|
|
||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Departments of 1Pediatrics and 2Medicine, Columbia University, New York, New York; and 3Department of Pediatrics, Hospital for Sick Children, Toronto, Ontario, Canada
Submitted 31 March 2007 ; accepted in final form 28 August 2007
| ABSTRACT |
|---|
|
|
|---|
pregnancy; cardiac hypertrophy; nitric oxide; mitogen-activated protein kinases
Mitogen-activated protein kinases (MAPKs) are intracellular signal transduction factors that promote cardiac hypertrophy through activation of downstream proteins like phosphatidylinositol 3-kinase/Akt, also known as protein kinase B. The three best-characterized MAPKs include p38; JNK, or c-Jun NH2-terminal kinase; and ERK1/2, or extracellular signal-regulated kinase (also known as p44/42). ERK1/2 has been proposed to regulate smooth muscle contraction and to promote hypertrophy (10). It is activated principally by mechanical stretch and mitogenic stimuli and has been associated with a growth of uterine vasculature and a dramatic increase in uterine artery blood flow during pregnancy (13). Recent reports demonstrate that persistent activation of p38 and JNK, on the other hand, can promote apoptosis, resulting in cardiac dilatation and dysfunction (22). In addition, Linke et al. (15) showed that endothelial nitric oxide synthase (eNOS) protein expression is elevated in the myocardium of pregnant rats and returns to control levels at 4 days after giving birth. Myocardial eNOS also seems to increase its expression in several vascular beds in pregnancy, and recent data suggest nitric oxide (NO) as a prime mediator for the fall in vascular resistance (8, 24). An important mediator of eNOS activation is Akt, or protein kinase B. While Akt increases eNOS phosphorylation, it can also have independent anti-growth effects by promoting apoptosis (7). ERK activation can have an inhibitory effect on Akt-mediated eNOS activation in some vascular beds (2). A delineation of these interacting molecular pathways in mediating both the physiological remodeling of pregnancy as well as the rapid reverse remodeling after removal of the pregnancy load is of considerable interest, since these signaling pathways may also play a role in the pathological remodeling seen in various disease conditions characterized by chronic volume overload. Thus the objective of this study was to determine the role of MAPKs, Akt, and NO in mediating cardiac remodeling during pregnancy and in the postpartum period.
| MATERIALS AND METHODS |
|---|
|
|
|---|
Echocardiography
All rats underwent transthoracic two-dimensional echocardiography after sedation with pentobarbital sodium (25 mg/kg ip). The rats were weighed, the left hemithorax was shaved, and transthoracic echocardiographic images were obtained in sedated animals using a Acuson Sequoia equipped with a 13-MHz linear transducer (15L8) in a phased-array format with real-time digital acquisition, storage, and review capabilities. During echocardiographic assessment, rats were laid supine with the body tilted to the left side to improve visualization of the heart as well as avoid compression of the inferior vena cava. The heart was imaged in two-dimensional mode using standard parasternal short-axis, apical four-chamber, and parasternal long-axis views to measure LV chamber dimensions and wall thickness and to calculate LV mass, volume, and ejection fraction. Three beats were averaged for each measurement. The details of these techniques have been previously published (1). All calculations were derived using standard formulas (21).
LV chamber dimensions and wall thickness. LV end-diastolic (LVEDD) and end-systolic diameters (LVESD) and wall thickness were measured from M-mode tracings.
LV volume and ejection fraction. LV volume and ejection fraction were calculated from the two-dimensional parasternal long-axis view using the single-plane area-length method.
LV mass. LV mass was calculated according to uncorrected cube assumptions using the equation, LV mass = 1.055[(IVST + LVEDD + PWT)3 – (LVEDD)3], where 1.055 is the specific gravity of the myocardium, IVST is the interventricular septal thickness, and PWT is the posterior wall thickness.
Myocardial Protein Expression
Four rats from each group were anesthetized with pentobarbital sodium (50 mg/kg ip) before extraction of the hearts by thoracotomy and euthanization of the animals. A piece of LV myocardium was obtained and immediately snap-frozen in liquid nitrogen. Proteins were prepared from myocardial cell lysates and separated on 10% precast Tris·HCl gels (Bio-Rad, Melville, NY). Separated proteins were transferred by a semidry method onto a polyvinylidene difluoride membrane (0.2 µm). The membranes were immunoblotted using the following primary antibodies in a buffer containing 10 mM Tris·HCl, 100 mM NaCl, 0.1% Tween 20, and 5% bovine serum albumin (pH 7.5): anti-eNOS antibody (1:1,000; Transduction Laboratories, Torreyana, CA), anti-phospho-Ser473 Akt antibody (1:500), Akt (1:500), anti-ERK antibody, anti-p38 MAPK antibody, and anti-JNK antibodies (phosphospecific and total, 1:1,000; Cell Signaling Technology, Beverly, MA, and Santa Cruz Biotechnology, Santa Cruz, CA).
-Actin was used as a loading control. After overnight incubation, the membranes were washed and incubated with the secondary antibody anti-rabbit IgG (1:2,000, Cell Signaling Technology). Protein abundance was quantified by scanning developed immunoblots containing known amounts of total homogenate protein. Signals were detected by enhanced chemiluminescence. Maximum light emission at a wavelength of 428 nm was detected by short exposure to blue light-sensitive autoradiography film, and optical density of scanned blots was measured using ImageQuant software and expressed in arbitrary units (AU).
Statistical Analysis
Echocardiographic and Western blot data were compared among different groups using multiway ANOVA and post hoc Tukey's test or Student's t-test as appropriate. Linear regression analyses were performed to determine the relationship between protein expression and echocardiographic measures of LV mass and volume. A P value <0.05 was considered significant.
| RESULTS |
|---|
|
|
|---|
During pregnancy, there was an increase in body weight, ventricular weight, LV mass (0.62 ± 0.03 to 1.1 ± 0.04 g, P < 0.001), LV mass/volume ratio (0.7 ± 0.02 to 1.3 ± 0.02 g/ml, P < 0.0001), and LV ejection fraction (64 ± 3 to 74 ± 2%) (Table 1, Fig. 1). Ventricular weight, LV mass and mass/volume ratio returned to nonpregnant values in the postpartum period. However, LV ejection fraction decreased to below the normal range immediately postpartum and this decrease persisted at 14 days postpartum (53 ± 3%, P < 0.05 vs. pregnant; P = 0.07 vs. nonpregnant).
|
|
The increase in LV mass and decrease in ejection fraction during pregnancy was paralleled by a decrease in the expression of the anti-hypertrophic factors, i.e., phosphorylated p38, JNK, and Akt (Figs. 2 and 3). The reversal of LV hypertrophy postpartum was accompanied by an increase in the anti-hypertrophic and decrease in prohypertrophic kinases. p38, JNK, and Akt phosphorylation increased to baseline levels, i.e., nonpregnant values, although JNK continued to increase beyond the immediate postpartum period resulting in higher than baseline levels by 14 days postpartum. Similarly, eNOS increased significantly by 14 days postpartum. ERK, a prohypertrophic kinase, on the other hand, did not change significantly during pregnancy but showed a rapid decrease postpartum with normalization by day 14 (Figs. 2 and 3). The decrease in phosphorylation ratios of p38 and JNK was due to a decrease in phosphorylation of the respective kinases rather than a decrease in total kinase. ERK expression, on the other hand, demonstrated a lower total and phosphorylated ERK, but the decrease in phosphorylated ERK was disproportionate to the decrease in total ERK.
|
|
There was a negative correlation between p38 and JNK activation and LV mass/volume ratio (r = –0.81 and –0.71, respectively; P < 0.05), i.e., higher levels of p38 and JNK were associated with lower LV mass/volume ratio (Fig. 4).
|
| DISCUSSION |
|---|
|
|
|---|
Pregnancy represents an excellent model of acute physiological LV hypertrophy and atrophy secondary to a transient, self-limited hemodynamic load. The molecular mechanisms underlying these processes are not well studied. Our phenotypic findings in pregnant rats were similar to those seen in human pregnancy (16). We demonstrated an increase in LV weight, mass, mass/volume ratio, and function during pregnancy. We demonstrated a 17% increase in ventricular weight during pregnancy and a 40% increase in LV mass during pregnancy. Published studies in humans and rats have reported between a 17 and 28% increase in LV mass during pregnancy (12, 17). The somewhat higher increase in LV mass when measured echocardiographically may reflect either the propensity for echocardiography to overestimate changes in LV mass or the study of relatively young rats (9-wk old), which may have shown an exaggerated LV hypertrophy in response to pregnancy load compared with older rats. We also demonstrated a strong association between the expression of growth kinases and LV hypertrophy during (and after) pregnancy. For example, during the hypertrophic phase of pregnancy, there was a significant decrease in p38, JNK, and Akt (which can have anti-growth effects) without a change in ERK, a progrowth kinase. Therefore, an altered balance between pro- and anti-growth kinases may have been responsible for the physiological hypertrophy of pregnancy. This was supported by the finding of an inverse correlation of p38 and JNK activation with LV mass/volume ratio. p38 and JNK were downregulated during the immediate postpartum period. This may reflect a delay in upregulation of MAPKs following removal of the hemodynamic load of pregnancy. By day 14 postpartum, however, p38 and JNK were upregulated, and this was associated with a reversal of cardiac hypertrophy and normalization by 14 days postpartum. Of further interest was the finding that, whereas ERK and p38 were normalized by 2 wk postpartum, JNK increased to higher than prepregnant levels. This was associated with a decrease in systolic function that persisted at 2 wk postpartum despite the removal of the hemodynamic load of pregnancy and despite the normalization of LV mass. The early decrease in LV function in the immediate postpartum period is likely triggered by the decrease in preload and increase in afterload caused by separation of the low-resistance placental circulation. Whether the upregulation of JNK contributed to the persistence of postpartum LV dysfunction requires further study. An increasing body of evidence suggests that persistent activation of JNK (and p38) can promote apoptosis and serve as a negative regulator of cardiac hypertrophy and function. Gain-of-function transgenic animals overexpressing these kinases have demonstrated progression from hypertrophy to dilated cardiomyopathy (11, 14). Longer follow-up postpartum is required to determine whether normalization of LV function is associated with normalization of JNK activation as well. Overall, our findings suggest that the early postpartum period is a vulnerable period, since the recovery of cardiac function appears to lag behind the normalization of LV morphology.
eNOS was not upregulated in rat myocardium during the third week of pregnancy. This is consistent with the findings of Jankowski et al. (12), who showed an increase in LV eNOS protein at day 14 but a decrease at day 21 of gestation in pregnant rats. This differs from the finding in pregnant dogs, wherein LV eNOS protein expression was increased at 60 days gestation (23). Many other studies have shown upregulation of eNOS in aorta and resistance arteries during pregnancy but often a downregulation in organs like the kidney (3, 4, 24). Therefore, our findings are consistent with published literature. We did not measure eNOS expression in systemic blood vessels in the present study, but it is possible that upregulation of eNOS during pregnancy may be organ specific and/or species specific. Also, eNOS was upregulated at 14 days postpartum. Although NO plays an important role in regulating systemic vascular tone and lowering systemic vascular resistance, which can benefit cardiac function, it can also have both anti-hypertrophic and negative inotropic effects at the myocardial level (19). The upregulation of eNOS, therefore, has the potential to contribute to persistent dysfunction postpartum. However, in the absence of a significant correlation between eNOS expression and LV dysfunction, there is insufficient evidence linking postpartum LV dysfunction to eNOS upregulation.
Clinical Relevance
Overall, the above findings suggest that MAPK signaling plays a key role in the physiological cardiac remodeling of pregnancy as well as reverse remodeling postpartum. The early postpartum period appears to be a particularly vulnerable period characterized by normalization of LV mass and volume but a delay in normalization of LV function and significant alterations in JNK signaling. These findings have potential clinical implications.
1) Peripartum cardiomyopathy is a rare but potentially lethal complication of pregnancy that is typically seen in women with risk factors including increased maternal age, multiparity, multiple pregnancy, and pregnancies complicated by preeclampsia and gestational hypertension (18). The hemodynamic decompensation is typically seen during the postpartum period, when the heart is particularly vulnerable to hemodynamic stressors (6). Although the underlying biological cause of peripartum cardiomyopathy is not known, an understanding of the physiological pathways that mediate cardiac remodeling will help our understanding of the disturbances in these processes that can potentially contribute to persistent postpartum cardiac dysfunction.
2) A more general application of our findings is that they provide a molecular signature for adaptive/physiological remodeling in response to load, against which other forms of pathological remodeling associated with chronic hemodynamic load, e.g., heart failure, can be compared. In addition, identifying the pathways that mediate reverse remodeling after pregnancy provides a basis for understanding how the heart adapts to removal of load. If these pathways could be manipulated in heart failure, either alone or in conjunction with mechanical unloading of the failing heart, it might help in optimizing myocardial recovery in heart failure (5, 9).
Limitations
1) This study did not establish a causal role for altered MAPK signaling in promoting cardiac remodeling during pregnancy and postpartum. The source of MAPK activation (i.e., vascular, myocytes, fibroblasts) was not studied. However, the finding of a significant correlation between MAPK activation and the degree of hypertrophy provides indirect support for a biological role for MAPKs in altering the cardiac phenotype. 2) We did not follow the animals long-term to determine whether LV function was normalized in the late postpartum period. 3) We did not investigate the role of the change in hormone levels during and after pregnancy in regulating MAPK, Akt, and eNOS activation. 4) We did not study myocyte ultrastructural changes to determine whether myocyte loss by apoptosis contributed to LV dysfunction postpartum.
| ACKNOWLEDGMENTS |
|---|
| 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 |
|---|
|
|
|---|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Visit Other APS Journals Online |