Infants of diabetic mothers (IDM) represent a high-risk group of patients with an increased incidence of perinatal morbidity and mortality [1]. Even with good maternal serum glucose control, the offspring are at increased risk for complex multisystem medical problems including macrosomia, hypoglycemia, hypocalcaemia, and respiratory distress syndrome [2]. Although not commonly seen in current medical practice, very poorly controlled gestational diabetes results in small for gestational age infants with relative sparing of internal organs such as the heart and significant multisystem medical problems [3].

The cardiomyopathy seen in very poorly controlled gestational diabetes is likely different from the hypertrophic cardiomyopathy (HCM) found in better controlled IDM [2, 4–6]. The HCM seen in infants of well controlled diabetic mothers is characterized by asymmetric septal hypertrophy with individual myocyte hypertrophy and limited myofiber disarray [6]. The functional significance of the HCM is variable, ranging from a clinically insignificant echocardiographic finding to congestive heart failure as a result of left ventricular outflow tract obstruction and diastolic dysfunction [7]. Regardless of severity, the cardiac hypertrophy is transient, with echocardiographic resolution by six months of age [6]. While the morphologic characteristics of the cardiomyopathy seen in IDM are well described, the molecular mechanisms regulating its development and resolution are currently unknown.

The heart is capable of remodeling in response to physiologic and pathologic stimulation. The mitogen activated protein kinase (MAPK) pathways play a key role in the response of cardiac myocytes to these stressful stimuli [8, 9]. The most studied terminal kinases of these pathways: extracellular signal regulated protein kinases (ERK), c-Jun NH₂-terminal kinases (JNK), and p38, activate transcription factors that ultimately affect gene expression [10]. In general, ERK activation (or phosphorylation) is pro-growth and has been implicated in hyperplasia of immature cardiomyocytes and hypertrophy of terminally differentiated cardiomyocytes [11]. The roles of JNK and p38 are less clear. In vitro, JNK and p38 activation promotes hypertrophy although in vivo these pathways appear to prevent hypertrophy [12]. In addition to the roles these pathways play in cell growth and proliferation, the MAPK signaling pathways are powerful regulators of apoptosis [13]. While apoptosis has been implicated in the development of some forms of cardiomyopathy, its role in the development and resolution of the HCM seen in IDM is unknown [14, 15].

In order to explore the role of the MAPK and apoptotic pathways in the development and resolution of the cardiomyopathy seen in infants of diabetic mothers, we developed two rat models of gestational hyperglycemia – one with very poorly controlled glucose with maternal ketosis and a second model with moderately well controlled glucose levels and no maternal ketosis. We hypothesized that the offspring of both the moderately and severely diabetic mothers would demonstrate a cardiomyopathy whose postnatal resolution was associated with activation of the MAPK signaling and apoptotic pathways.

Maternal diabetes creates an adverse environment for fetal development resulting in a number of anatomic, hematologic, and physiologic abnormalities in the newborn [19]. The effects of uncontrolled or long-standing diabetes on the fetus and the neonate are quite different from those found in well-controlled maternal diabetes. Prior to the use of insulin, few diabetic women conceived and even fewer carried a pregnancy to term [19]. In performing a post mortem quantitative morphologic study on infants of diabetic mothers and comparing them to controls matched for gestational and postnatal age, Naeye identified two populations of infants of diabetic mothers [3]. The first group of infants had a body weight that was 141% of the control values. In addition, these infants had organomegally with heart weights that were 174% of the control values. This group likely represented mothers with residual insulin function whose serum glucose values were elevated but did not become ketotic. The second group consisted of underweight infants with body weights 61% of control values and heart weights that were 51% of controls. This group was likely more severely ill and was thought to have placental insufficiency due to their degree of illness. The decrease in cardiac mass in the newborns was attributed to myocyte death [3].

The findings of our study are consistent with Naeye's observations. Specifically, OSDM had a decrease in body weight and heart weight when compared to control pups. Due to a relative sparing of heart weight compared to body weight, however, there was a significant increase in the heart weight to body weight ratio at both time points. These findings were associated with decreased left ventricular systolic function echocardiographically in the immediate newborn period. The difference between control and OSDM heart weight-to-body weight and cardiac function is likely related to the poorly controlled maternal diabetes and associated maternal anorexia, undernutrition and/or placental insufficiency. This theory is supported by both human and animal data. Echocardiographic studies have found a relative sparing of cardiac weight when compared to body weight in human pregnancies complicated by poorly controlled maternal diabetes and intrauterine growth restriction (IUGR) [20]. IUGR alone is also associated with decreased cardiac function that worsens with advancing gestation [21].

Several studies have described the cardiomyopathy seen in the macrosomic offspring of mothers with moderately well controlled gestational diabetes [2, 4–6, 22]. The HCM in these infants is characterized by asymmetric hypertrophy of the ventricular septum [6]. While increased septal thickness was not found by echo in the OMDM in this study, an increase in cardiac mass was seen at one-day of age in the OMDM compared to controls. As in humans, variability in septal thickness between animals may have prevented the identification of significant differences in this measurement between groups.

While multiple studies have described the morphologic and histologic characteristics of the cardiomyopathy seen in infants of diabetic mothers, no study to our knowledge has described the molecular mechanisms that correlate with the cardiomyopathic changes. The striking finding of this study is that regardless of the extent of intrauterine insult to the heart due to maternal hyperglycemia, the postnatal remodeling process was accompanied by activation of the MAPK signaling pathways. The MAPK signaling pathways have multiple effects on cardiac myocyte growth, proliferation and apoptosis and have been implicated in other forms of cardiomyopathy [14, 15]. We found that at one day of age, the hypertrophic cardiomyopathy in OMDM and the dilated cardiomyopathy in OSDM was associated with limited changes in MAPK activation, although small changes were found at NB1 in OSDM in total ERK and active (phosphorylated) JNK (Figures 1 and 2). In both models, the resolution of the cardiomyopathy at NB21 was associated with significant up regulation of both active JNK (pJNK) and active ERK (pERK). Levels of active P38 were unchanged at both time points. The downstream targets of these signaling proteins that contribute to the cardiac remodeling in the offspring of diabetic dams are currently not known.

ERK has the most well defined role of the MAPK proteins. In general, ERK is considered to be progrowth, promoting hyperplasia in uninucleated myocytes and hypertrophy in binucleated myocytes [11, 23]. Bueno et. al. examined the effects of ERK activation in transgenic mice with cardiac specific expression of the ERK activator, MEK1. In addition, they activated ERK in cultured myocytes by transfecting with MEK1 adenovirus. Both experiments demonstrated that the MEK1/ERK signaling pathway stimulated physiologic hypertrophy associated with augmented cardiac function and partial resistance to apoptosis [24]. In addition, ERK was found to play a key role in IGF-1 induced cardiac myocyte proliferation [25]. The significant increase in pERK in NB21 offspring in both models may reflect a general decrease in myocyte cell number in utero due to maternal hyperglycemia that is compensated for by an increase in pERK-driven myocyte proliferation postnatally. Measurement of myocyte size in the OMDM would help to define whether individual myocyte hypertrophy is present in these animals, which would be necessary if cardiac hypertrophy, as seem in the NB1 pups, was accompanied by decreased myocyte number.

The roles of JNK and p38 are less well defined than that of ERK. As recently reviewed by Liang and Molkentin, p38 and JNK have differing roles in vivo and in vitro [26]. In vivo, activation of p38 and JNK appears to inhibit hypertrophy and promote apoptosis, while in vitro, activation of these pathways appears to promote hypertrophy [8, 12, 27, 28]. Although phosphorylated p38 protein levels were unchanged, there was a striking increase in the levels of active JNK (pJNK) at NB21 in both models. Thus, at 21 days of age there appears to be activation of competing regulatory proteins in the hearts of the offspring of diabetic mothers, regardless of the degree of maternal hyperglycemia, with ERK being prohypertrophic and antiapoptotic and JNK being antihypertrophic and proapoptotic. Whether these competing influences work to appropriately couple myocyte hyperplasia with myocyte apoptosis, as has been suggested occurs during normal cardiac development [29], is not known.

We found indications that activation of apoptotic pathways was important to the cardiac remodeling seen in both OMDM and OSDM (Figure 3). Up-regulation of apoptosis was assessed by measuring protein levels of the initiator caspase 8 and the effector caspase 3. In the OMDM, a significant increase was found in active caspase 8 while both active caspase 3 and 8 were significantly increased in OSDM in NB21 hearts. This suggests that despite activation of competing MAP kinase signaling pathways in the NB21 offspring of diabetic mothers, the JNK signaling to promote apoptosis was more robust than ERK signaling.

The degree of maternal hyperglycemia during pregnancy greatly impacts the type of cardiomyopathy found in the newborn. Moderately well controlled hyperglycemia, as occurs with most human pregnancies, results in hypertrophy of the neonatal heart. When the hyperglycemia is poorly controlled and impacts maternal, and likely fetal, nutrition, a dilated cardiomyopathy results. Both the cardiomyopathies are transient, and in the rat, have largely resolved by three weeks of age. The similar activation of the MAPK signaling and apoptotic pathways in these two models suggests that these pathways may have a wide-ranging impact on the ability of the heart to remodel. Exploration of MAP kinase and apoptotic activity in the remodeling process that occurs with other cardiomyopathies may help to understand how the heart responds to other insults that impact cardiac structure and function.

BR, JS, and TS are all faculty in the Department of Pediatrics at the University of Iowa and members of the Program on Developmental Origins of Disease (PDOD). Among the interests of the PDOD are the short- and long-term consequences of being exposed to the adverse intrauterine environment caused by gestational diabetes mellitus.