in Y757F;STAT3low mice. The increase in total and phosphorylated (activated) cardiac STAT3 proteins was reduced compared with Y757F (Figure 3A), a feature that was also confirmed by immunohistochemistry. The prominent staining of activated P-STAT3 in cardiomyocyte nuclei of Y757F mice was absent in Y757F;STAT3low mice (Figure 3C). In turn, post-MI cardiac STAT3 and P-STAT3 levels in Y757F;STAT3low mice were comparable to WT mice (Figure 3A and 3C).
Y757F;STAT3low Mice Display Smaller Infarcts and Less LV Dilatation Compared With Y757F Mice
Two weeks after MI, surviving Y757F mice displayed marked LV dilatation and infarct scar elongation with larger infarct sizes (40±17%; n=17; P<0.05 versus WT) compared with WT (27±17%; n=20) or Y757F;STAT3low (31±9%; n=9; P=NS versus WT; Figure 4A and 4B and the Table) mice. Despite larger infarcts, no significant differences in LV apoptosis were observed between Y757F and WT mice (Figure II of the online-only Data Supplement).
Post-MI Cardiac Dysfunction Is Attenuated in Y757F;STAT3low Compared With Y757F Mice
Three days after MI, echocardiograms showed similar LV dimensions and function for WT, Y757F, and Y757F;STAT3low mice (Figure 4B and Table II of the online-only Data Supplement). Two weeks after MI, echocardiograms of mice with matched infarct sizes (MI size: WT, 35±8%; Y757F, 33±7%; Y757F;STAT3low, 33±8% as determined retrospectively by histological analysis in hematoxylin and eosin–stained LV sections) showed that Y757F mice displayed significantly depressed cardiac function compared with WT mice, whereas cardiac function in Y757F;STAT3low mice was comparable to that in WT mice (the Table). Sham-operated mice of all genotypes were similar with respect to cardiac function 2 weeks after MI (Table III of the online-only Data Supplement).
Post-MI Cardiomyocyte Hypertrophy and Capillary Density Are Altered in Y757F but Not Y757F;STAT3low Mice
Morphometric analysis in sections from in situ fixed LVs showed decreased septum thickness, cardiomyocyte cross-sectional area and cardiomyocyte length in nonischemic LVs of Y757F compared with WT mice, whereas in Y757F;STAT3low mice, these parameters were comparable to WT mice (the Table and Figure 4C). Despite these differences, whole genome microarrays of nonischemic LVs 2 weeks after MI showed no alterations in the classic marker profile associated with pathophysiological hypertrophy, ie, atrial natriuretic peptide, brain-type natriuretic peptide, skeletal muscle α-actin, SERCA2a, and α- and β-MHC between WT and Y757F mice (data not shown). However, nonischemic LVs of Y757F mice displayed a higher ratio between capillaries and cardiomyocytes compared with WT mice, whereas it was similar in Y757F;STAT3low mice (the Table and Figure 4C).
Enhanced Collagen Fragmentation and Degradation and Increased Expression of Matrix Metalloproteinase-1 and -13 in the Infarct Border Zone of Y757F Mice Is Corrected in Y757F;STAT3low Mice
Moderate interstitial fibrosis and similar expression of matrix metalloproteinase-1 (MMP1) and MMP13 (data not shown) were present in nonischemic LVs of WT, Y757F, and Y757F;STAT3low mice, although collagen fibers in Y757F mice appeared to be more fragmented (Figure 4D). In contrast, in the infarct border zone, MMP1 and MMP13 expression, collagen fragmentation and degradation, and loose tissue organization were increased in Y757F compared with WT mice, whereas it appeared dense and intact in Y757F;STAT3low mice (Figure 4D through 4F). Tissue inhibitor of metalloproteinase-1 (TIMP1) protein levels in border zones were comparable between all genotypes (Figure 4F).
Enhanced Cardiac Inflammation in Y757F Compared With Y757F;STAT3low Mice in the Subacute Phase of MI
Two weeks after MI, the degree of inflammation indicated by increased numbers of CD45-positive cells, mainly leukocytes and macrophages, was markedly higher in the nonischemic LV and border zone of Y757F compared with WT or Y757F;STAT3low mice (Figure 5A through 5C). Compared with WT mice, the degree of post-MI inflammation tended to be even lower in KO mice (Figure 5A through 5C).
Whole Genome Microarray Analysis Confirms a Proinflammatory Gene Expression in Infarcted Y757F Hearts
Whole genome microarray analysis comparing the mRNA pattern in nonischemic LVs 2 weeks after MI showed that 555 genes were upregulated and 397 were downregulated in Y757F compared with WT. Many of the upregulated genes in Y757F were associated with an enhanced inflammatory status, including cytokines, IL-6, chemokines, chemokine ligand 5, CXCL13, chemokine receptor 5, factors associated with the complement system, mannose-binding lectin A and C (MBL-C), mannan-binding lectin serine peptidase 2, complement component 3 (C3), C3ar1, C4bp, C6, C8b, C8g, and serum amyloid A3 and 4 (Table IV of the online-only Data Supplement). Expression of tumor necrosis factor-α or IL-1β did not differ between WT and Y757F mice (confirmed by qRT-PCR; data not shown).
The Post-MI Kinetics of IL-6 and MBL-C Expression Differs Among Y757F, WT, and Y757F;STAT3low Mice
IL-6 mRNA expression (qRT-PCR) in nonischemic LV was similarly upregulated in WT (16-fold; P<0.01 versus sham), Y757F (18-fold; P<0.01 versus sham), and Y757F;STAT3low (16-fold; P<0.01 versus sham) mice 1 day after MI. Two weeks later, IL-6 mRNA levels in WT and Y757F;STAT3low mice had decreased (WT: 1.2-fold, P=NS versus sham; Y757F;STAT3low: 4-fold, P<0.01 versus WT sham), whereas it had increased further in Y757F mice (56-fold; P<0.01 versus sham).
MBL-C mRNA levels in infarcted WT and Y757F;STAT3low mice did not differ between sham and nonischemic LVs during the 2-week follow-up. In Y757F mice, however, MBL-C mRNA levels increased steadily (Figure 6A).
Post-MI Accumulation of MBL-C and Complement C3 Proteins in LVs of Y757F but Not in WT or Y757F;STAT3low Mice
MBL-A/MBL-C double-knockout mice served as controls for MBL-C antibody specificity in Western blots and immunohistochemistry. Twenty-four hours after MI, MBL-C protein content decreased in nonischemic LVs of WT and Y757F mice compared with shams (Figure 6C). MBL-C protein levels remained low in WT mice over the 2-week follow-up (Figure 6D through 6F). In contrast, MBL-C protein levels were restored to sham levels in Y757F mice within 7 days after MI (Figure 6D through 6F). This restoration of MBL-C protein was not observed in Y757F;STAT3low mice (Figure 6F).
Immunohistochemistry confirmed the presence of MBL-C protein in LVs of control and sham mice (WT, Y757F, and Y757F;STAT3low mice; Figure 6B and Figure III of the online-only Data Supplement). Immunostaining was markedly reduced in nonischemic LVs of WT and Y757F;STAT3low mice 2 weeks after MI but remained abundant in Y757F mice with patchy staining in cardiomyocytes (Figure 6G). MBL-C is known to induce deposition of circulating C3 to tissues.14 Little C3 immunostaining was detected in sham LVs (shams of all genotypes were similar; Figure 6H). Two weeks after MI, cardiac C3 deposition was slightly increased in WT and Y757F;STAT3low mice but markedly increased in Y757F mice (Figure 6H).
Cardiac MBL-C Protein Levels in Patients With End-Stage Heart Failure Caused by Ischemic Heart Disease
Similar to infarcted Y757F mice, MBL-C protein levels in LV tissue of patients with end-stage heart-failure due to ICM were comparable to levels in nonfailing donor hearts (90±27% in ICM versus 100±18% in nonfailing; P=NS; Figure III of the online-only Data Supplement).
CVF Improves Postinfarct Outcome in Y757F Mice
CVF, derived from Naja melanoleuca cobra venom, inhibits complement activation.15 To test the effect of complement inhibition, CVF therapy was applied. Three days after MI, 22 Y757F mice with echocardiographically confirmed LV infarction were randomized into 2 groups (n=11 each; 1 IU · kg−1 · /d−1 CVF versus NaCl) for 11 days. After 2 weeks, the survival rate of the CVF group was significantly (P<0.01) improved with no fatalities (100% survived), whereas only 64% survived in the NaCl group with LV rupture in 50% of dead mice. Furthermore, LV function was improved and CD45+ inflammatory infiltrates were substantially reduced by CVF treatment (Figure 7A and 7B).
Important cardioprotective roles of STAT3 after MI with respect to angiogenesis, hypertrophy, fibrosis, and heart failure have been implicated previously in mice with cardiomyocyte-restricted deficiencies of STAT3.12 Yet, our present work indicates that not only STAT3 deficiency but also upregulated expression and activation of STAT3 mediated by the gp130 receptor system predominantly in subacute infarction can be detrimental for the heart. These detrimental effects of prolonged STAT3 activation after MI are caused in part by sustained inflammation leading to LV rupture in the earlier subacute phase and adverse remodeling and heart failure in the later phase. This scenario is mimicked in Y757F mice with impaired termination of JAK/STAT signaling in cardiomyocytes, resulting in a positive feedback circuit that increases both STAT3 expression and activation. Prolonged and detrimental inflammation by gp130-STAT3 activation after MI seems to depend at least in part on upregulation of MBL-C/lectin/complement activation. These conclusions are supported by the observations that a reduction of cardiomyocyte STAT3 activity improved post-MI survival because of the prevention of cardiac IL-6 and MBL-C expression, C3 deposition, cardiac inflammation, LV rupture, and heart failure and that targeting the complement system by CVF improved cardiac function and survival in infarcted Y757F mice by reducing cardiac inflammation and LV rupture. Because unleashing gp130-STAT3 signaling after MI promoted LV rupture and heart failure, our observations suggest the need for a precise and timely regulation of gp130-mediated STAT3 activation after MI to securely exploit its beneficial effects.
ERK and Akt activation and termination of JAK/STAT signaling depend critically on the Y757 of gp130. Binding of SHP2 and SOCS3 to the gp130 receptor is simultaneously eliminated by the point mutation gp130Y757F. This shifts the balance of gp130-mediated signaling toward an enhanced and prolonged JAK/STAT activation,9 illustrated by continuously increased STAT3 activation after infarction that is associated with induction of IL-6 and STAT3 expression in the subacute phase of infarction, suggesting a positive feedback loop of the gp130-STAT3 signaling pathway. Positive feedback loops of STAT3 and IL-6 inducing a proinflammatory state have been described in human tumors with somatic mutations at the gp130 locus.16 In WT mice, STAT3 and IL-6 are transiently induced after MI with a gradual decrease in the subacute phases of infarct healing, indicating that gp130-mediated STAT3 activation is actively limited by WT gp130 receptor signaling. In contrast, prolonged and enhanced gp130-STAT3 activation in subacute MI of Y757F mice induces enhanced inflammation and poor post-MI outcome because (1) genetic ablation of the entire gp130 signaling in cardiomyocytes of KO mice resulted in less inflammation and improved post-MI survival compared with Y757F mice, suggesting that it is not lack of gp130-ERK and/or Akt activation but rather sustained and unlimited gp130-JAK-STAT signaling that caused LV rupture and heart failure, and (2) in gene dose studies, the reduction of STAT3 signaling in cardiomyocytes not only normalized STAT3 expression and activation in the subacute phase after MI but also corrected angiogenesis, hypertrophy, inflammation, LV rupture, adverse remodeling, cardiac function, and ultimately survival.
It is important to note that infarcted Y757F mice display a specific transcriptional pattern of proinflammatory mediators in which L-6 was significantly induced, whereas neither tumor necrosis factor-α nor IL-1β expression differed from WT. In addition, a pattern of enhanced activation of the MBL/lectin complement system emerged in infarcted Y757F mice, including MBL-A, MBL-C, and mannan-binding lectin serine peptidase 2.17 MBL-C mRNA especially appeared to be specifically upregulated by gp130-STAT3 signaling because it was corrected to WT levels in Y757F;STAT3low mice.
MBL-C protein levels decreased in nonischemic LVs of all genotypes within 24 hours after MI without changes in MBL-C mRNA levels, suggesting that posttranslational mechanism(s) independent of gp130 signaling are involved in the acute MBL-C clearing or consumption. Interestingly, MBL-C mRNA and MBL-C protein content remained low in WT and Y757F;STAT3low mice during the 2-week follow-up, but they increased steadily in Y757F mice, suggesting a JAK/STAT-dependent pathway controlling MBL-C bioavailability after MI.
The potential importance of MBL-C for cardiac inflammation has been previously suggested on the basis of observations that genetic variants in humans with high MBL-C serum levels have an increased risk for chronic rheumatic valvular disease, coronary artery disease, and MI.18,19 Our observation that MBL-C levels in LVs from patients with ICM were comparable to levels in LVs from nonfailing donors resembles the situation in infarcted Y757F mice, suggesting that increased cardiac MBL-C levels may also be associated with an adverse outcome in patients with ICM.
Maroko et al20 showed that inhibition of complement activation after acute MI by CVF (applied 30 minutes after MI induction) reduces inflammation and myocardial necrosis within the first 24 hours after MI. Because WT and Y757F mice did not differ in survival and function up to 3 days after MI, we allowed the early acute inflammation and started CVF treatment at post-MI day 3 after the acute inflammatory phase had passed. CVF reduced inflammation, which was associated with attenuated functional impairment and improved survival in Y757F mice 2 weeks after MI, supporting the notion that enhanced complement activation in the subacute phase of MI promotes adverse post-MI outcome.
High capillary density and low cardiomyocyte apoptosis in infarcted Y757F mice indicate that enhanced capillary formation and prevention of cardiomyocyte loss alone may not be sufficient to improve outcome after MI. Indeed, cardiac dimensions and function briefly after MI were not different between Y757F and WT mice, suggesting that impaired blood supply and cardiac cell death did not cause high mortality and heart failure in infarcted Y757F mice. In turn, the high degree of inflammation may be responsible for weakening the border zone and elongating the infarct scar in Y757F mice, thereby promoting LV rupture and the development of unfavorable LV geometry during post-MI remodeling. This unfavorable remodeling during removal of necrotic tissue and digestion of extracellular matrix components (collagens) in the border zone and later the infarct scar are likely mediated by MMPs released from inflammatory cells. Indeed, levels of active MMP1 and MMP13 were increased but were not counterbalanced by upregulated expression of their inhibitor TIMP1 in infarcted Y757F mice.
The observation of increased post-MI capillary density in Y757F mice and its reduction in Y757F;STAT3low is in line with known proangiogenic effects of STAT3.12 Whether enhanced angiogenesis further promoted inflammation or vice versa cannot be distinguished with our tools. However, it suggests that increasing cardiac angiogenesis as a sole therapeutic goal may not suffice to improve prognosis after MI.
The mechanisms behind smaller cardiomyocyte dimensions in infarcted Y757F mice compared with WT and its prevention in Y757F;STAT3low mice is currently the focus of ongoing studies.
Our observations shed new light on the role of gp130 downstream signaling pathways in the infarcted heart and suggest that outcome after MI is, at least in part, determined by proper regulation of cardiac gp130-STAT3 signaling. Although, as reported previously, transiently upregulated STAT3 activity in acute cardiac ischemia is cardioprotective,12,21,22 failure in downregulation or a continuous activation of gp130-STAT3 signaling in the subacute infarction seems detrimental because of enhanced inflammation with adverse remodeling, high rupture rates, and heart failure. Such a scenario may take place in patients with high IL-6 serum levels after MI and may offer an explanation for their poor prognosis.
We thank Suzanne Dorfmann for editing the manuscript and Birgit Brandt and Silvia Gutzke for excellent technical assistance.
Sources of Funding
This study was supported by the Deutsche Forschungsgemeinschaft (DR 148/9–3, HI 842/3–1) and the Fondation Leducq.
Fischer P, Hilfiker-Kleiner D. Survival pathways in hypertrophy and heart failure: the gp130-STAT3 axis. Basic Res Cardiol. 2007; 153 (suppl 1): S414–S427.
Eiken HG, Oie E, Damas JK, Yndestad A, Bjerkeli V, Aass H, Simonsen S, Geiran OR, Tonnessen T, Christensen G, Froland SS, Gullestad L, Attramadal H, Aukrust P. Myocardial gene expression of leukaemia inhibitory factor, interleukin-6 and glycoprotein 130 in end-stage human heart failure. Eur J Clin Invest. 2001; 31: 389–397.
Podewski EK, Hilfiker-Kleiner D, Hilfiker A, Morawietz H, Lichtenberg A, Wollert KC, Drexler H. Alterations in Janus kinase (JAK)-signal transducers and activators of transcription (STAT) signaling in patients with end-stage dilated cardiomyopathy. Circulation. 2003; 107: 798–802.
Zolk O, Ng LL, O'Brien RJ, Weyand M, Eschenhagen T. Augmented expression of cardiotrophin-1 in failing human hearts is accompanied by diminished glycoprotein 130 receptor protein abundance. Circulation. 2002; 106: 1442–1446.
Tsutamoto T, Hisanaga T, Wada A, Maeda K, Ohnishi M, Fukai D, Mabuchi N, Sawaki M, Kinoshita M. Interleukin-6 spillover in the peripheral circulation increases with the severity of heart failure, and the high plasma level of interleukin-6 is an important prognostic predictor in patients with congestive heart failure. J Am Coll Cardiol. 1998; 31: 391–398.
Zhang EE, Chapeau E, Hagihara K, Feng GS. Neuronal Shp2 tyrosine phosphatase controls energy balance and metabolism. Proc Natl Acad Sci U S A. 2004; 101: 16064–16069.
Dierssen U, Beraza N, Lutz HH, Liedtke C, Ernst M, Wasmuth HE, Trautwein C. Molecular dissection of gp130-dependent pathways in hepatocytes during liver regeneration. J Biol Chem. 2008; 283: 9886–9895.
Hilfiker-Kleiner D, Hilfiker A, Fuchs M, Kaminski K, Schaefer A, Schieffer B, Hillmer A, Schmiedl A, Ding Z, Podewski E, Poli V, Schneider MD, Schulz R, Park JK, Wollert KC, Drexler H. Signal transducer and activator of transcription 3 is required for myocardial capillary growth, control of interstitial matrix deposition, and heart protection from ischemic injury. Circ Res. 2004; 95: 187–195.
Fuchs M, Hilfiker A, Kaminski K, Hilfiker-Kleiner D, Guener Z, Klein G, Podewski E, Schieffer B, Rose-John S, Drexler H. Role of interleukin-6 for LV remodeling and survival after experimental myocardial infarction. FASEB J. 2003; 17: 2118–2120.
Schepers A, de Vries MR, van Leuven CJ, Grimbergen JM, Holers VM, Daha MR, van Bockel JH, Quax PH. Inhibition of complement component C3 reduces vein graft atherosclerosis in apolipoprotein E3-Leiden transgenic mice. Circulation. 2006; 114: 2831–2838.
Walsh MC, Bourcier T, Takahashi K, Shi L, Busche MN, Rother RP, Solomon SD, Ezekowitz RA, Stahl GL. Mannose-binding lectin is a regulator of inflammation that accompanies myocardial ischemia and reperfusion injury. J Immunol. 2005; 175: 541–546.
Keller TT, van Leuven SI, Meuwese MC, Wareham NJ, Luben R, Stroes ES, Hack CE, Levi M, Khaw KT, Boekholdt SM. Serum levels of mannose-binding lectin and the risk of future coronary artery disease in apparently healthy men and women. Arterioscler Thromb Vasc Biol. 2006; 26: 2345–2350.
Negoro S, Kunisada K, Fujio Y, Funamoto M, Darville MI, Eizirik DL, Osugi T, Izumi M, Oshima Y, Nakaoka Y, Hirota H, Kishimoto T, Yamauchi-Takihara K. Activation of signal transducer and activator of transcription 3 protects cardiomyocytes from hypoxia/reoxygenation-induced oxidative stress through the upregulation of manganese superoxide dismutase. Circulation. 2001; 104: 979–981.
High serum interleukin-6 (IL-6) levels in patients with myocardial infarction or heart failure are consistently associated with reduced cardiac function and a poor outcome (ie, increased morbidity and mortality). This observation contrasts with experimental data that suggest cardioprotective effects mediated by the common receptor subunit glycoprotein-130 (gp130), of IL-6 cytokines. IL-6 cytokine–mediated gp130 signaling involves a complex network of activating and terminating mechanisms controlling 3 major downstream signaling pathways: janus kinase/signal transducer and activator of transcription (STAT), cytoplasmic protein tyrosine phosphatase/extracellular signal-regulated kinase, and phosphoinositide-3-kinase/Akt. The degree and duration of downstream signaling activation depend at least in part on a single tyrosine residue, Y-757, at the gp130 receptor. In cardiomyocyte-specific gp130Y757F mutant mice (Y757F), the Y-757–dependent control is ablated by a point mutation leading to prolonged and enhanced JAK/STAT activation, whereas ERK and Akt signaling is absent. In response to myocardial infarction, Y757F mice displayed higher mortality associated with an increased left ventricular rupture rate, sustained cardiac inflammation, and heart failure. These adverse effects result from prolonged and enhanced STAT3 activation, increased expression of IL-6, and upregulated mannose-binding-lectin-C–mediated complement activation. Pharmacological inhibition of the complement system by cobra venom factor or genetic reduction of STAT3 prevented sustained inflammation and lowered left ventricular rupture rate, heart failure, and mortality in subacute myocardial infarction. This study unravels the potential adverse effects of high IL-6 cytokine levels after myocardial infarction in settings with impaired gp130 downstream signaling. In this regard, sustained IL-6–dependent and gp130-mediated STAT3 activation in subacute infarction promotes cardiac inflammation, adverse remodeling, and heart failure.
The online-only Data Supplement is available with this article at http://circ.ahajournals.org/cgi/content/full/CIRCULATIONAHA.109.933127/DC1.