Interleukin-1α stimulates proinflammatory cytokine expression in human cardiac myofibroblasts

Neil A. Turner, Anupam Das, Philip Warburton, David J. O'Regan, Stephen G. Ball, Karen E. Porter

Abstract

Cardiac myofibroblasts (CMF) play a key role in infarct repair and scar formation following myocardial infarction (MI) and are also an important source of proinflammatory cytokines. We postulated that interleukin-1α (IL-1α), a potential early trigger of acute inflammation post-MI, could stimulate human CMF to express additional proinflammatory cytokines. Furthermore, we hypothesized that these effects may be modulated by the anti-inflammatory cytokine interleukin-10 (IL-10). Human CMF were cultured from atrial biopsies from multiple patients. Interleukin-1β (IL-1β), tumor necrosis factor-α (TNF-α), interleukin-6 (IL-6), and cardiotrophin-1 (CT-1) mRNA expression and secretion were measured using quantitative real-time RT-PCR and enzyme-linked immunosorbent assay. IL-1α (0.001–10 ng/ml, 0–6 h) stimulated IL-1β, TNF-α, and IL-6 mRNA expression with distinct temporal and concentration profiles, resulting in increased cytokine secretion. The response to IL-1α was much greater than with TNF-α. Neither IL-1α nor TNF-α treatment modulated CT-1 mRNA expression. Immunoblotting with phosphospecific antibodies revealed that IL-1α stimulated the extracellular signal-regulated kinase (ERK)-1/2, p38 mitogen-activated protein kinase (MAPK), c-Jun NH2-terminal kinase (JNK), phosphatidylinositol 3-kinase (PI 3-kinase)/protein kinase B (Akt), and nuclear factor (NF)-κB signaling pathways. Pharmacological inhibitor studies indicated roles for PI 3-kinase/Akt and NF-κB pathways in mediating IL-1β expression, and for NF-κB and p38 MAPK pathways in mediating TNF-α expression. IL-1α-induced IL-6 mRNA expression was reduced by p38 MAPK inhibition, but increased by ERK and JNK pathway inhibitors. IL-10 produced a consistent but modest reduction in IL-1α-induced IL-6 mRNA levels (not IL-1β or TNF-α), but this was not reflected by reduced IL-6 protein secretion. In conclusion, IL-1α stimulates human CMF to express IL-1β, TNF-α, and IL-6 via specific signaling pathways, responses that are unaffected by IL-10 exposure.

  • cardiac fibroblasts
  • inflammation
  • signal transduction
  • cytokines
  • interleukin-10

myocardial infarction (MI) arises from coronary artery occlusion and subsequent localized myocardial ischemia. Pathophysiological remodeling of the infarcted myocardium is characterized by several distinct phases, including cardiomyocyte death (necrosis and apoptosis), acute inflammation (elevated chemokine and proinflammatory cytokine levels and infiltration of inflammatory cells in the infarct area), formation of granulation tissue [cardiac myofibroblast (CMF) and macrophage accumulation, degradation of extracellular matrix, and neovascularization], extracellular matrix deposition (scar formation), and finally resolution of the inflammatory response (production of anti-inflammatory cytokines and apoptosis of scar CMF) (5, 15, 16, 39). Thus the infarcted myocardial tissue is dynamic and hosts several different cell types (neutrophils, monocyte/macrophages, mast cells, CMF, and vascular cells) that sequentially infiltrate the injured myocardium through a highly coordinated series of events.

The initial triggers of the post-MI inflammatory response are thought to include complement activation, generation of reactive oxygen species, and increased levels of proinflammatory cytokines that act as stimuli for inflammatory cell infiltration to the reperfused myocardial infarct (5, 15, 16, 39). Release of the proinflammatory cytokine interleukin-1α (IL-1α) from necrotic cells was recently identified as a key inducer of chemokine secretion and subsequent neutrophilic infiltration in necrotic liver (10). Although such a role for IL-1α in the infarcted myocardium has not yet been firmly established, it seems likely based on the observations that IL-1α mRNA and protein are expressed by cardiomyocytes (47) and that IL-1α protein levels are increased in the infarcted myocardium following myocyte necrosis (41). Characterizing the response of myocardial cells to IL-1α is therefore likely to be of importance in delineating the early events that occur in the myocardium post-MI.

Cardiac fibroblasts are the most abundant cell type in the adult human heart, accounting for two-thirds of total myocardial cell number (6, 9, 21, 46). In damaged myocardium (e.g., post-MI), the normally quiescent fibroblasts undergo phenotypic alteration to become myofibroblasts that play a key role in infarct repair and scar formation (39). CMF are an important source of proinflammatory cytokines, including tumor necrosis factor-α (TNF-α), interleukin-1β (IL-1β), and interleukin-6 (IL-6). Cultured cardiac (myo)fibroblasts secrete these cytokines in response to specific stimuli that occur in the post-MI heart, including hypoxia, mechanical stretch, and increased levels of cytokines (32). However, the ability of IL-1α to stimulate synthesis and secretion of other proinflammatory cytokines by CMF has not been investigated previously.

The detrimental effects of chronic proinflammatory cytokine expression are tempered by the action of anti-inflammatory cytokines, such as the macrophage-derived cytokine interleukin-10 (IL-10). Deletion of the IL-10 gene increases myocardial proinflammatory cytokine levels and increases infarct size and mortality rates in transgenic mouse models (20, 26, 48). One mechanism by which IL-10 exerts anti-inflammatory effects is by reducing macrophage expression of proinflammatory cytokines, including TNF-α, IL-1β, and IL-6 (24). We postulated that IL-10 may act on CMF in a similar way to reduce proinflammatory cytokine expression.

The aims of this study were to determine the effects of IL-1α on proinflammatory cytokine expression in cultured human CMF, to identify the intracellular signaling pathways responsible, and to investigate modulation by, IL-10.

METHODS

Reagents.

All cell culture reagents were purchased from Invitrogen (Paisley, Scotland, UK), except FCS, which was from Biosera (Ringmer, East Sussex, UK). Recombinant human IL-1α, TNF-α, and IL-10 were from BioSource (Paisley, Scotland, UK). Actinomycin D was from Sigma (Poole, UK). Pharmacological signaling pathway inhibitors were obtained from Calbiochem (Nottingham, UK), except LY-294002, which was from Alexis Biochemicals (Nottingham, UK).

Cell culture.

Right atrial appendage biopsies from patients undergoing elective coronary artery bypass surgery at the Leeds General Infirmary were obtained following local ethical committee (The Leeds Teaching Hospitals Local Research Ethics Committee) approval and informed patient consent. Primary cultures of cardiac fibroblasts were harvested, characterized, and cultured as we have described previously (28, 33, 44). Cells exhibited a CMF phenotype as determined by positive staining for both α-smooth muscle actin and vimentin. The CMF phenotype was observed at passage 1 and maintained through at least passage 5 (28). Experiments were performed on cells from passage 3 to passage 5 from a cohort of 30 patients (26 male, 4 female) with a mean age of 62.8 ± 1.8 (range 45–83) yr, 5 of whom were receiving treatment for type 2 diabetes. Cells were serum-starved for 48 h before performing experiments in basal medium (DMEM supplemented with 0.4% FCS), unless stated otherwise.

Quantitative RT-PCR.

Cellular RNA was extracted from cells at the end of the incubation period, and cDNA was prepared as described previously (43). Real-time PCR was performed in duplicate using the Applied Biosystems 7500 Real-Time PCR System. Intron-spanning primers and Taqman probes for human IL-1β (Hs00174097_m1), TNF-α (Hs00174128_m1), IL-6 (Hs00174131_m1), cardiotrophin (CT)-1 (Hs00173498_m1), leukemia inhibitory factor (LIF) (Hs00171165_m1), oncostatin M (OSM) (Hs00171165_m1), and IL-10 (Hs00174086_m1) were from Applied Biosystems (www.appliedbiosystems.com). Data were expressed as a percentage of glyceraldehyde-3-phosphate dehydrogenase (GAPDH) endogenous control levels (Hs99999905_m1 primers) using the formula 2Math× 100, or expressed relative to control sample using the formula 2Math, in which CT is the threshold cycle number.

Enzyme-linked immunosorbent assay.

Cells cultured in six-well plates were serum-starved for 48 h before exposure to appropriate stimuli in 1.5 ml medium for 24 h before collecting conditioned media. Media were centrifuged to remove cellular debris, and samples were stored at −40°C for subsequent analysis. Enzyme-linked immunosorbent assays (ELISAs) were performed according to the manufacturer's instructions (R&D Systems, Abingdon, UK) using undiluted (IL-1β and TNF-α ELISAs) or 1:200 diluted (IL-6 ELISA) samples.

Immunoblotting.

For signaling experiments, serum-starved cells were exposed to serum-free medium containing 10 ng/ml IL-1α for 5–60 min before preparing whole cell homogenates and immunoblotting as we have described previously (42). Activation of the extracellular signal-regulated kinase (ERK-1/2), p38 mitogen-activated protein kinase (p38-MAPK), c-Jun NH2-terminal kinase (JNK), phosphoinositide 3-kinase (PI 3-kinase)/protein kinase B (Akt), and nuclear factor-κB (NF-κB) pathways was determined using phosphospecific antibodies (nos. 9106, 9211, 9255, 4058, and 9246, respectively; Cell Signaling Technology, Hitchin, UK). Total expression levels were assessed using appropriate expression antibodies (nos. 9102, 2371, 9252, 9272, and 9242, respectively; Cell Signaling Technology). Signal transducer and activator of transcription 3 (STAT3) activation was determined using phosphospecific STAT3 (Ser727) antibody (no. 9136; Cell Signaling Technology). We have previously confirmed the effectiveness of this antibody by its ability to detect phospho-STAT3 following ANG II treatment of HEK293 cells transfected with the angiotensin type 1A receptor (data not shown). Equal sample loading was confirmed using monoclonal β-actin antibody (no. ab8226; Abcam, Cambridge, UK). Immunolabeled bands were visualized by the SuperSignal West Pico chemiluminescence kit (Perbio, Cramlington, UK). Densitometric analysis was performed using a flat-bed scanner and ImageQuant software (Amersham Life Sciences, Amersham, UK).

Statistical analysis.

Results are expressed as means ± SE with n representing the number of experiments on cells from individual patients. Differences between treatment groups were analyzed using paired t-tests (absolute data) or paired ratio t-tests (normalized data). Dose-response and time course data were compared using repeated-measures one-way ANOVA and the Newman-Keul's post hoc test. All statistical analyses were performed using GraphPad Prism software (www.graphpad.com). P < 0.05 was considered statistically significant.

RESULTS

Effect of IL-1α on proinflammatory cytokine expression in human CMF.

Cultured human CMF from five different patients were treated with a range of concentrations of IL-1α (0.001–10 ng/ml) in basal medium for 6 h before extracting RNA and performing real-time RT-PCR to measure relative mRNA levels of IL-1β, TNF-α, IL-6, and CT-1 (Fig. 1). Basal levels of mRNA expression were highest for IL-6 (3.4% GAPDH) and were approximately one, two, and four orders of magnitude lower for IL-1β, CT-1, and TNF-α, respectively. IL-1α increased IL-1β, TNF-α, and IL-6 mRNA levels in a concentration-dependent manner, with maximal effects observed in response to the 1 ng/ml concentration for IL-1β and IL-6 mRNA (Fig. 1, A–C). In contrast, CT-1 mRNA levels did not increase in response to IL-1α treatment at any of the concentrations used (Fig. 1D).

Fig. 1.

Concentration profile of proinflammatory cytokine mRNA expression in response to interleukin (IL)-1α. Cardiac myofibroblasts (CMF) from 5 patients were exposed to 0.001–10 ng/ml IL-1α for 6 h before extracting RNA and performing real-time RT-PCR with primers for IL-1β (A), tumor necrosis factor (TNF)-α (B), IL-6 (C), and cardiotrophin (CT)-1 (D). Data are expressed as percentage relative to glyceraldehyde-3-phosphate dehydrogenase (GAPDH) endogenous control. ANOVA: P < 0.001 (IL-1β), P < 0.001 (TNF-α), P < 0.01 (IL-6), and P > 0.05 (CT-1). Newman-Keul's post hoc test: ***P < 0.001, **P < 0.01, and *P < 0.05 for effect of IL-1 (n = 5). Note different scales on y-axes.

Analysis of the time course of the IL-1α response (Fig. 2) revealed different temporal expression profiles for each of the proinflammatory cytokines. IL-1β mRNA levels were elevated 140-fold after 2 h and continued to rise, peaking at 320-fold after 4–5 h (Fig. 2A). In marked contrast, TNF-α mRNA levels were maximal after 2 h (>4,600-fold increase) and returned rapidly toward basal levels thereafter, with expression after 6 h being 125 times higher than basal (Fig. 2B). IL-6 mRNA levels were elevated 13-fold within 2 h of IL-1α treatment (10 ng/ml), and remained steady for at least 6 h (Fig. 2C). CT-1 mRNA levels were not increased at any time point following IL-1α treatment (Fig. 2D).

Fig. 2.

Temporal profile of proinflammatory cytokine mRNA expression in response to IL-1α. CMF from 3 patients were exposed to 10 ng/ml IL-1α for 2–6 h before extracting RNA and performing real-time RT-PCR with primers for IL-1β (A), TNF-α (B), IL-6 (C), and CT-1 (D). Data are normalized to GAPDH mRNA levels and expressed relative to control (0 h) sample. ANOVA (ratios): P < 0.001 (IL-1β), P < 0.001 (TNF-α), P < 0.001 (IL-6), and P > 0.05 (CT-1). Newman-Keul's post hoc test: ***P < 0.001 for effect of IL-1 (n = 3). Note different scales on y-axes. E: cells from 3 patients were stimulated with IL-1α for 2 h before inhibiting transcription with 5 μg/ml actinomycin D and extracting RNA at regular intervals thereafter. Real-time RT-PCR was performed with specific cytokine primer sets, and data were expressed relative to GAPDH levels, which were unaffected by actinomycin D treatment.

Given the disparate effects of IL-1α on expression of IL-6 and CT-1, two members of the IL-6 family of cytokines, we investigated the temporal effect of IL-1α stimulation on two further members of the IL-6 family, LIF and OSM. Of the four IL-6 family cytokines studied, LIF exhibited the highest basal mRNA level (∼6 times higher than IL-6), and OSM exhibited the lowest basal mRNA level (Table 1). IL-1α stimulated LIF mRNA expression over a 2- to 6-h period, with peak expression (18-fold increase) observed after 4 h (Table 1). In contrast, OSM mRNA expression was not detectable in unstimulated cells and was not significantly induced by IL-1α treatment (Table 1).

View this table:
Table 1.

Time course of IL-1α stimulation on mRNA levels of IL-6 family cytokines in human cardiac myofibroblasts

The rapid decline in TNF-α mRNA levels (Fig. 2B) relative to those of IL-1β (Fig. 2A) and IL-6 (Fig. 2C) prompted us to explore the underlying mechanism. CMF were stimulated with IL-1α for 2 h before inhibiting transcription with 5 μg/ml actinomycin D and measuring the decay of individual cytokine transcripts over the next 2 h. GAPDH mRNA levels were unaffected by actinomycin D treatment over the time course of the experiment, so cytokine mRNA levels were expressed relative to those of GAPDH. The mean half-life of TNF-α mRNA was 28 min, whereas those of IL-1β and IL-6 transcripts were in excess of 2 h (Fig. 2E).

In a recent study, we reported that TNF-α, another key proinflammatory cytokine involved in the post-MI remodeling process, could stimulate IL-1α, IL-1β, and IL-6 gene expression in human CMF (43). We therefore compared the relative potencies of IL-1α and TNF-α on proinflammatory cytokine mRNA expression and investigated the effects of their coadministration (Fig. 3). IL-1α (10 ng/ml) was ∼10 times more potent than an equivalent concentration of TNF-α in stimulating IL-1β, TNF-α, and IL-6 gene expression (Fig. 3, A–C). When TNF-α and IL-1 were added together, a further modest increase was observed that reached statistical significance for IL-1β only (Fig. 3, A–C). Neither TNF-α nor IL-1, alone or in combination, stimulated an increase in CT-1 mRNA expression (Fig. 3D). Further experiments revealed that several other stimuli, including FCS, isoproterenol, and insulin, were unable to modulate CT-1 mRNA levels in human CMF (data not shown).

Fig. 3.

Effect of IL-1α and TNF-α on proinflammatory cytokine expression. A–D: CMF from 5 patients were exposed to 10 ng/ml TNF-α or 10 ng/ml IL-1α (alone or in combination) for 6 h before extracting RNA and performing real-time RT-PCR with primers for IL-1β (A), TNF-α (B), IL-6 (C), and CT-1 (D). Data are normalized to GAPDH mRNA levels and expressed relative to IL-1-treated sample. Paired ratio t-test: ***P < 0.001, **P < 0.01, *P < 0.05, or not significant (NS) for effect of TNF-α and/or IL-1α vs. vehicle control (C), or for IL-1 vs. IL-1 + TNF as indicated (n = 5). E-F: cells from a further 6 patients were exposed to 10 ng/ml TNF-α or 10 ng/ml IL-1α (alone or in combination) for 24 h before collecting conditioned media and measuring IL-1β (E) and IL-6 (F) levels by enzyme-linked immunosorbent assay (ELISA). Note that IL-1β levels are expressed as pg/ml, whereas IL-6 levels are in ng/ml. Paired t-test: **P < 0.01, *P < 0.05, or NS for effect of TNF-α and/or IL-1α vs. vehicle control, or for IL-1 vs. IL-1 + TNF as indicated (n = 6).

ELISAs were performed to determine whether increased mRNA expression resulted in increased protein secretion. CMF were treated for 24 h in basal medium supplemented with TNF-α and/or IL-1α before collecting conditioned media and quantifying IL-1β, TNF-α, and IL-6 protein levels by ELISA. IL-1β was present at very low levels (0.2 pg/ml) in medium from unstimulated control cells, and increased to 0.5 pg/ml following exposure to TNF-α alone (Fig. 3E). In comparison, IL-1α treatment elicited a much greater increase in secreted IL-1β levels (8-fold over basal), and addition of TNF-α and IL-1α together stimulated a synergistic increase (18-fold), resulting in a mean concentration of 3.2 pg/ml IL-1β (Fig. 3E).

Basal levels of TNF-α were very low in control conditioned medium (1.9 ± 0.2 pg/ml) and rose slightly to 8.5 ± 2.8 pg/ml (P = 0.07, n = 6) after 24 h IL-1α treatment. Measurement of TNF-α-induced TNF-α secretion was not feasible because of interference of exogenous TNF-α with the assay.

Considerable levels of basal IL-6 secretion were observed in unstimulated control cells, resulting in a mean concentration of 1.8 ng/ml in the culture medium after 24 h (Fig. 3F), almost four orders of magnitude higher than basal IL-1β levels. IL-6 secretion was further increased in response to TNF-α (3.5-fold) or IL-1α (19-fold) alone, and addition of both cytokines together stimulated a synergistic increase (33-fold), resulting in a mean final concentration of 58.3 ng/ml IL-6 (Fig. 3F). We did not measure secreted CT-1 protein levels because of the lack of effect of IL-1α or TNF-α on CT-1 mRNA levels (Figs. 1D, 2D, and 3D).

Role of signaling pathways in IL-1α-induced proinflammatory cytokine expression.

IL-1α acting at the IL-1 receptor (IL-1R1) has the potential to stimulate multiple intracellular signaling pathways (13). However, the specific pattern of activation depends largely on the cell type studied. We investigated the ability of IL-1α to activate MAPK (ERK, p38, JNK), PI 3-kinase/Akt, and NF-κB pathways in human CMF by immunoblotting cell homogenates with phosphospecific antibodies (Fig. 4). IL-1α stimulated phosphorylation (activation) of ERK-1/2, p38, JNK, and Akt over a 10- to 60-min period. Phosphorylation of inhibitory factor κB-α (IκB-α), the inhibitory binding partner of NF-κB, was observed after 5 min and declined rapidly thereafter, likely because of proteolytic degradation of IκB-α protein, as evidenced by the IκB-α expression immunoblot (Fig. 4).

Fig. 4.

Signaling pathways activated by IL-1α. CMF were exposed to 10 ng/ml IL-1α for 5–60 min before preparing whole cell homogenates and immunoblotting with phosphospecific and expression antibodies for components of different signaling pathways. β-Actin expression was assessed to confirm equal sample loading. Approximate molecular sizes of bands (kDa) are indicated on right. Line graphs depict densitometric analysis of phosphorylated proteins expressed as a percentage of maximum (n = 3). ANOVA values: phospho (p)-extracellular signal-regulated kinase (ERK)-1/2 (p44: P < 0.001, p42: P < 0.001), phospho-p38 (P < 0.05), phospho-c-Jun NH2-terminal kinase (JNK) (p54: P < 0.05, p46: P < 0.05), phospho-protein kinase B (Akt) (P < 0.01), phospho-inhibitory factor κB-α (IκB-α) (P < 0.001), and β-actin (P > 0.05).

We then employed pharmacological signaling inhibitors to determine which of these signaling pathways was important for IL-1α-induced cytokine gene expression (Fig. 5A). Inhibitors were used at concentrations that we have previously shown to effectively and selectively inhibit relevant pathways in this cell type (43). Both PD-98059 (ERK pathway inhibitor) and SP-600125 (JNK inhibitor) had comparable effects on cytokine expression. Both agents significantly increased basal and IL-1α-induced IL-6 mRNA levels (Fig. 5A, left), and there was an apparent trend toward increased IL-1α-induced TNF-α mRNA levels with these inhibitors (Fig. 5A, middle), although this was not statistically significant (P = 0.08 and P = 0.14 for PD-98059 and SP-600125, respectively). Neither inhibitor significantly affected basal or IL-1α-stimulated IL-1β mRNA levels (Fig. 5A, left). The p38 MAPK inhibitor SB-203580 significantly reduced IL-1α-stimulated TNF-α (Fig. 5A, middle) and IL-6 (Fig. 5A, right) mRNA levels, but again had no effect on IL-1β mRNA levels (Fig. 5A, left). The PI 3-kinase inhibitor LY-294002 selectively reduced IL-1α-induced IL-1β mRNA levels (Fig. 5A, left), without modulating TNF-α or IL-6 expression (Fig. 5A, middle and right). Finally, the IKK-2 inhibitor IMD-0354 reduced IL-1α-induced IL-1β and TNF-α mRNA (Fig. 5A, left and middle), suggesting important roles for the NF-κB pathway in mediating IL-1β and TNF-α gene expression in response to IL-1α. IMD-0354 also increased basal IL-6 mRNA expression, although it did not affect IL-1α-induced IL-6 expression (Fig. 5A, right). A summary of the role of the different signaling pathways in mediating IL-1α-induced cytokine expression in human CMF is depicted in Fig. 5B.

Fig. 5.

Role of signaling pathways in IL-1α-induced proinflammatory cytokine expression. A: CMF from 4 patients were pretreated for 1 h with the following signaling pathway inhibitors: 30 μM PD-98059 (PD), 10 μM SB-203580 (SB), 10 μM SP-600125 (SP), 10 μM LY-294002 (LY), or 10 μM IMD-0354 (IMD). Ctrl, control. Cells were then exposed to basal medium alone (unstimulated, open bars) or basal medium supplemented with 10 ng/ml IL-1α (stimulated, filled bars) for 2 h. RNA was extracted, and real-time RT-PCR was performed with primers specific for IL-1β (left), TNF-α (middle), and IL-6 (right) mRNA expression. Data are normalized to GAPDH mRNA levels and expressed relative to IL-1-treated sample. Paired ratio t-test: ***P < 0.001, **P < 0.01, and *P < 0.05 for effect of inhibitor on IL-1α-stimulated cells; ###P < 0.001 and ##P < 0.01 for effect of inhibitor on unstimulated cells (n = 4). B: summary of the role of different signaling pathways in mediating IL-1α-induced cytokine expression in human CMF.

Effects of IL-10 on IL-1α-induced proinflammatory cytokine expression.

IL-10 inhibits expression of proinflammatory cytokines in macrophages (24), but its effects on cardiac (myo) fibroblasts are unreported. We therefore investigated whether IL-10 (20 ng/ml) could modulate IL-1α-induced proinflammatory cytokine expression in human CMF (Fig. 6). IL-10 had no effect on basal or IL-1α-induced levels of IL-1β or TNF-α mRNA (Fig. 6, A and B). However, a small (20%) but consistent inhibitory effect of IL-10 on basal and IL-1α-induced IL-6 mRNA expression was observed (Fig. 6C). Similar results were obtained if cells were treated with TNF-α rather than IL-1α as the stimulus (data not shown). To determine whether the decrease in IL-6 mRNA levels resulted in decreased protein secretion, we collected conditioned media from cells treated with IL-1α and IL-10 for 24 h and measured IL-6 protein levels by ELISA. In contrast to the RT-PCR data, IL-10 had no effect on basal or IL-1α-induced IL-6 protein secretion (Fig. 6D).

Fig. 6.

Effects of IL-10 on IL-1α-induced proinflammatory cytokine expression. A–C: CMF from 5 patients were exposed to IL-1α (10 ng/ml) with or without IL-10 (20 ng/ml) for 6 h before extracting RNA and performing RT-PCR with primers for IL-1β (A), TNF-α (B), or IL-6 (C). Data are normalized to GAPDH mRNA levels. Paired ratio t-test: *P < 0.05 or NS for effect of IL-10 on various treatment groups (n = 5). D: cells from a further 5 patients were exposed to IL-1α (10 ng/ml) with or without IL-10 (20 ng/ml) for 24 h before collecting conditioned media and measuring IL-6 levels by ELISA. Paired t-test: NS for effect of IL-10 on various treatment groups (n = 5). E: CMF from 3 patients were exposed to IL-10 (20 ng/ml) for 5–60 min before preparing whole cell homogenates and immunoblotting with phosphospecific signal transducer and activator of transcription (STAT) 3 (Ser727) antibody. β-Actin expression was assessed to confirm equal sample loading. Line graph depicts densitometric analysis of phosphorylated STAT3 expressed as relative to β-actin and as a percentage of time 0. ANOVA: P < 0.05 (n = 3). F: CMF were exposed to 10 ng/ml IL-1α for 2–6 h before extracting RNA and measuring IL-10 mRNA levels by RT-PCR. Data are expressed as a percentage of GAPDH levels. ANOVA: P > 0.05 (n = 3).

Further studies investigated whether human CMF express functional IL-10 receptors (IL-10R). The major signaling pathway coupling the IL-10R to cellular responses is the JAK1-STAT3 pathway (29). We therefore measured STAT3 (Ser727) phosphorylation in human CMF as a marker of IL-10R activation. Basal STAT3 phosphorylation was observed in unstimulated cells, but was not modulated markedly by 5–60 min treatment with 20 ng/ml IL-10 (Fig. 6E). A possible explanation for the lack of effect of IL-10 on human CMF could be that the cells secrete a high level of endogenous IL-10 that reduces the potency of exogenously applied IL-10. However, this did not appear to be the case, since IL-10 mRNA was undetectable in CMF under basal conditions and was not affected markedly by IL-1α treatment (Fig. 6F).

DISCUSSION

The results of our study reveal that human CMF cultured from different patients consistently respond to IL-1α through increased expression of specific proinflammatory cytokines (IL-1β, TNF-α, IL-6, and LIF), but not others (CT-1 and OSM). The cytokine response to IL-1α was reproducibly higher than that induced by TNF-α, and addition of both cytokines together induced additive or synergistic responses. IL-1α stimulated the ERK, p38, JNK, PI 3-kinase/Akt, and NF-κB pathways, and signaling inhibitor experiments revealed differential roles for these pathways in inducing expression of individual cytokines. Finally, the anti-inflammatory cytokine IL-10 had little or no effect on proinflammatory cytokine expression, likely reflecting a lack of functionally coupled IL-10R in this cell type.

IL-1α levels are elevated in the infarcted myocardium (41), and transgenic mice with cardiac-specific overexpression of IL-1α develop left ventricular hypertrophy (31), whereas those with ubiquitous IL-1α overexpression die of heart failure (18). In contrast to IL-1β, IL-1α is not secreted from human cells but accumulates intracellularly and at the plasma membrane, being released only when cells undergo necrotic cell death (10, 11). Although IL-1α and IL-1β are products of different genes, they signal via the same cell surface receptor, IL-1R1 (11), activation of which is detrimental in post-MI myocardial remodeling (1, 7, 40). In one recent example, genetically targeted IL-1R1-null mice were shown to exhibit reduced ventricular dilatation following experimental MI, without any change in infarct size (7). This protective effect of IL-1R1 deficiency was not due to reduced myocyte injury, but rather to suppression of the acute inflammatory response (reduced cytokine and chemokine levels and reduced neutrophil infiltration) and attenuated fibrosis [less myofibroblast infiltration, lower transforming growth factor-β levels, less collagen deposition, decreased matrix metalloproteinase (MMP) levels] (7). Our current observation that IL-1α induces CMF to express proinflammatory cytokines may help to explain the reduced inflammatory response observed in IL-1R1 knockout mice (7) and in rats overexpressing the IL-1 receptor antagonist (40).

Although both IL-1α and TNF-α could induce cytokine expression, IL-1α was consistently much more potent than TNF-α. Previous studies in rat cardiac fibroblasts have reported larger responses to IL-1β compared with TNF-α at the levels of cell migration (27), MMP expression (37), angiotensin receptor expression (17), and reduced collagen synthesis (37). Thus both human and rat cardiac (myo)fibroblasts appear to be more responsive to IL-1 than TNF-α.

Cardiac fibroblasts are an important source of TNF-α and IL-1β in the myocardium, and their expression has been shown to be induced by a diverse range of stimuli, including hypoxia, ANG II, catecholamines, cytokines, and mechanical stretch (32). However, little is known of the underlying signaling pathways that regulate TNF-α and IL-1β gene expression in cardiac fibroblasts. We recently demonstrated that TNF-α could stimulate IL-1β mRNA expression in human CMF via a mechanism involving activation of the p38 MAPK, PI 3-kinase/Akt, and NF-κB pathways, but not the ERK pathway (43). Our current observations implicate the NF-κB and PI 3-kinase/Akt pathways, but not the ERK, JNK, or p38 MAPK pathways, in IL-1α-induced IL-1β gene expression in this cell type. Moreover, IL-1α-induced TNF-α mRNA expression was mediated via the p38 MAPK and NF-κB pathways. The half-life of TNF-α mRNA was significantly less than IL-1β or IL-6, likely accounting for the transient nature of the increase in TNF-α mRNA levels compared with the other cytokines.

IL-6 is synthesized by both cardiomyocytes and cardiac fibroblasts (2). In cardiac fibroblasts, IL-6 expression can be stimulated by catecholamines (8, 49), ANG II (36), and TNF-α (43). The signal transduction pathways that mediate changes in IL-6 gene expression in cardiac fibroblasts include the ERK, p38 MAPK, NF-κB, and PI 3-kinase/Akt cascades (30, 36, 43, 49). In the present study, we show for the first time that IL-1α can also induce IL-6 mRNA expression and secretion in CMF and that this occurs predominantly via the p38 MAPK pathway. PD-98059 and SP-600125, inhibitors of the ERK and JNK pathways, respectively, increased basal and IL-1α-stimulated IL-6 mRNA levels, suggesting that IL-6 mRNA expression may be negatively regulated by the ERK and JNK pathways in these cells. SP-600125 is less selective than the other inhibitors used in this study and can inhibit several other protein kinases in addition to JNK (4); hence, the data obtained with this inhibitor should be interpreted with a degree of caution.

We did not observe any modulatory effect of IL-1α or TNF-α on CT-1 mRNA levels. In a prior study on cultured human cardiac fibroblasts, CT-1 protein was not detected in conditioned medium, despite detection of other IL-6 family members (IL-6, LIF, IL-11) at nanogram per milliliter concentrations (2). However, CT-1 mRNA and protein have been detected in cultures of neonatal rat cardiac fibroblasts (19, 22, 35). These discrepancies may be due to species and/or developmental differences (adult human vs. neonatal rat), although these studies did not directly address this. It is noteworthy that the stimuli that evoked CT-1 expression in neonatal rat cardiac fibroblasts included ANG II and basic fibroblast growth factor (19, 35), which are noninflammatory mitogenic stimuli. In other cell types, CT-1 expression has been shown to be stimulated by norepinephrine, insulin, glucose, and reactive oxygen species (38). The CT-1 promoter does not contain an NF-κB site (14), whereas the promoter regions of the IL-1β, IL-6, and TNF-α genes contain several such sites (3, 12, 45), and this may explain the inability of inflammatory stimuli (IL-1α and TNF-α) to induce CT-1 expression in our study. It is noteworthy that we also did not observe any effect of isoproterenol, insulin, or FCS on CT-1 mRNA expression, suggesting that regulation of CT-1 expression in human CMF is different from other cell types (38).

Analysis of additional members of the IL-6 family revealed that LIF, but not OSM, mRNA expression was induced by IL-1α in human CMF. Adult human cardiac fibroblasts in culture have been previously reported to secrete IL-6, LIF, and IL-11, but not CT-1, OSM, or ciliary neurotrophic factor (2). Our data concur with that earlier report and further establish that IL-6 and LIF (but not CT-1 or OSM) can be induced by IL-1α stimulation in this cell type. In common with CT-1, the promoter region of the human OSM gene does not contain an NF-κB site, but rather is regulated by activator protein-1 and STAT5 (25), likely explaining its lack of sensitivity to proinflammatory stimuli such as IL-1α.

There are several lines of evidence suggesting that IL-10 is important in the post-MI remodeling of the myocardium (20, 26, 48). For example, IL-10 knockout mice exhibit increased myocardial IL-6, IL-1β, and TNF-α levels in response to an inflammatory stimulus compared with wild-type mice (26). Moreover, deletion of the IL-10 gene increases infarct size and subsequent mortality rates following experimental MI (20, 48). Conversely, a more recent report showed that left ventricular remodeling in IL-10 knockout mice was similar to that in wild-type mice (50). Although IL-10 is synthesized primarily by macrophages, there is also evidence of its secretion by cardiac fibroblasts (23, 34). However, in our study, IL-10 mRNA expression was undetectable in unstimulated human CMF and was not markedly increased by IL-1α treatment. We found that IL-10 did not modulate basal or IL-1α-induced expression of IL-1β and TNF-α mRNA, and the modest, yet consistent, inhibitory effect on IL-6 mRNA levels did not lead to reduced IL-6 secretion. Thus IL-10 is unable to modulate proinflammatory cytokine secretion from CMF, at least in response to an IL-1α stimulus. There are no previous reports describing IL-10R expression in cardiac fibroblasts. One explanation for the muted effect of IL-10 on human CMF could be a lack of expression of the IL-10R, a theory supported by our observed lack of IL-10-induced STAT3 phosphorylation in this cell type.

In conclusion, we have characterized the intracellular signaling pathways by which IL-1α, a proinflammatory cytokine released by necrotic cardiomyocytes, induces expression of additional proinflammatory cytokines in human CMF. Furthermore, we have demonstrated that IL-10 does not modulate IL-1α-induced proinflammatory cytokine expression in this important cell type, most likely because of lack of functionally coupled IL-10R. Understanding the mechanisms by which early markers of myocardial damage initiate the inflammatory response remains of key significance in designing strategies to reduce adverse post-MI remodeling in humans.

GRANTS

N. A. Turner is funded by a Research Councils United Kingdom Academic Fellowship.

Acknowledgments

We are grateful to Stacey Galloway for cell culture expertise.

REFERENCES

View Abstract