Biochemical Journal

Research article

MCP-1 causes cardiomyoblast death via autophagy resulting from ER stress caused by oxidative stress generated by inducing a novel zinc-finger protein, MCPIP

Craig W. Younce , Pappachan E. Kolattukudy

Abstract

MCP-1 (monocyte chemotactic protein-1) plays a critical role in the development of heart failure that is known to involve apoptosis. How MCP-1 contributes to cell death involved in the development of heart disease is not understood. In the present study we show that MCP-1 causes death in cardiac myoblasts, H9c2 cells, by inducing oxidative stress which causes ER stress leading to autophagy via a novel zinc-finger protein, MCPIP (MCP-1-induced protein). MCPIP expression caused cell death, and knockdown of MCPIP attenuated MCP-1induced cell death. It caused induction of iNOS (inducible NO synthase), translocation of the NADPH oxidase subunit phox47 from the cytoplasm to the membrane, production of ROS (reactive oxygen species), and induction of ER (endoplasmic reticulum) stress markers HSP40 (heat-shock protein 40), PDI (protein disulfide-isomerase), GRP78 (guanine-nucleotide-releasing protein 78) and IRE1α (inositol-requiring enzyme 1α). It also caused autophagy, as indicated by beclin-1 induction, cleavage of LC3 (microtubule-associated protein 1 light chain 3) and autophagolysosome formation, and apoptosis, as indicated by caspase 3 activation and TUNEL (terminal deoxynucleotidyltransferase-mediated dUTP nick-end labelling) assay. Inhibitors of oxidative stress, including CeO2 nanoparticles, inhibited ROS formation, ER stress, autophagy and cell death. Specific inhibitors of ER stress inhibited autophagy and cell death as did knockdown of the ER stress signalling protein IRE1. Knockdown of beclin-1 and autophagy inhibitors prevented cell death. This cell death involved caspase 2 and caspase 12, as specific inhibitors of these caspases prevented MCPIP-induced cell death. Microarray analysis showed that MCPIP expression caused induction of a variety of genes known to be involved in cell death. MCPIP caused activation of JNK (c-Jun N-terminal kinase) and p38 and induction of p53 and PUMA (p53 up-regulated modulator of apoptosis). Taken together, these results suggest that MCPIP induces ROS/RNS (reactive nitrogen species) production that causes ER stress which leads to autophagy and apoptosis through caspase 2/12 and IRE1α–JNK/p38–p53–PUMA pathway. These results provide the first molecular insights into the mechanism by which elevated MCP-1 levels associated with chronic inflammation may contribute to the development of heart failure.

  • apoptosis
  • autophagy
  • endoplasmic reticulum stress
  • monocyte chemotactic protein-1 (MCP-1)
  • monocyte chemotactic protein-1-induced protein (MCPIP)

INTRODUCTION

Inflammation plays a critical role in the development of cardiovascular diseases [1,2]. The role of leucocyte infiltration and the role of MCP-1 (monocyte chemotactic protein-1) in cardiovascular disease progression are widely recognized [3,4]. There is overwhelming evidence that MCP-1 is involved in the development of cardiovascular diseases in humans [4,5]. Many lines of experimental evidence from animal models also provide strong evidence for a critical role for MCP-1 and its receptor CCR2 in the development of cardiovascular diseases. These include results obtained with CCR2-deficient mice [6], MCP-1-deficient mice [7] and transgenic mice with targeted expression of MCP-1 [8]. Death of cardiomyocytes is known to be involved in heart failure. We previously reported that treatment of monocytes with MCP-1 induces a novel zinc-finger protein, MCPIP (MCP-1-induced protein), which can induce cell death [9]. The presence of MCPIP transcripts and protein was associated with the apoptotic cardiomyocytes in a murine model of heart failure with cardiomyocyte-targeted expression of MCP-1. Association of MCPIP with ischaemic heart disease was also found in the human disease [9]. Although such indirect indications suggest a role for MCPIP in causing cell death associated with heart failure, how MCPIP might cause cell death is not known.

In transgenic mice with cardiomyocyte-targeted expression of MCP-1, the development of heart failure was found to be associated with elevated levels of MCPIP [9] and histochemically detectable ROS (reactive oxygen species) and RNS (reactive nitrogen species) production [10]. Oxidative stress is a known inducer of ER (endoplasmic reticulum) stress [11,12] and the MCP-1 transgenic mice showed ER stress prior to developing heart failure [13]. ER stress is known to cause autophagy [14,15] that can lead to cell death [16,17]. On the basis of such observations on MCP-transgenic mice and the literature, we postulate that MCP-1 would cause cell death via MCPIP induction, ROS production, ER stress and autophagy. Since it is more convenient to test the validity of such molecular mechanisms in cell cultures, we tested this hypothesis in a cardiomyoblast cell line, H9c2 cells. We show that MCP-1, via MCPIP expression, induces ROS/RNS production that causes ER stress which leads to autophagy and apoptosis in H9c2 cells. In the present study we show that MCPIP-induced ER stress leads to the death of H9c2 cells through caspase 2/12 and the IRE1α (inositol-requiring enzyme 1α)–JNK (c-Jun N-terminal kinase)/p38–p53–PUMA (p53 up-regulated modulator of apoptosis) pathway. These results provide the first molecular insight into the probable mechanism by which elevated levels of MCP-1 associated with chronic inflammation may cause cardiomyocyte death and thus contribute to the development of heart failure.

MATERIALS AND METHODS

Cell culture, adenoviral infection and measurement of cell death

H9c2 rat cardiomyoblasts (A.T.C.C.) were grown in DMEM (Dulbecco's modified Eagle's medium) supplemented with 2% (v/v) FBS (fetal bovine serum), 1% penicillin and 1% streptomycin, and infected with either MCPIP–GFP- (green fluorescent protein) or GFP-expressing adenovirus (see Supplementary Materials and methods at http://www.BiochemJ.org/bj/426/bj4260043add.htm) at an MOI (multiplicity of infection) of 20. After 6 h of infection, cells were grown in DMEM supplemented with 10% (v/v) FBS, 1% penicillin and 1% streptomycin. Cells were treated with appropriate inhibitors 3 h prior to adenovirus transfection (Supplementary Table S1 at http://www.BiochemJ.org/bj/426/bj4260043add.htm). Cell viability and death were measured by MTT [3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyl-2H-tetrazolium bromide], Trypan Blue and TUNEL (terminal deoxynucleotidyltransferase-mediated dUTP nick-end labelling) assays using standard procedures.

ROS measurements

H9c2 cells were infected with MCPIP–GFP- or GFP-expressing adenovirus for 24 h and, after different periods, ROS were measured with 1 μmol/l DHR123 (dihydrorhodamine 123) for 30 min at 37 °C and 5% CO2, and were evaluated with a fluorimetric plate reader (excitation wavelength 550 nm and emission wavelength 590 nm).

RT (reverse transcription)–PCR

Total RNA was isolated with the RNAeasy kit (Invitrogen) from H9c2 cells expressing MCPIP–GFP or GFP alone, and first-strand cDNA was synthesized using 1 μg of total RNA (DNase-treated) using the I Script cDNA synthesis kit (Bio-Rad Laboratories); GAPDH (glyceraldehyde-3-phosphate dehydrogenase) served as an internal control. Primers designed for RT–PCR are listed in Supplementary Table S2 (at http://www.BiochemJ.org/bj/426/bj4260043add.htm).

In vivo measurement of caspase 3 activity

H9c2 cells (5×105 cells) were plated on to six-well plates for 24 h and were infected with MCPIP–GFP- or GFP-expressing adenovirus as indicated above. After 72 h, cells were treated with 1 μl of Red-DEVD-FMK (Red-Asp-Glu-Val-Asp-fluoromethylketone) and incubated for 1 h at 37 °C at 5% CO2. Cells were washed twice, and were then assessed on a fluorimetric plate reader (excitation wavelength 540 nm and emission wavelength 570 nm).

Immunoblot analysis

H9c2 cells were treated with cell lysis buffer [20% glycerol, 0.1% Triton X-100, 8% 0.5 M EDTA and 1% 1 M DTT (dithiothreitol)] and protein samples were collected and subjected to immunoblot analysis using the polyclonal antibodies listed in Supplementary Table S3 (at http://www.BiochemJ.org/bj/426/bj4260043add.htm). Immunoblots were quantified as a ratio over GAPDH.

siRNA (small interfering RNA) treatment

H9c2 cells were transfected with 100 nmol/l of a chemically synthesized siRNA targeted for the appropriate gene (Ambion) or with 100 nmol/l non-specific siRNA (Ambion) using DharmaFECT (Dharmacon) for 24 h prior to infection with MCPIP–GFP or GFP-expressing adenovirus.

Statistical analysis

The experimental data were analysed using SPSS statistical software under Windows XP. All values are presented as means±S.E.M. Results were compared between groups by ANOVA analysis followed by Student's t tests. Differences were considered significant at a P value of < 0.05.

RESULTS

MCP-1 treatment of H9c2 cells caused cell death via MCPIP induction

Since elevated MCP-1 expression is associated with ischaemic cardiomyopathy which is known to involve death of cardiomyocytes, we tested whether MCP-1 treatment of H9c2 cells can cause cell death. MCP-1 treatment of H9c2 cells caused cell death (Figure 1A). To determine whether the death caused by MCP-1 treatment was mediated via MCPIP, cells were treated with MCP-1 and siRNA specific for MCPIP. We found that MCP-induced cell death in H9c2 cells was attenuated by siRNA specific for MCPIP (Figure 1A), whereas non-specific siRNA showed no effect. Adenovirus-mediated expression of MCPIP–GFP or GFP was observed by the appearance of fluorescence after infection of the H9c2 cells. The expression level of the adenovirus-mediated MCPIP–GFP was maximal at 1 day after infection and subsequently decreased (Figure 1B). Immunoblot analyses revealed the production of the expected fusion protein (Figure 1C). Adenoviral expression of MCPIP–GFP in H9c2 decreased viability as measured by the MTT assay (Figure 1D) and caused cell death as measured by Trypan Blue assay (Figure 1E). Since MCPIP-expressing cardiomyocytes were found to undergo apoptosis in transgenic mice with cardiomyocyte-targeted expression of MCP-1 [9], we tested whether MCPIP expression induced apoptosis in H9c2 cells. Expression of MCPIP–GFP in H9c2 cells resulted in TUNEL positivity from 1 to 3 days after infection, with maximal positivity occurring at 3 days (Figure 1F). These results indicate that MCP-1 caused cell death via MCPIP.

Figure 1 MCP-1 induces H9c2 cell death that is mediated via MCPIP

(A) Cells were treated with 100 nmol/l siRNA specific for MCPIP, or 100 nmol/l non-specific siRNA 12 h prior to treatment with 10 ng/ml MCP-1. Cells were evaluated 24 h post-MCP-1 treatment for cell death using Trypan Blue (P<0.01). (B) Cells were infected with 20 MOI of MCPIP–GFP- or GFP-alone-containing adenovirus. Fluorescence was monitored using microscopy and the percentage of cells positive for expression was recorded from 1 to 5 days. (C) H9c2 cells were infected with adenovirus expressing MCPIP–GFP or GFP alone, and 50 μg of cell lysate was subjected to immunoblot analysis with an antibody against MCPIP. The molecular mass in kDa is indicated on the right-hand side. (DF) H9c2 cells expressing MCPIP–GFP or GFP alone were examined for cell viability using the MTT assay (D) or for cell death using Trypan Blue (E) (P<0.02) and TMR TUNEL (F).

MCPIP induces the production of ROS

Since ROS is a major signalling component often associated with cell death, we tested whether MCPIP expression caused ROS production. Immunoblot analysis showed that MCPIP-expressing cells had increased iNOS (inducible NO synthase) protein levels beginning on day 1 (Figures 2A and 2B). An increase in the NADPH oxidase phox47 subunit protein levels was observed (Figure 2C). Activation of phox47 occurred 1 day after MCPIP-adenoviral infection, as indicated by translocation of phox47 from the cytoplasm into the plasma membrane (Figure 2D). MCPIP-induced ROS production was demonstrated with the fluorescent dye DHR123 (Figure 2E). Inhibition of iNOS and NADPH oxidase was able to attenuate ROS production (Figures 2E–2G). These results indicate that MCPIP expression induces iNOS and NADPH oxidase, which leads to ROS production.

Figure 2 MCPIP induces ROS/RNS production

(A) MCPIP–GFP or GFP alone were expressed in the H9c2 cell line using adenovirus-mediated gene delivery at 20 MOI. Cell lysate (20 μg) was subjected to immunoblot analysis using an antibody against iNOS. (B) Quantification of immunoblot normalized against GAPDH (P<0.05). (C) Protein (40 μg) was evaluated via immunoblot analysis with an antibody specific for phox47 and quantification of the immunoblot was carried out and normalized against GADPH (P<0.02). (D) Plasma membrane was isolated from total cell lysate, and 20 μg of protein was evaluated via immunoblot for phox47 or the plasma-membrane-specific protein cox4. Quantification of the immunoblot was carried out and normalized to cox4. (EG) Cells were treated with or without (E) iNOS inhibitor 1400W, (F) the NADPH oxidase inhibitor tiron or (G) the free radical scavenger CeO2 3 h prior to infection. Total ROS was detected by treating the MCPIP–GFP- or GFP-expressing cells 48 h post-infection with 10 μmol/l DHR123 for 30 min. Cells were then examined fluorimetrically (excitation wavelength 550 nm and emission wavelength 590 nm; P<0.01).

MCPIP-induced ROS causes an ER stress response

Since MCP-1 has been reported to induce an ER stress response in the heart tissue of mice with cardiomyocyte-targeted expression of MCP-1 [13], we tested whether MCP-1 treatment of H9c2 cells causes ER stress. MCP-1 treatment of H9c2 cells resulted in increased expression of the ER stress markers HSP40 (heat-shock protein 40), PDI (protein disulfide-isomerase) and GRP78 (guanine-nucleotide-releasing protein 78) at both the transcript (Figure 3A) and protein (Figures 3B and 3C) levels. This induction was attenuated by knockdown of MCPIP expression with siRNA, indicating that MCP-1 induced ER stress via MCPIP. These results indicate that MCP-1 causes an ER response via induction of MCPIP. Since IRE1α is known to be an major signalling component involved in ER-stress-mediated events [18,19], we measured the level of this protein. Cells expressing MCPIP–GFP showed significantly elevated levels of this protein (Figure 4A). Moreover, the ER-stress-associated caspase 2 [20] was activated in cells expressing MCPIP (Figures 4B and 4C). The caspase 2-specific inhibitor Z-VDVAD-FMK (benzyoxycarbonyl-Val-Asp-Val-Ala-Asp-fluoromethylketone) inhibited MCPIP-induced activation of caspase 2 in H9c2 cells (Figure 4B).

Figure 3 MCPIP induces an ER stress response

(A) H9c2 cells were treated with siRNA specific for MCPIP or non-specific siRNA (NS) prior to treatment with or without 10 ng/ml MCP-1. After 24 h, cells were evaluated for the ER stress proteins HSP40, PDI and GRP78 using RT–PCR (P<0.05). (B and C) Cell lysate from H9c2 cells with adenovirus-mediated expression of MCPIP–GFP or GFP alone (20 MOI) was evaluated using immunoblot analysis to detect HSP40, PDI and GRP78.

Figure 4 MCPIP induces the ER stress protein IRE1α and induces cleavage of caspase 2

(A) MCPIP–GFP or GFP alone was expressed in the H9c2 cell line using adenovirus-mediated gene delivery at 20 MOI. Cell lysate (20 μg of protein) was subjected to immunoblot analysis using an antibody against IRE1α. (B and C) H9c2 cells were treated with or without 10 μmol/l of the caspase 2 inhibitor Z-VDVAD-FMK (Kamiya Biomedical) 1 h prior to infection with GFP- or MCPIP–GFP-expressing adenovirus. Medium was changed daily with or without Z-VDVAD-FMK for a period of 3 days. After 3 days, cells were collected and evaluated for caspase 2 activity via immunoblot analysis of caspase 2-cleaved product. The immunoblot results were quantified against GAPDH.

We next tested whether MCPIP-induced ROS production is required for the induction of ER stress in H9c2 cells. We found that inhibition of iNOS with 1400W, and inhibition of ROS production with tiron resulted in a significant decrease in the expression levels of HSP40, PDI and GRP78 in MCPIP-expressing cells (Figures 5A and 5B). Furthermore, inhibition of MCPIP-induced ROS/RNS production with CeO2 also resulted in a decrease in expression levels of HSP40, PDI and GRP78 (Figure 5C). Interestingly, inhibition of ER stress with the ER-stress-specific inhibitor TUDC (tauroursodeoxycholate) also resulted in reduced ROS production (Figure 5D). Moreover, the ER stress protein ERO1 increased in MCPIP-expressing cells (Figure 6A). Furthermore, treatment of these cells with siRNA for ERO1, a critical player in ER stress, resulted in decreased ROS production (Figure 6B).

Figure 5 Inhibition of ROS production attenuates MCPIP-induced ER stress and inhibition of ER stress attenuates ROS production

(A and B) H9c2 cells were treated with or without 1400W or tiron 1 h prior to infection with adenovirus expressing MCPIP–GFP or GFP alone, and real time RT–PCR was used to assess levels of the ER stress proteins HSP40, PDI and GRP78 2 days post-infection (*P<0.03). (C) H9c2 cells were treated with or without CeO2 3 h prior to infection with adenovirus expressing MCPIP–GFP or GFP alone, and real-time RT–PCR was used to assess levels of the ER stress proteins HSP40, PDI and GRP78 2 days post-infection. (D) Cells were also treated with TUDC 1 h prior to adenovirus infection and were evaluated for ROS production using 10 μmol/l DHR123, incubated for 30 min and then examined fluorimetrically (excitation wavelength 550 nm and emission wavelength 590 nm; P<0.01).

Figure 6 MCPIP induces the ER protein ERO1

(A) H9c2 cells were infected with adenovirus expressing MCPIP–GFP or GFP alone. Cell lysate prepared 48 h after infection (20 μg of protein) was evaluated using immunoblot to detect ERO1. (B) H9c2 cells were treated with siRNA specific for ERO1 prior to adenovirus infection, and knockdown was confirmed using immunoblot analysis with an antibody against ERO1 (top panels). ROS were evaluated fluorimetrically using DHR123 (bottom panel; excitation wavelength 550 nm and emission wavelength 590 nm; P<0.01).

Expression of MCPIP in H9c2 cells induces autophagy

Since ER stress is known to lead to autophagy and autophagy has been reported to occur in cardiomyocytes involved in cardiovascular disease [17,21], we tested whether MCPIP-induced death in H9c2 cells involves increased formation of autophagolysomes characteristic of autophagy. Cells with MCPIP expression showed higher amounts of autophagosomes than did the GFP control (Figure 7A). Knockdown of beclin-1 with siRNA attenuated the accumulation of MCPIP-induced autophagolysomes (Figure 7A). We found that beclin-1, an autophagy marker, significantly increased 2 days after adenovirus infection (Figures 7B and 7C). LC3 (microtubule-associated protein 1 light chain 3) cleavage to LC3 II was also induced by MCPIP (Figures 7D and 7E). To test whether the autophagy induced by MCPIP was mediated via ER stress, we tested whether the ER stress inhibitor TUDC, would inhibit autophagy. Induction of beclin-1 by MCPIP was inhibited by TUDC and by knockdown of IRE1 with siRNA (Figure 7F).

Figure 7 MCPIP induces autophagy via ER stress

H9c2 cells were infected with adenovirus expressing GFP or MCPIP–GFP and were evaluated 2 days later. (A) At 48 h after infection, cells were stained with monodansylcadaverine (MDC) to detect autophagolysosome formation (white arrows) and with DAPI (4′6-diamidino-2-phenylindole; black arrowheads) to stain the nucleus. Cells were then assessed microscopically. (B) Cell lysates (50 μg of protein) were subjected to immunoblot analysis with antibody against beclin-1. (C) Quantification of immunoblot data in (B) normalized against GAPDH (P<0.01). (D) Protein (20 μg) was evaluated via immunoblot analysis to detect cleavage of LC3 I into LC3 II. (E) Quantification of immunoblot data in (D) normalized against GAPDH (P<0.02). (F) Cells were treated with 100 nmol/l siRNA of IRE1 or beclin-1 prior to adenovirus infection. At 48 h after adenovirus treatment, cell lysate was collected. Cell lysates (50 μg) were subjected to immunoblot analysis with an antibody against beclin-1 (P<0.05).

Expression of MCPIP in H9c2 cells induced apoptotic cell death via ROS, ER stress and autophagy

Next we tested whether cell death resulting from MCPIP expression is mediated via ROS-induced ER stress that leads to autophagy. Inhibition of oxidative stress with either the iNOS-specific inhibitor 1400 W or CeO2 resulted in the attenuation of MCPIP-induced cell death (Figure 8A). Specific inhibitors of ER stress, 4-PBA (4-phenylbutyric acid) and TUDC [22] (Figures 8B and 8C) and knockdown of IRE1 with siRNA (Figures 9A and 9B) attenuated MCPIP-induced cell death. We tested whether a specific inhibitor of ER-stress-associated caspases 2 and 12 would inhibit MCPIP-induced cell death. In fact, Z-VDVAD-FMK and Z-ATD-FMK (benzyoxycarbonyl-Ala-Thr-Asp-fluoromethylketone) attenuated MCPIP-induced cell death in H9c2 cells (Figures 8D and 8E). Moreover, 3′MA (3-methyladenine) (Figure 8F), which is known to inhibit autophagy, and knockdown of beclin-1 (Figures 9A and 9B), also resulted in the attenuation of MCPIP-induced cell death. In support of this conclusion, treatment of MCPIP–GFP-expressing H9c2 cells with 1400 W, TUDC or 3′MA showed marked inhibition of apoptosis as detected by the TMR TUNEL assay (Figures 9C and 9D). Furthermore, treatment of MCPIP–GFP-expressing H9c2 cells with 1400 W, TUDC or 3′MA resulted in decreased caspase 3 activity (Figure 9C). We conclude that MCPIP induces ROS that results in ER stress which leads to autophagy and eventual apoptosis.

Figure 8 MCPIP-induced cell death occurs via ROS induction of ER stress that leads to autophagy

H9c2 cells were infected with adenovirus expressing GFP or MCPIP–GFP. (A) Cells were treated with or without the iNOS inhibitor 1400W, or with or without CeO2 3 h prior to infection. Cell death was evaluated using Trypan Blue at 72 h after infection (P<0.02) with or without treatment with 1400W. Cells were treated with 4-PBA (B) (P<0.01), TUDC (C) (P<0.02) or Z-ATD-FMK (D) (P<0.01) 1 h prior to infection with MCPIP–GFP or GFP-expressing adenovirus (20 MOI). Samples were examined for cell death 3 days post-infection using Trypan Blue. (E) Cells were treated with 50 μM 3′MA 1 h prior to adenovirus-mediated expression of MCPIP–GFP or GFP. Cells were evaluated 1–3 days post-infection for cell death using Trypan Blue (P<0.03).

Figure 9 MCPIP-induced cell death involves caspase 3 activation and apoptosis

(A) Cells were treated with or without 1400W, TUDC or 3′MA 1 h prior to infection at 20 MOI with MCPIP–GFP- or GFP-expressing adenovirus, and were evaluated for apoptosis using TMR TUNEL. (B) Left-hand panel: the results in (A) were quantified (P<0.04). Right-hand panel: caspase 3 was monitored in cells using Red-DEVD-fmk (P<0.02). (C) Cells were treated with or without siRNA for IRE1 or beclin-1 1 h prior to infection at 20 MOI with MCPIP–GFP- or GFP-expressing adenovirus, and were evaluated for apoptosis using TMR TUNEL. (D) Left-hand panel: the results in (C) were quantified (P<0.04). Right-hand panel: caspase 3 was monitored in cells using Red-DEVD-fmk (P<0.03). Non-specific (NS) siRNA served as a control.

MCPIP expression causes activation of JNK and p38, and induction of p53 and PUMA

Microarray analysis performed with RNA isolated 3 days after adenoviral infection revealed that MCPIP induced many of the genes that are known to be important in signalling cell death (Supplementary Table S4 at http://www.BiochemJ.org/bj/426/bj4260043add.htm). Beclin-1 and p53 showed the highest fold increases in expression. Caspase 3 increased 2.3-fold and caspase 9 increased 5.2-fold. The ER-associated caspases 4 and 12 also increased.

Because IRE1α signals JNK activation [18,19], we examined its role in MCPIP-induced death. We also examined p38, which is known to be induced by ER stress [24]. Inhibition of both JNK and p38 attenuated MCPIP-induced cell death (Supplementary Figure 1A at http://www.BiochemJ.org/bj/426/bj4260043add.htm). Furthermore, levels of phosphorylated c-Jun increased in cells expressing MCPIP–GFP as compared with GFP controls, whereas overall levels of c-Jun remained unchanged (Supplementary Figure 1B). Moreover, the downstream proteins p53 and PUMA, regulated by JNK and p38 activation [25], increased in MCPIP-expressing cells compared with GFP controls (Supplementary Figure 1C). Thus MCPIP-induced ER stress probably signals cell death through JNK/p38 activation, and consequent induction of p53 and PUMA.

DISCUSSION

If and how MCP-1 plays a role in the death of cardiomyocytes, which is known to be involved in heart failure, are unknown. In addition, the signalling events resulting from the MCP-1–CCR2 interaction might lead to gene expression changes that may play a critical role in the development of cardiovascular diseases. The nature and the role of such gene expression changes in the pathophysiological changes associated with heart disease are poorly understood. We previously reported that MCP-1 induces MCPIP, a novel zinc-finger protein that has transcription factor-like activity [9]. In mice with cardiac-targeted expression of MCP-1, MCPIP expression was elevated and was associated with cardiomyocyte apoptosis and heart failure [9]. If and how MCPIP might cause cardiomyocyte death is not known. We postulate that MCPIP induces oxidative and nitrosative stress that would cause ER stress, which would lead to autophagy and cell death, and provide experimental evidence to support this hypothesis in the present study.

Previous studies have demonstrated the involvement of ROS and RNS in myocardial infarction, cardiac hypertrophy, cardiomyopathy and heart failure in animals and human patients [2628]. Moreover, it was found that ROS and RNS production in ischaemia/reperfusion was attenuated in Ccr2−/− mice [28a]. In the present study we show that MCPIP induces ROS production in the cardiomyoblast cell line H9c2 via induction and activation of phox47. This idea was further supported by the finding that inhibition of iNOS with 1400W and NADPH oxidase with tiron attenuated ROS production and cell death. ROS and RNS production is known to cause ER stress [11,12,29,30]. It was recently reported that hypoxia in cultured myocytes and myocardial infarction in intact animals caused ER stress in cardiomyocytes [30]. ER stress results from the accumulation of misfolded proteins which lead to the induction of the UPR (unfolded protein response) [31]. The present results demonstrate that MCP-1 treatment of H9c2 cells results in increased expression of the ER stress proteins ERO1 (ER oxidoreductin 1), HSP40, PDI and GRP78 at both transcript and protein levels. That induction of ER stress by MCP-1 was mediated via MCPIP was shown by the finding that knockdown of MCPIP inhibited MCP-1-induced ER stress. Furthermore, expression of MCPIP in H9c2 cells induced ER stress chaperones. We showed that MCPIP-induced ER stress was mediated via ROS and RNS production. Treatment with CeO2 nanoparticles attenuated oxidative stress, ER stress and cell death. Furthermore, use of the previously reported specific ER stress inhibitors [22] also resulted in decreased ROS production in cells expressing MCPIP. This result can be explained by the induction of ERO1 which is known to generate ROS production by the reduction of the disulfide bonds in PDI [19,32]. In fact, knockdown of ERO1 with siRNA resulted in a decrease of ROS production. Thus MCPIP induces ROS production which causes ER stress that leads to further amplication of ROS production through ERO1. In addition, inhibition of ER stress by specific inhibition or by knockdown of IRE1 resulted in the attenuation of MCPIP-induced cell death. Thus our results indicate that MCPIP induces ROS production which causes ER stress that leads to cell death.

Even though the critical role of MCP-1 in the development of cardiovascular diseases is well recognized, there are observations that suggest a protective role for MCP-1. Our finding that MCP-1 induces the expression of ER stress chaperones via MCPIP can reconcile these seemingly contradictory results. In isolated neonatal mouse cardiac myocytes, treatment with MCP-1 protected the cells from death caused by subsequent 18 h of hypoxia [33]. This protection could arise from MCP-1-induced ER stress chaperones. Many reports have demonstrated that expression of ER stress chaperones have cardioprotective effects. For example, protection by ischaemic preconditioning by the myocardial expression of GRP78 and protection from ischaemia/reperfusion injury by cardiac expression of ER stress chaperone have been demonstrated [34,35]. It has been demonstrated that expression of ER stress chaperone GRP78 prior to ischaemic injury protects cells [36], whereas sustained and prolonged stress would result in the breakdown of this protection. Cardiac-targeted expression of MCP-1 in transgenic mice shows cardioprotective effects against myocardial infarction and ischaemia/reperfusion damage in young animals [3739]. In such animals, ER stress chaperones have been shown to be elevated [13] and these elevated chaperone levels probably account for the observed protection. However, as the animals age, the sustained myocardial expression of MCP-1 would lead to prolonged oxidative and nitrosative stress under which protection would break down with a decrease in ER stress chaperones as was found previously [13]. This loss of protection would cause autophagy and cell death leading to the development of heart failure and death of the animal at approx. 6 months of age.

ER stress is known to cause cell death through multiple signalling pathways [19,24,40]. Caspase 2 activation has been reported to be involved in oxidative-stress-induced apoptosis [41]. Caspase 2 and caspase 12 are ER-related caspases that have been shown to be activated upon prolonged ER stress resulting in cell death [20,41a]. Caspase 2 has been recently shown to cleave and activate Bid, a BH3 pro-apoptotic protein that activates the mitochondrial apoptotic proteins Bax and Bak [20]. Our microarray analysis showed that MCPIP expression induced production of caspase 12, and immunoblot analysis showed that caspase 2 was cleaved into its active form in MCPIP-expressing H9c2 cells. In addition, specific inhibition of caspase 12 or caspase 2 resulted in attenuation of MCPIP-induced cell death. Expression of MCPIP in cardiomyoblasts resulted in the induction of IRE1 protein which is a well-known signalling molecule involved in ER-stress-induced cell death [18,19]. IRE1 can induce autophagy and apoptosis [18,40]. ER stress is known to cause autophagy [14,15]. Autophagy is known to be involved in the death of cardiomyocytes that leads to heart failure [16,17,21]. Beclin-1 is the key protein involved in autophagy and can interact with the anti-apoptotic protein Bcl-2, and its binding state may play a key part in co-ordinating the cellular decision to undergo autophagy [42]. Activation of autophagy through beclin-1 results in the accumulation of autophagosomes, an event that requires the cleavage of the autophagy protein LC3 [40]. The results of the present study showed that expression of MCPIP in cardiomyoblasts resulted in increased expression of beclin-1 protein levels and cleavage of LC3. The finding that inhibition of autophagy by 3′MA and knockdown of beclin-1 with siRNA resulted in attenuation of MCPIP-induced death strongly implicates autophagy in cell death. This conclusion was supported by our results showing that inhibition of MCPIP-induced ER stress with TUDC and by knockdown of IRE1 with siRNA resulted in decreased expression levels of beclin-1 and cell death.

The precise role of autophagy in cell death is unclear. Some reports suggest that autophagy attempts to protect the cell from death by degrading misfolded proteins [43,44]. Autophagy is also known to cause cell death under the influence of developmental or stress signals [45]. It is known that signalling components of apoptosis can, in fact, induce autophagy [46]. Apoptosis has been associated with heart failure [47]. Our results show that MCPIP expression in cardiomyoblasts resulted in apoptosis, as detected by TUNEL and caspase 3 cleavage. Moreover, many of the genes that were induced by MCPIP, as revealed by our microarray analysis, can be either apoptotic or related to autophagy. Inhibition of ER stress with TUDC resulted in decreased caspase 3 activity, demonstrating that MCPIP-induced ER stress does cause apoptosis. Signalling components previously shown to be induced by the ER stress protein IRE1 were also found to be activated or induced in MCPIP-expressing cells. These included phosphorylation of c-Jun and increased expression levels of p53 and PUMA. Moreover, a role for caspase 2 in p53-mediated apoptosis has recently been reported [48].

It is likely that the processes elucidated in the present study in H9c2 cells occur in vivo. In the transgenic mice with cardiomyocyte-targeted expression of MCP-1, the failing heart tissue showed ROS production, as indicated by staining of heart tissue with chloromethyl dichlorofluorescein diacetate, and RNS production, as indicated by staining for nitrotyrosine [10]. These hearts showed elevated expression of ER stress chaperones prior to the development of significant left ventricular dysfunction [13]. Inhibition of oxidative stress by administration of CeO2 nanoparticles to the transgenic mice inhibited ER stress development, cardiomyocyte apoptosis and development of heart disease [8]. Thus the molecular mechanisms underlying the MCP-1-induced death in H9c2 cells elucidated in the present study are probably involved in intact animals and form the molecular basis of how the elevated levels of MCP-1 associated with chronic inflammation contributes to the development of ischaemic heart failure by causing the death of cardiomyocytes.

AUTHOR CONTRIBUTION

Craig Younce performed the experiments and collected the data, prepared the Figures, helped with experimental design and interpretation of the data and wrote the initial draft of the manuscript. Pappachan Kolattukudy conceptualizatized the project, designed the experiment, interpreted the data and revised the manuscript.

FUNDING

This work was supported by the National Institutes of Health [grant number HL-69458].

Abbreviations: DHR123, dihydrorhodamine 123; DMEM, Dulbecco's modified Eagle's medium; ER, endoplasmic reticulum; ERO1, ER oxidoreductin 1; FBS, fetal bovine serum; GAPDH, glyceraldehyde-3-phosphate dehydrogenase; GFP, green fluorescent protein; GRP78, guanine-nucleotide-releasing protein 78; HSP40, heat-shock protein 40; iNOS, inducible NO synthase; IRE1α, inositol-requiring enzyme 1α; JNK, c-Jun N-terminal kinase; LC3, microtubule-associated protein 1 light chain 3; 3′MA, 3-methyladenine; MCP-1, monocyte chemotactic protein-1; MCPIP, MCP-1-induced protein; MOI, multiplicity of infection; MTT, 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyl-2H-tetrazolium bromide; 4-PBA, 4-phenylbutyric acid; PDI, protein disulfide-isomerase; PUMA, p53 up-regulated modulator of apoptosis; Red-DEVD-FMK, Red-Asp-Glu-Val-Asp-fluoromethylketone; ROS, reactive oxygen species; RNS, reactive nitrogen species; RT, reverse transcription; siRNA, small interfering RNA; TUDC, tauroursodeoxycholate; TUNEL, terminal deoxynucleotidyltransferase-mediated dUTP nick-end labelling; Z-ATD-FMK, benzoxycarbonyl-Ala-Thr-Asp-fluoromethylketone; Z-VDVAD-FMK, benzoxycarbonyl-Val-Asp-Val-Ala-Asp-fluoromethylketone

References

  1. 1.
  2. 2.
  3. 3.
  4. 4.
  5. 5.
  6. 6.
  7. 7.
  8. 8.
  9. 9.
  10. 10.
  11. 11.
  12. 12.
  13. 13.
  14. 14.
  15. 15.
  16. 16.
  17. 17.
  18. 18.
  19. 19.
  20. 20.
  21. 21.
  22. 22.
  23. 23.
  24. 24.
  25. 25.
  26. 26.
  27. 27.
  28. 28.
  29. 28a.
  30. 29.
  31. 30.
  32. 31.
  33. 32.
  34. 33.
  35. 34.
  36. 35.
  37. 36.
  38. 37.
  39. 38.
  40. 39.
  41. 40.
  42. 41.
  43. 41a.
  44. 42.
  45. 43.
  46. 44.
  47. 45.
  48. 46.
  49. 47.
  50. 48.
View Abstract