Oxidative stress and adenosine A1 receptor activation differentially modulate subcellular cardiomyocyte MAPKs

Cherry Ballard-Croft, Adam C. Locklar, Byron J. Keith, Robert M. Mentzer Jr, Robert D. Lasley

Abstract

The mechanism by which distinct stimuli activate the same mitogen-activated protein kinases (MAPKs) is unclear. We examined compartmentalized MAPK signaling and altered redox state as possible mechanisms. Adult rat cardiomyocytes were exposed to the adenosine A1 receptor agonist 2-chloro-N6-cyclopentyladenosine (CCPA; 500 nM) or H2O2 (100 μM) for 15 min. Nuclear/myofilament, cytosolic, Triton-soluble membrane, and Triton-insoluble membrane fractions were generated. CCPA and H2O2 activated p38 MAPK and p44/p42 ERKs in cytosolic fractions. In Triton-soluble membrane fractions, H2O2 activated p38 MAPK and p42 ERK, whereas CCPA had no effect on MAPK activation in this fraction. The greatest difference between H2O2 and CCPA was in the Triton-insoluble membrane fraction, where H2O2 increased p38 and p42 activation and CCPA reduced MAPK activation. CCPA also increased protein phosphatase 2A activity in the Triton-insoluble membrane fraction, suggesting that the activation of this phosphatase may mediate CCPA effects in this fraction. The Triton-insoluble membrane fraction was enriched in the caveolae marker caveolin-3, and >85% of p38 MAPK and p42 ERK was bound to this scaffolding protein in these membranes, suggesting that caveolae may play a role in the divergence of MAPK signals from different stimuli. The antioxidant N-2-mercaptopropionyl glycine (300 μM) reduced H2O2-mediated MAPK activation but failed to attenuate CCPA-induced MAPK activation. H2O2 but not CCPA increased reactive oxygen species (ROS). Thus the adenosine A1 receptor and oxidative stress differentially modulate subcellular MAPKs, with the main site of divergence being the Triton-insoluble membrane fraction. However, the adenosine A1 receptor-mediated MAPK activation does not involve ROS formation.

  • signal transduction
  • compartmentation

mitogen-activated protein kinases (MAPKs) are ubiquitous signaling proteins that play an important role in the response of a cell to extracellular signals. The three MAPK subtypes, extracellular signal-regulated kinases (ERKs), p38 MAPK, and c-Jun NH2-terminal kinases (JNKs), are activated by a wide variety of stimuli such as hyperosmotic shock, inflammation, growth factors, reactive oxygen species (ROS), and G protein-coupled receptors (GPCRs) (29). In the heart, ROS and GPCRs modulate the cardiomyocyte response to ischemia-reperfusion through MAPK activation. Excessive ROS formation is deleterious, whereas stimulation of the GPCR, adenosine A1, is cardioprotective (20, 39). The underlying mechanism by which these diverse stimuli activate the same MAPKs to yield distinct physiological effects is unclear. The elucidation of the mechanism responsible for this stimulus-specific MAPK signaling may provide a better understanding of how MAPKs could be involved in both injurious and cardioprotective processes in the heart (3, 29, 46).

Stimulus-specific MAPK signaling may occur through compartmentation since MAPK stimulation in different subcellular compartments may lead to the activation of distinct downstream targets resulting in different physiological effects (28). Stimulus-dependent differences in cytosolic and nuclear ERK activation have been reported in noncardiac cells (49, 51), but the possibility that distinct stimuli may also differentially modulate cardiomyocyte ERK and p38 MAPK in these compartments has not been investigated. Furthermore, the differential activation of MAPKs in cardiomyocyte sarcolemmal membranes by different stimuli such as ROS and adenosine A1 receptor agonists has not been explored.

Although some downstream targets of MAPKs are localized in plasma membranes (9, 29), MAPK signaling in cardiomyocyte membrane compartments is poorly understood. Our studies in intact myocardium show that an adenosine receptor agonist modulates p38 MAPK and ERK in membrane fractions and that ischemia-reperfusion has a different effect on the MAPKs in this fraction (5, 43). Thus the sarcolemmal membrane may indeed be a site for divergence of MAPK signals from various stimuli. Sarcolemmal membranes can be further subdivided into Triton-soluble and -insoluble components. The Triton-insoluble membrane fraction contains the scaffolding protein caveolin-3, which is found in cardiomyocyte caveolae (2, 32). Caveolae are thought to play an important role in signal transduction by providing the scaffolding for various signaling molecules to interact (2, 24a, 34, 52). Moreover, we have recently reported that ischemia-reperfusion activates ERK and p38 MAPK in myocardial caveolin-3-enriched fractions (6). The stimulus-specific modulation of these MAPKs may occur in cardiomyocyte Triton-insoluble membrane fractions, which are also enriched in caveolin-3, but this remains to be determined.

Another mechanism by which stimulus-specific MAPK signaling may occur is through transient ROS formation. Whereas excessive ROS is injurious to the heart, transient ROS formation has been shown by several groups to be cardioprotective (11, 35, 38, 54). There are numerous reports that myocardial GPCR activation is associated with ROS formation (1, 20, 35, 36), suggesting that increased oxidative stress may mediate MAPK activation in GPCR-stimulated cells. Studies showing that GPCR-mediated cardioprotection is blocked with the antioxidant N-2-mercaptopropionyl glycine (MPG) provide additional indirect evidence supporting this idea (11, 36, 38). In contrast, A1 adenosine receptor activation has been shown to reduce the formation of ROS and increase the activities of antioxidant enzymes (18, 31). In the present study, the effects of adenosine A1 receptor stimulation and H2O2 on the compartmentation of MAPK signaling and ROS formation were examined in adult rat ventricular myocytes. The results from this study show that stimulus-specific differences in MAPK signaling do occur in cardiomyocyte membrane fractions, but ROS formation does not mediate adenosine A1 receptor-mediated MAPK activation.

MATERIALS AND METHODS

All animals in this study received humane care according to guidelines in The Principles of Laboratory Animal Care formulated by the National Society for Medical Research and the National Institutes of Health (NIH) Guide for the Care and Use of Laboratory Animals (NIH Publication No. 86-23, Revised 1996). The protocols used to handle the animals were approved by the University of Kentucky Institutional Animal Care and Use Committee.

Isolation of adult rat ventricular myocytes.

Ventricular myocytes were enzymatically dissociated from adult male Sprague-Dawley rats (300–325 g; Charles River, Portage, MI), as described previously (21). The cardiomyocytes were suspended in an experimental buffer containing 140 mM NaCl, 4.8 mM KCl, 1.2 mM KH2PO4, 1.2 mM MgSO4, 11.0 mM glucose, 1 mM CaCl2, and 10 mM HEPES (pH 7.4). This protocol yielded >75% viable, rod-shaped cardiomyocytes.

Myocyte drug treatments.

The cardiomyocytes were exposed to an experimental buffer (control; n = 5 hearts), 100 μM H2O2 (n = 5), or 500 nM 2-chloro-N6-cyclopentyladenosine (CCPA; n = 5 hearts), the A1 adenosine receptor agonist, for 15 min at 37°C. Previous studies have shown that these concentrations of H2O2 and CCPA have a maximal effect on preconditioning and MAPK activation (22, 31, 44, 50). The 15-min time point was chosen for this study because it is the time at which 100 μM H2O2 exposure produces maximal MAPK activation (22, 50). The cardiomyocytes were also exposed to the antioxidant MPG (300 μM) 10 min before treatment with either 100 μM H2O2 (n = 3) or 500 nM CCPA (n = 3) for 15 min at 37°C (11).

Cardiomyocyte subcellular fractionation.

After drug treatment, the myocytes were rapidly resuspended in an ice-cold homogenization buffer containing 250 mM sucrose, 10 mM KCl, 1.5 mM MgSO4, 1 mM EDTA, 1 mM EGTA, 1 mM DTT, 0.1 mg/ml PMSF, 45 μg/μl aprotinin, 0.5 mM β-glycerophosphate, 1 mM sodium vanadate, and 20 mM HEPES (pH 7.4). The sample was homogenized with a dounce homogenizer and sonicated three times in 10-s bursts. The homogenate was then centrifuged for 10 min at 750 g to pellet the nuclei and myofilaments. The resulting supernatant was centrifuged at 100,000 g for 30 min to obtain the cytosol. The pellet from the 100,000-g spin was resuspended in an homogenization buffer containing 1% Triton X-100 and incubated on ice for 30 min. The sample was then centrifuged again for 30 min at 100,000 g. The supernatant was the Triton-soluble membrane fraction, and the pellet was the Triton-insoluble fraction. The total protein in each fraction was determined with a Lowry protein assay (Bio-Rad, Hercules, CA).

Western blot analysis.

Western blot analysis was performed as described previously (5). The protein samples (30 μg) were separated and transferred to nitrocellulose (Bio-Rad). Ponceau S staining was used to verify equal protein loads. The membrane was then blocked in Tris-buffered saline containing 0.1% Tween and 0.2% I-Block (Tropix, Bedford, MA). The phospho-p38 MAPK (Santa Cruz), the phospho-ERK1/2 (Cell Signaling, Beverly, MA), α-actinin (Santa Cruz), or the caveolin-3 antibody (Transduction Laboratories, Lexington, KY) in 5% BSA was then incubated with the membrane. After secondary antibody incubation, the bound antibodies were visualized by enhanced chemiluminescence (Amersham, Piscataway, NJ). The phospho-p38 MAPK and the phospho-ERK1/2 membranes were stripped to reprobe with the p38α antibody (Santa Cruz) and total ERK antibody (Santa Cruz), respectively. Scion Image software (Frederick, MD) was used for densitometric analysis of the protein bands. The results were normalized to a positive control included on each membrane (H2O2-treated cardiomyocyte whole cell lysate).

Measurement of phosphatase activity.

Protein phosphatase 2A (PP2A) activity in Triton-insoluble fractions was measured with a Malachite Green Phosphatase assay kit (Upstate). The sample (100 μl) was rocked with G25 Sephadex for 20 min at 4°C to remove free phosphate. The sample was then centrifuged at 200 g for 5 min, and the resulting supernatant was centrifuged for 30 min at 100,000 g. The phosphate-free samples were diluted to 25 μg with a phosphatase assay buffer (containing 20 mM MOPS, 0.1 M NaCl, 60 mM β-mercaptoethanol, 1 mM MgCl2, 1 mM EGTA, 0.1 mM MnCl2, 1 mM DTT, 10% glycerol, and 0.1 mg/ml serum albumin) and added to the wells of a 96-well plate. After the addition of 5 μl of the phosphopeptide, the samples were incubated for 30 min at room temperature. The reaction was terminated by the addition of 100 μl of Malachite green solution. The color was allowed to develop for 15 min at room temperature, and the free phosphate levels were determined by measuring absorbance at 650 nm. The signal that was inhibited by okadaic acid (10 nM) was considered PP2A activity.

Measurement of oxidative stress in cardiomyocytes.

The extent of oxidative stress in cardiomyocytes was determined with the fluorescent dye 5-(and 6-)chloromethyl-2′,7′-dichlorodihydrofluorescein (CMDCF) diacetate (CMDCF-DA), as previously described (21). The cardiomyocytes were loaded with CMDCF-DA (5 μM; Molecular Probes, Eugene, OR) for 30 min at room temperature in an experimental buffer. The cells were then incubated in an experimental buffer for an additional 30 min to allow for intracellular deacetylation of CMDCF-DA. Baseline CMDCF fluorescence was measured 15 min after the cells were placed in a temperature-controlled chamber (37°C) on the stage of an Olympus inverted microscope (Olympus America, Melville, NY). The experimental buffer (control; n = 11 cells from 2 hearts), 100 μM H2O2 (n = 19 cells from 4 hearts), or 500 nM CCPA (n = 14 cells from 3 hearts) was suffused into the chamber, and CMDCF fluorescence was recorded at 5, 10, and 15 min of treatment. Epifluorescence was collected with a charge-coupled device camera, and the fluorescence intensity analysis was done with custom-made software. Excitation-dependent oxidation of the dye was reduced with a neutral density filter (ND12). The cell fluorescence values were background corrected and expressed relative to baseline fluorescence of each cell.

Statistical analysis.

All data are expressed as means ± SE. For the Western blot data, statistical significance among the groups was determined using a one-way ANOVA, followed by Tukey's post hoc test. A one-way ANOVA followed by Dunnett's post hoc test was used to determine significant differences in oxidative stress between the control and experimental groups. The statistical significance was defined as P < 0.05.

RESULTS

Characterization of cardiomyocyte subcellular fractions.

Since it is well known that scaffolding proteins, which play an important role in signal transduction (2, 19, 34, 47), are Triton-insoluble (2, 32, 53), the membrane fraction was further subdivided into Triton-soluble and -insoluble components. The cardiomyocyte fractions were then analyzed for the distribution of the scaffolding protein caveolin-3 and the cytoskeletal scaffolding protein α-actinin (Fig. 1). Caveolin-3 immunoreactivity was greatest in the Triton-insoluble membrane fraction (55%) with the cytosolic (2%), Triton-soluble membrane (17%) and nuclear/myofilament (26%) fractions containing significantly less caveolin-3 protein (Fig. 1, A and B). This caveolin-3 expression pattern did not change with H2O2 or CCPA treatment (data not shown). The distribution of α-actinin is shown in Fig. 1, C and D. The nuclear/myofilament fractions contained 58% of the total α-actinin expression, with the Triton-insoluble membrane fraction also containing a significant portion of this protein.

Fig. 1.

Distribution of the scaffolding proteins caveolin-3 and α-actinin. Cytosolic (Cyto), Triton-soluble membrane (Sol Memb), Triton-insoluble membrane (Insol Memb), and nuclear/myofilament (Nuc/Myo) fractions from control rats were analyzed for presence of the caveolae marker caveolin-3 and α-actinin by Western blot analysis. A: representative Western blot showing that caveolin-3 was primarily present in the Insol Memb fraction. B: densitometric summary of caveolin-3 Western blots (n = 3 animals). *P < 0.05 vs. insoluble membrane fraction. C: representative Western blot of α-actinin, which was predominantly found in the Nuc/Myo and Insol Memb fractions. D: densitometric summary of pooled α-actinin Western blots (n = 2 animals). *P < 0.05 vs. Triton-insoluble fractions.

We have previously shown that MAPKs are present in caveolin-3-enriched cardiac caveolar fractions (6); therefore, we determined the extent to which the MAPKs interact with this scaffolding protein in the Triton-insoluble fraction. Caveolin-3 was immunoprecipitated from the Triton-insoluble fraction, and the amount of MAPK bound to caveolin-3 was determined by Western blot analysis. The majority of p38α in the Triton-insoluble fraction (87 ± 3%) was bound to caveolin-3, and this amount did not change with H2O2 (91 ± 2%) or CCPA (91 ± 3%) treatment. Likewise, 89 ± 5% of total p42 ERK was also bound to caveolin-3 in the Triton-insoluble membrane fraction, which was not altered by H2O2 (94 ± 2%) or CCPA (94 ± 2%). By contrast, p44 ERK was not bound to caveolin-3 in the Triton-insoluble membrane fraction (data not shown).

Differential p38 MAPK activation in cardiomyocyte subcellular fractions.

The effects of H2O2 and CCPA on subcellular phospho-p38 MAPK levels are shown in Fig. 2. Both H2O2 and CCPA activated p38 MAPK in cytosolic fractions, but to varying magnitudes. Cytosolic p38 MAPK was activated by 55% in the H2O2 group, whereas CCPA caused a significantly greater activation (96%). H2O2 and CCPA also increased p38 MAPK activation in the nuclear/myofilament fraction, but no difference between the treatment groups was observed in this fraction. H2O2 stimulated p38 MAPK by 53% in the Triton-soluble membrane fraction and by 29% in the Triton-insoluble membranes. By contrast, CCPA had no effect on p38 activation in the Triton-soluble fraction and reduced p38 phosphorylation by 43% in the Triton-insoluble fraction.

Fig. 2.

Activation of phosphorylated p38 MAPK in cardiomyocyte subcellular fractions. Activation of p38 MAPK was determined by Western blot analysis using an antibody that recognizes only the active, dually phosphorylated form of this kinase. A: representative Western blot showing expression of the phosphorylated, active form of p38 MAPK (p-p38) in Cyto, Sol Memb, Insol Memb, and Nuc/Myo fractions. C, control; H, H2O2; and CP, 2-chloro-N6-cyclopentyladenosine (CCPA). B: densitometric analysis of antiactive p38 MAPK Western blots in control (n = 4 animals), H2O2 (n = 6 animals), and CCPA-treated (n = 4 animals) groups. The greatest difference between H2O2 and CCPA occurred in the Insol Memb compartment. *P < 0.05 vs. control and +P < 0.05 vs. H2O2.

p38α Localization in cardiomyocyte subcellular fractions.

The analysis of the p38α protein distribution indicated that the majority of this protein was located in the cytosolic fraction (44%; Fig. 3). The Triton-insoluble fraction had the least p38α immunoreactivity (11%), with the nuclear/myofilament and Triton-soluble membrane fractions possessing similar amounts (25% and 20%). In the CCPA group, there was a significant 27% reduction in the p38α protein in the Triton-insoluble fraction, whereas there was a corresponding tendency for this protein to be increased in the Triton-soluble fraction. These results may explain the diminished p38 MAPK activation observed in the Triton-insoluble fraction.

Fig. 3.

Subcellular localization of p38α. A: representative Western blot showing p38α expression in Cyto, Sol Memb, Insol Memb, and Nuc/Myo fractions. B: densitometric summary of pooled p38α Western blots from control (n = 4 animals), H2O2 (n = 5 animals), and CCPA-treated (n = 3 animals) samples. *P < 0.05 vs. control and +P < 0.05 vs. H2O2.

Differential p42 ERK activation in cardiomyocyte subcellular fractions.

The activation of p44/p42 ERK was next investigated (Fig. 4). Both H2O2 and CCPA activated p44 ERK only in cytosolic fractions, with no significant difference between these stimuli (Fig. 4, A and B). In contrast, p42 ERK activation exhibited stimulus-dependent differences. H2O2 activated p42 ERK in all fractions, whereas CCPA activated p42 only in the cytosolic fractions (Fig. 4, A and C). The greatest difference between H2O2 and CCPA occurred in the Triton-insoluble fraction, where H2O2 increased p42 activation by 2.4-fold, whereas CCPA significantly reduced p42 ERK phosphorylation by 61%. H2O2 also activated p42 ERK by 62% in the Triton-soluble fraction with no effect of CCPA in this fraction.

Fig. 4.

Activation of p44/p42 ERKs in cardiomyocyte intracellular fractions. Activation of p44 and p42 ERK in Cyto, Sol Memb, Insol Memb, and Nuc/Myo fractions was assessed via Western blots using an antibody that recognizes the dually phosphorylated form of ERK. A: representative antiactive ERK Western blot showing p44/p42 activation in C, H, and CP groups. B: densitometric analysis of phospho-p44 ERK (n = 4–6 animals). p44 Activation occurred only in the cytosol, and there was no difference between H2O2 and CCPA. *P < 0.05 vs. control. C: densitometric analysis of phospho-p42 ERK signal (n = 4–6 animals). The greatest difference between H2O2 and CCPA occurred in the Insol Memb compartment. *P < 0.05 vs. control and +P < 0.05 vs. H2O2.

p44/p42 ERK expression in cardiomyocyte subcellular fractions.

The distribution of the total p44/p42 ERK protein is shown in Fig. 5. The total p44 ERK protein was present primarily in the cytosolic (61%) and nuclear/myofilament (23%) fractions, with very little in the Triton-soluble (11%) and -insoluble (5%) membrane fractions (Fig. 5, A and B). The p44 ERK protein was significantly elevated in the Triton-soluble membrane fraction with H2O2 treatment, an effect not due to unequal protein loading. Similar to p44 ERK, the greatest p42 ERK expression was in the cytosolic (40%) and nuclear/myofilament (28%) fractions, with the total p42 ERK protein also present in the Triton-soluble (24%) and -insoluble (8%) membranes (Fig. 5, A and C). Although the Triton-insoluble fraction contained the least p42 ERK protein, the most significant stimulus-dependent differences in p42 ERK activation occurred in this fraction (Fig. 4). The total p42 ERK protein was not altered with H2O2 or CCPA.

Fig. 5.

Localization of p44/p42 ERKs. Distribution of p44/p42 ERKs in Cyto, Sol Memb, Insol Memb, and Nuc/Myo fractions was measured by Western blot analysis. A: representative Western blot of total p44/p42 expression in C, H, and CP groups. B: densitometric summary of total p44 protein signal (n = 3–5 animals). *P < 0.05 vs. control and +P < 0.05 vs. H2O2. C: densitometric summary of total p42 ERK immunoreactivity from pooled Western blots (n = 3–5 animals). There were no changes in p42 protein with drug treatment.

CCPA activates PP2A in Triton-insoluble membranes.

Adenosine A1 receptor stimulation has been reported to activate PP2A (25, 30), which may explain the reduced p38 MAPK and p42 ERK phosphorylation in Triton-insoluble fractions in the presence of CCPA. The results indicated that CCPA increased PP2A activity 3.4-fold in the Triton-insoluble fraction, whereas H2O2 had no effect on the activity of this phosphatase (Fig. 6). PP2A activity was measured only in the Triton-insoluble fraction since MAPK phosphorylation was not reduced by CCPA or H2O2 in other fractions (Figs. 2 and 4).

Fig. 6.

CCPA activates protein phosphatase 2A (PP2A) in Insol Memb fractions. PP2A activity was measured with a commercially available phosphatase assay kit. PP2A activity was defined as the okadaic acid-inhibitable portion of the signal. The A1 agonist CCPA significantly increased PP2A activity. *P < 0.05 vs. control.

Effect of MPG on subcellular p38 MAPK and p42 ERK activation.

The role of ROS in H2O2- and CCPA-mediated MAPK signaling was next investigated (Fig. 7). In these experiments, the ability of the antioxidant, MPG, to attenuate MAPK activation in the H2O2 and CCPA groups was examined. Pretreatment with MPG reduced H2O2-mediated p38 MAPK phosphorylation by 43% in the soluble membrane and by 55% in the insoluble membrane fractions, whereas there was no effect of this antioxidant in the cytosolic and nuclear/myofilament fractions (Fig. 7A). The effect of MPG on CCPA-mediated p38 MAPK phosphorylation was only determined in the cytosolic and nuclear/myofilament fractions since these were the only fractions in which p38 was activated by this A1 agonist. MPG had no effect on the CCPA-mediated p38 activation in these fractions (Fig. 7A).

Fig. 7.

Effect of the antioxidant 2-mercaptopropionyl glycine (MPG) on p38 MAPK and p42 ERK signaling. A: densitometric summary of pooled antiactive p38 MAPK Western blots (n = 3 animals). MPG attenuated H2O2-mediated p38 MAPK activation in the Sol Memb and Insol Memb fractions. No effect of MPG on CCPA-mediated p38 MAPK signaling was observed. *P < 0.05 vs. H2O2. B: densitometric summary of phospho-p42 ERK signal. H2O2-mediated p42 ERK activation was attenuated by MPG pretreatment in the Cyto, Sol Memb, Insol Memb, and Nuc/Myo fractions. MPG had no effect on p42 activation in the CCPA-treated samples. *P < 0.05 vs. H2O2.

H2O2-mediated p42 ERK activation was attenuated with MPG in all fractions, with the greatest reduction (88%) occurring in Triton-insoluble fractions (Fig. 7B). ERK activation in the CCPA + MPG group was examined in the cytosolic fraction since this fraction was the only one in which ERK was activated by this agonist. MPG had no effect on the CCPA-mediated p42 or p44 activation in the cytosol.

ROS formation in H2O2- and CCPA-treated cardiomyocytes.

To support the signaling results obtained with MPG, the effects of H2O2 and CCPA on cardiomyocyte ROS formation were investigated (Fig. 8). H2O2 elevated oxidative stress by 2.0-fold, 3.5-fold, and 4.1-fold after 5, 10, and 15 min, respectively, as measured by dichlorodihydrofluorescein (DCF) fluorescence. By contrast, ROS levels during CCPA treatment did not increase from baseline and were not different from the control group at any time point.

Fig. 8.

Oxidative stress in cardiomyocytes. Formation of reactive oxygen species/reactive nitrogen species was detected by dichlorofluorescein (DCF) fluorescence (n = 3 to 4 hearts/treatment group). All values are expressed relative to baseline. *P < 0.05 vs. control. H2O2 progressively increased the DCF signal, whereas CCPA had no effect.

DISCUSSION

The most significant finding of this study is that adenosine A1 receptor activation and oxidative stress exhibited different patterns of cardiomyocyte subcellular MAPK signaling. Whereas both H2O2 and the adenosine A1 receptor agonist, CCPA, activated cytosolic p38 MAPK and p42 ERK, only H2O2 stimulated these kinases in Triton-soluble and -insoluble membrane fractions. Furthermore, CCPA significantly reduced p38 MAPK and p42 ERK activation in Triton-insoluble membrane fractions, an effect likely due to an A1 receptor agonist-induced increase in PP2A activity in this fraction. Thus cardiomyocyte membrane fractions, specifically the Triton-insoluble component, are the site of divergence in MAPK signaling by different stimuli. Another difference between these two stimuli is that adenosine A1 receptor-mediated MAPK activation was not ROS dependent. In contrast, MPG attenuated H2O2-induced MAPK signaling, and ROS levels were also increased with H2O2. Thus these results suggest that the activation of MAPKs in different subcellular compartments may allow diverse stimuli to elicit distinct physiological effects through the same MAPK.

Existing studies of myocardial MAPK signaling have focused primarily on nuclear and cytosolic fractions (5, 13, 14, 40, 41, 43), and studies in cardiomyocytes have been based on whole cell lysates (22, 26, 50). Both adenosine A1 receptor activation and H2O2 have been reported to activate p38 in cardiomyocyte whole cell lysates (22, 26, 50), but the modulation of this MAPK in specific cardiomyocyte compartments by these stimuli has not been previously studied. In the current study, little difference between oxidative stress and A1 receptor-mediated MAPK activation occurred in cardiomyocyte cytosolic and nuclear fractions, suggesting that these fractions do not mediate stimulus-specific signaling.

Our previous studies indicated that an adenosine receptor agonist and ischemia-reperfusion differentially modulated MAPKs in myocardial membrane fractions (5, 43). The results of the present study indicate that the primary differences between oxidative stress and adenosine A1 receptor activation of cardiomyocyte MAPKs also occur in the membrane fractions, where the least amount of MAPKs were detected. H2O2 activated p38 MAPK and p42 ERK in the Triton-soluble and -insoluble membrane compartments, whereas CCPA had no effect on MAPK phosphorylation levels in the Triton-soluble fraction and significantly decreased MAPK activation in the Triton-insoluble fraction. Thus the Triton-insoluble portion of cardiomyocyte membranes exhibited the greatest signal specificity. Protein scaffolding may mediate this stimulus-specific signaling since several components in the MAPK signaling cascade are associated with scaffolding proteins (42), some of which are Triton insoluble (2, 24, 32). Harding et al. (17) reported that ERK signaling at the plasma membrane had a lower threshold for activation than the cytosol. This observation may also be due to protein scaffolding in the membrane, which would result in signaling molecules from the same pathway being in close proximity for more efficient activation (17).

It is well recognized that two key scaffolding proteins, caveolin-3 (2, 24, 32) and α-actinin (24, 32), are Triton insoluble. Caveolin-3 is found in cardiomyocyte caveolae and T tubules (2, 24a, 34), whereas α-actinin anchors parallel filaments of F-actin throughout the cytoskeleton (24a, 34). Caveolin-3 was enriched in Triton-insoluble membrane fractions in the current study, as was a significant proportion of α-actinin. Protein scaffolding through caveolin-3 is thought to play an important role in MAPK signaling (2, 24a, 32, 34, 52). The MAPKs possess a caveolin-binding domain (34) and changes in caveolin-3 expression alter MAPK activation (15). In this study, about 90% of total p38α MAPK and p42 ERK protein was bound to caveolin-3 in the Triton-insoluble fraction. This finding suggests that the bulk of MAPKs was associated with caveolae in this fraction. We have previously reported that caveolar MAPK activation also occurs with myocardial ischemia-reperfusion (6).

Protein scaffolding through the actin cytoskeleton may also play an important role in MAPK signaling (19, 47). Colocalization of an upstream activator of ERK and p38 MAPK with α-actinin has been reported (10), suggesting that a portion of these MAPKs may be associated with α-actinin in Triton-insoluble membranes. There are also reports that the actin cytoskeleton is altered by oxidative stress and simulated ischemia through a p38 MAPK-dependent mechanism involving heat shock protein 27 (12, 33). Thus H2O2 may have also activated p38 MAPK and p42 ERK in the actin cytoskeletal component of the Triton-insoluble membranes.

The disparate effects of adenosine A1 receptor stimulation and oxidative stress on Triton-insoluble membrane MAPK signaling in the present study could be due to the differential modulation of protein phosphatases. CCPA reduced p38 MAPK and p42 ERK phosphorylation below the control levels in this fraction, consistent with our observations that adenosine A1 receptor activation increased PP2A activity in the Triton-insoluble membrane fraction. In addition, PP2A inhibitors block the antiadrenergic effect of A1 agonists, suggesting that this phosphatase is indeed involved in membrane adenosine A1 receptor signaling mechanisms (25, 30). Liu and Hofmann (26) reported that adenosine A1 receptor stimulation induced the translocation of PP2A to crude particulate fractions, an effect which may be indicative of phosphatase activation. The results in the present study extend this observation since the activation of PP2A by an A1 agonist in caveolin-enriched, Triton-insoluble membrane fractions has not been previously reported.

Other significant findings in this study were the differences observed between p44 and p42 ERKs. The expression of p42 was 2 to 3-fold greater than of p44, and the activation pattern of the two ERK isoforms was also different. H2O2 activated p42 in all fractions, whereas p44 was only activated by this stimulus in the cytosolic and nuclear/myofilament fractions. There was no difference in H2O2- and CCPA-mediated MAPK activation of p44, whereas the activation pattern of p42 was quite different with these stimuli. These results suggest that the differential compartmentation of p42 and p44 ERKs may produce different, isoform-specific physiological effects. This hypothesis is supported by a report that p42 knockout mice show increased myocardial infarct size, whereas p44 knockout mice exhibit no difference in injury (23).

Further evidence that the differential compartmentation of cardiomyocyte MAPKs may produce distinct downstream effects is found in the regulation of the Na+/H+ exchanger. The activation of this exchanger by H2O2 occurs via an ERK-dependent mechanism (50). In contrast, Avkiran and Yokoyama (4) reported an A1 receptor-induced inhibition of α1-adrenergic-mediated stimulation of the Na+/H+ exchange. It has also been shown that adenosine A1 receptor-mediated inhibition of this exchanger may occur via the activation of PP2A (45). Since the exchanger has been found in caveolin-enriched, Triton-insoluble membranes (8), it is possible that the differential modulation of Triton-insoluble membrane-associated p42 ERK by CCPA and H2O2 may mediate the distinct effects of these stimuli on Na+/H+ exchanger activity. Furthermore, the differential modulation of this exchanger in Triton-insoluble membranes may explain the deleterious effect of excessive oxidative stress and the protective effect of adenosine A1 receptor activation since increased Na+/H+ exchanger activity exacerbates myocardial ischemia-reperfusion injury (45, 50).

Another major finding in this study is that differential subcellular MAPK signaling in response to adenosine A1 receptor activation and oxidative stress is associated with differences in the cardiomyocyte redox state. There are reports that stimulation of certain GPCRs is associated with increased cardiomyocyte ROS production (1, 35, 36, 48) and MAPK activation (7, 26, 29, 37), suggesting a linkage between the two. However, the results in the present study indicate that stimulation of the cardiomyocyte adenosine A1 receptor is not associated with ROS production, consistent with a previous report from our laboratory (31). Since ROS formation may occur in discrete compartments and differentially regulate MAPKs (16, 48), it is possible that CCPA could have increased ROS production in a specific subcellular compartment that was undetectable by total cellular DCF fluorescence measurements. However, this is unlikely since A1 receptor-induced MAPK activation was not altered by the antioxidant MPG. In contrast, H2O2 increased ROS formation and MPG reduced H2O2-mediated MAPK signaling. These results suggest that our observed differences in subcellular signaling may be due to differences in the myocyte redox state. Our findings are consistent with Palomeqmeque et al. (37), who recently reported that angiotensin II activated p38 MAPK without increasing ROS in adult cardiomyocytes. Bogoyevitch et al. (7) also found that the antioxidant N-acetyl-l-cysteine blocked ERK activation by H2O2 but had no effect on phenylephrine-mediated activation of this kinase.

In summary, the results of the present study indicate that although H2O2 and CCPA had similar effects on MAPK activation in cytosolic and nuclear fractions, MAPK signaling differed significantly in membrane fractions. The greatest difference between the two stimuli occurred in the Triton-insoluble portion of the membrane, where H2O2 increased and CCPA decreased the level of p38 MAPK and p42 ERK phosphorylation via PP2A. In addition, CCPA-mediated MAPK activation did not appear to involve ROS production. Thus these results suggest that the differential compartmentation of signaling may allow various stimuli to exert distinct physiological effects through the same signaling pathway. These results further indicate that protein scaffolding may play an important role in the stimulus-specific effects on signaling.

GRANTS

This research was supported by National Heart, Lung, and Blood Institute Grants 66132 (to R. Lasley) and 34579 (to R. Mentzer Jr) and the American Heart Association Grant 0465166B (to C. Ballard-Croft).

Footnotes

  • The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

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. 24a.
  26. 25.
  27. 26.
  28. 28.
  29. 29.
  30. 30.
  31. 31.
  32. 32.
  33. 33.
  34. 34.
  35. 35.
  36. 36.
  37. 37.
  38. 38.
  39. 39.
  40. 40.
  41. 41.
  42. 42.
  43. 43.
  44. 44.
  45. 45.
  46. 46.
  47. 47.
  48. 48.
  49. 49.
  50. 50.
  51. 51.
  52. 52.
  53. 53.
  54. 55.
View Abstract