Inhibition of p38 reduces myocardial infarction injury in the mouse but not pig after ischemia-reperfusion

Robert A. Kaiser, Jefferson M. Lyons, Jodie Y. Duffy, Connie J. Wagner, Kelly M. McLean, Timothy P. O'Neill, Jeffrey M. Pearl, Jeffery D. Molkentin


The MAPK family member p38 is activated in the heart after ischemia-reperfusion (I/R) injury. However, the cardioprotective vs. proapoptotic effects associated with p38 activation in the heart after I/R injury remain unresolved. Another issue to consider is that the majority of past studies have employed the rodent as a model for assessing p38's role in cardiac injury vs. protection, while the potential regulatory role in a large animal model is even more uncertain. Here we performed a parallel study in the mouse and pig to directly compare the extent of cardiac injury after I/R at baseline or with the selective p38 inhibitor SB-239063. Infusion of SB-239063 5 min before ischemia in the mouse prevented ischemia-induced p38 activation, resulting in a 25% reduction of infarct size compared with vehicle-treated animals (27.9 ± 2.9% vs. 37.5 ± 2.7%). In the pig, SB-239063 similarly inhibited myocardial p38 activation, but there was no corresponding effect on the degree of infarction injury (43.6 ± 4.0% vs. 41.4 ± 4.3%). These data suggest a difference in myocardial responsiveness to I/R between the small animal mouse model and the large animal pig model, such that p38 activation in the mouse contributes to acute cellular injury and death, while the same activation in pig has no causative effect on these parameters.

  • signaling
  • kinase
  • infarction
  • heart

the mapk family member p38 functions as a stress- and membrane receptor-responsive signaling effector that transduces signals from the cell membrane to the cytoplasm and nucleus. In the heart, p38 is activated within 10 min of ischemia, but it is not significantly reactivated during reperfusion (13). Although the kinetics of p38 activation in myocardial ischemia are described (4), the ultimate biological effects associated with p38 activation in the heart remain controversial. In some reports, p38 activation reduces cardiomyocyte apoptosis, and its inhibition is associated with increased cell death (6, 7, 25). Still others observe that p38 activation induces apoptosis and inhibition promotes cell survival in vitro (5, 15) and reduces the extent of myocardial infarction injury or functional deficits thereafter in vivo (8, 11, 18, 21, 24). Although most in vivo studies that have examined the biological effects associated with p38 inhibition have employed rodent models, some translational studies in the pig have been performed (1, 3). Because of profound differences in design of these studies and others, there is no clear consensus regarding the role of p38 in infarction injury.

Rodent myocardium is physiologically different in many respects from the large animal (pig and human) myocardium, which is especially relevant when translating animal experiments to human disease (12). Therefore, encouraging reports in rodent models of ischemia-reperfusion (I/R) may not be substantiated when tested in larger animals like the pig. Here we attempted to address this and other issues by performing parallel observations in closely matched mouse and pig models of I/R to examine the functional role of p38 in each context.


Mouse I/R model.

All procedures with animals were approved by the Institutional Animal Care and Use Committee. Mouse left anterior descending coronary artery (LAD) was isolated and occluded as described previously (11). SB-239063 (7 μg, n = 15) or PPCES vehicle (PEG-400:propylene glycol:cremaphor-EL:ethanol:saline 30:20:15:5:30; n = 17) alone was injected into the tail vein 5 min before ischemia. Suture was tied in a slipknot around the LAD, and mice were revived by removal from anesthetic during 60 min of ischemia, after which the knot was released and the heart was reperfused for 24 h. Mice were killed by CO2 asphyxiation, and hearts were analyzed as previously described using 2% triphenyltetrazolium chloride (TTC) in saline (11). Briefly, ligatures were retied around the LAD at the original occlusion, and the heart was injected with 2% Evan's blue in saline. Hearts were sectioned and stained in 1.5% TTC in saline, fixed in 10% formalin in PBS, and photographed. Myocardial area not at risk (stained blue), area at risk (AAR, white or red), and infarcted area (white area) were quantified using ImageJ software (Scion, Frederick, MD).

Porcine I/R model.

Farm-raised pigs weighing 25–30 kg (n = 25) were initially sedated with ketamine and acepromazine, intubated, and maintained on intravenous pentobarbital and fentanyl. Pigs were mechanically ventilated, and the heart was accessed via a midline sternotomy. Amiodarone was given intravenously at 150-mg bolus and maintained at 50 mg/h until the end of ischemia. Pigs were injected with either 7 mg SB-239063 (n = 11) or PPCES vehicle (n = 7) directly into the left ventricular (LV) lumen, and the drug was allowed to perfuse for 5 min before LAD occlusion by a snare at one-third the distance from the apex. Ischemia was maintained for 60 min, and the snare was then released for 4 h (10). Blood samples were obtained at baseline, end of ischemia, and end of reperfusion. Pig hearts were analyzed for infarction as described above for mice using TTC. A total of 23% inhibitor and 25% vehicle pigs required internal defibrillation during the ischemia, respectively. Animals requiring more than three intrathoracic 20-J pulses or more than 1 ml 1:1,000 epinephrine (1 mg total) for conversion were excluded.

Western blotting.

Mouse surgeries (n = 6/treatment group) were performed as described, and hearts were excised at 10 min of ischemia for protein determinations. Similarly, pig heart biopsies (n = 6/treatment group) were taken at baseline, 10 and 20 min ischemia, and 10 and 20 min reperfusion. Mouse and pig samples were homogenized in protein buffer (11) and centrifuged at 13,500 rpm to pellet debris. One-hundred micrograms of each protein sample were separated on 12% polyacrylamide gels and blotted with antibodies for total p38, ERK1/2, JNK1/2, phosphorylated (p)-p38, p-ERK1/2, p-JNK1/2, p-heat shock protein 27 (p-HSP27), and p-MAPK-activated protein kinase 2 (p-MAPKAPK2) (Cell Signaling, Beverly MA). Quantitation was performed using ImageQuant software to determine the relative amounts of phosphorylation.

Statistical analysis.

All data were analyzed with Graphpad Prism software (Graphpad, San Diego, CA) using Student's t-test for unpaired data or ANOVA for repeated measures followed by Bonferroni posttest for serial protein phosphorylation observations and creatine kinase levels. Significance was considered at P < 0.05. Data are represented as means ± SE.


Given the controversy surrounding the effect of p38 inhibition on myocardial infarction injury, here we employed the most specific p38 inhibitor to date, SB-239063 (2), to evaluate p38 in mouse and porcine models of I/R. We first examined p38 activation in mice treated with vehicle or SB-239063 after 10 min of ischemia. The data demonstrated augmented p38 phosphorylation in untreated or vehicle-treated mice but no activation in mice treated with SB-239063 (Fig. 1, A and B), supporting previous reports that p38 inhibition can prevent p38 phosphorylation (9). No differences were observed in the phosphorylation status of ERK1/2 or JNK1/2 between SB-239063- and vehicle-treated mice (data not shown). To determine the effect of SB-239063 on p38 signaling, we blotted for phosphorylation of the known p38 substrate MAPKAPK2 (19). Phosphorylation of MAPKAPK2 was also reduced by SB-239063 infusion compared with vehicle controls (Fig. 1, A and C). Furthermore, SB-239063 pretreatment significantly reduced infarction area (vehicle 37.5 ± 2.8%, SB-239063 28.0 ± 2.9%) after 60 min ischemia and 24 h reperfusion (Fig. 1D), consistent with a previous study in rats (8). These results support the conclusion that p38 activation is a factor in infarction development in the mouse myocardium, consistent with our own previous study in dominant negative p38 mice, a separate study in heterozygous p38-targeted mice, and observations made with the related drug SB-203580 (11, 18, 24).

Fig. 1.

SB-239063 treatment effectively reduces p38 signaling in the murine heart. A: Western blots of protein samples from hearts of naive mice (sham operated or ischemia) or mice given vehicle or SB-239063 and subjected to 10 min of left anterior descending coronary artery (LAD) occlusion. p-p38, phosphorylated p38; p-MAPKAPK2, phosphorylated MAPK-activated protein kinase 2. B: histogram of p-p38 levels from the drug treatment groups as in A. AU, arbitrary units. C: histogram of p-MAPKAPK2 levels from the drug treatment groups as in A. Histograms present data from 3 separate experiments with duplicate observations in each group; n = 6 total per group. #P < 0.05 vs. sham operated, *P < 0.05 vs. vehicle treated. D: histogram of infarct size presented as a scattergram and average ± SE for both vehicle- and SB-239063-treated mice. Infarct area is presented as the percentage of the area at risk that infarcted as a result of 1-h ischemia and 24-h reperfusion in each group (n = 17 vehicle; n = 15 SB treated, *P < 0.05).

The I/R protocol used in the porcine model is detailed in Fig. 2A. Needle biopsies were taken from the ischemic region of the LV during the equilibration phase (arrow A), 10 and 20 min of ischemia (arrows B and C), and 10 and 20 min of reperfusion (arrows D and E). Western blotting demonstrated increased p38 phosphorylation in vehicle-treated animals after ischemia but not in SB-239063-treated and nonischemic sham-operated animals (Fig. 2B). No differences were observed in the phosphorylation status of ERK1/2 or JNK1/2 in the SB-239063- and vehicle-treated animals (data not shown). Additionally, SB-239063 treatment reduced MAPKAPK2 phosphorylation (Fig. 2, B and C) and phosphorylation of a MAPKAPK2 substrate, heat shock protein (HSP27) (19) at 10 and 20 min of ischemia (Fig. 2, B and D), suggesting depression of the p38 pathway during ischemic time points known to correlate with increased p38 activity (4). No differences were observed in the AAR normalized to LV area in any of the cohorts (data not shown), and serum creatine kinase (CK) levels were similarly elevated between vehicle- and SB-239063-treated animals, indicating similarity of injuries in both treatment groups (Fig. 3A). Surprisingly, even though SB-239063 inhibited p38 pathway activation in the ischemic pig heart, there was no difference in infarct size between vehicle- and SB-239063-treated animals (Fig. 3B).

Fig. 2.

SB-239063 inhibits p38 pathway activation in ischemic porcine hearts. A: schematic of the pig ischemia-reperfusion protocol. Needle biopsies were taken from the at-risk myocardium at the times indicated by the arrows (arrow A, during equilibration; arrow B, 10 min of ischemia; arrow C, 20 min of ischemia; arrow D, 10 min of reperfusion; arrow E, 20 min of reperfusion). B: Western blotting of protein from porcine hearts at the times indicated in A. p-HSP27, phosphorylated heat shock protein 27. C: quantitation of p-MAPKAPK2 from multiple Western blots for the treatments shown in B. Shading denotes the ischemic period. D: quantitation of p-HSP27 from multiple Western blots for the treatments shown in B. Shading denotes the ischemic period. (*P < 0.05 compared with sham operated and SB-239063; n = at least 6 individual hearts examined for each time point.)

Fig. 3.

Inhibition of p38 does not preserve viability of the porcine myocardium. A: blood levels of creatine kinase (CK) measured at equilibrium (zero), end of ischemia (60 min), and end of reperfusion (300 min) for sham-operated, vehicle-treated, and SB-239063-treated pigs. Shading denotes the ischemic period. *P < 0.01 vs. sham operated. B: infarcted area of porcine heart is represented as the percentage of the area at risk that proceeded to infarct after 60 min ischemia and 240 min reperfusion period. Average ± SE and individual determinations are presented for vehicle- and SB-239063-treated pigs (n = 7 and 11, respectively).

Collectively, these observations suggest that p38 plays a more important role in signaling myocyte death in the ischemic mouse heart than in the pig. Indeed, in many large animal models, p38 activation does not appear to be as clearly proapoptotic as the majority of rodent studies would suggest. In dogs, for example, another p38 inhibitor (SB-203580) infused into the coronary artery during ischemia had a similar cardioprotective effect as ischemic preconditioning, while the same dose administered during preconditioning abrogated protection (20). In pigs, SB-203580 infused systemically at similar doses to those reported here limited infarction acutely but had no inhibitory effect on the benefits of preconditioning (1). It was further reported that two structurally distinct p38 inhibitors (including SB-203580) could not affect infarct size alone but could abrogate ischemic preconditioning (22), apparently by a mechanism involving p38β and connexin 43 (23). In the rabbit, infusion of SB-203580 exerted a significant cardioprotective effect during ischemia in a Langendorff preparation (14), although this same inhibitor again abolished the cardioprotective effect of ischemic preconditioning (17). Differences between this report and previous reports, as well as between the various studies cited herein regarding surgical technique, drug administration, anesthetic regimen, and end point assessment, result in profound discrepancies between basal levels of infarct and efficacy of pharmacological perturbations, even within species. For example, one study cited herein used a myocardial infusion of an older p38 inhibitor (SB-203580) that bypassed the vasculature entirely (1) and observed myocardial protection after only a 60-min reperfusion period. The discrepancy with the current report with regard to infarction may be in part due to anesthetic regimen, reduced reperfusion period, or even differences in the defined AAR compared with the drug-treated area, which are not rigorously related in the infusion protocol. In the current report we eliminate many of these variables and conclude that p38 in the mouse contributes to infarction, while in the pig, p38 has no effect on acute injury development.

The mechanism of p38 activation that results in either proapoptotic or prosurvival signaling is likely to be dependent on timing, duration, and intensity of activation. In fact, cardiac p38 can be activated by multiple stimuli, including pacing, stretch, osmotic stress, inflammatory cytokines, and activation of Gαq-coupled receptors (16), many of which do not elicit cell death. Furthermore, the ramifications of p38 activation appear to be heavily dependent on the animal model employed. In any pharmacological study, it is possible that unpredicted and/or nonspecific effects of a compound can confound the interpretation of results. SB-239063, employed in the current study, has been described as highly selective for p38 over other known kinases in rodent (2), although this information is not yet available for porcine kinase isoforms. Additionally, the surgical procedures employed herein were somewhat different between the murine and porcine models because of technicalities of the individual preparations. Despite these concerns, the lack of an acute protective effect in pig suggests a potential lack of benefit for p38 inhibition in preventing acute myocardial infarct damage in large mammals, although p38 inhibition might still benefit long-term remodeling and development of cardiomyopathy.


This work was supported by the National Institutes of Health (NIH) and a Pew charitable Trust Scholar Award and a Translational Research Initiative Grant from the Children's Hospital Research Foundation (J. D. Molkentin). R. A. Kaiser was supported by NIH National Research Service Award HL-073550.


  • 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.


View Abstract