Epidemiological studies have shown that infants exposed to an increased supply of nutrients before birth are at increased risk of type 2 diabetes in later life. We have investigated the hypothesis that fetal overnutrition results in reduced expression and phosphorylation of the cellular fuel sensor, AMP-activated kinase (AMPK) in liver and skeletal muscle before and after birth. From 115 days gestation, ewes were fed either at or ∼55% above maintenance energy requirements. Postmortem was performed on lamb fetuses at 139–141 days gestation (n = 14) and lambs at 30 days of postnatal age (n = 21), and liver and quadriceps muscle were collected at each time point. The expression of AMPKα1 and AMPKα2 mRNA was determined by quantitative RT-PCR (qRT-PCR). The abundance of AMPKα and phospho-AMPKα (P-AMPKα) was determined by Western blot analysis, and the proportion of the total AMPKα pool that was phosphorylated in each sample (%P-AMPKα) was determined. The ratio of AMPKα2 to AMPKα1 mRNA expression was lower in fetuses compared with lambs in both liver and muscle, independent of maternal nutrition. Hepatic %P-AMPKα was lower in both fetuses and lambs in the Overfed group and %P-AMPKα in the lamb liver was inversely related to plasma glucose concentrations in the first 24 h after birth (r = 0.73, P < 0.025). There was no effect of maternal overnutrition on total AMPKα or P-AMPKα abundance in liver or skeletal muscle. We have, therefore, demonstrated that AMPKα responds to signals of increased nutrient availability in the fetal liver. Suppression of hepatic AMPK phosphorylation may contribute to increased glucose production, and basal hyperglycemia, present in lambs of overfed ewes in early postnatal life.
the incidence of being either overweight (BMI > 25 kg/m2) or obese (BMI > 30 kg/m2) among U.S. adults is currently greater than 35% (3). Obesity, along with its comorbidities—insulin resistance, type 2 diabetes, and cardiovascular disease—incurs significant health care costs (3). Epidemiological studies have shown that infants exposed to an increased supply of nutrients before birth are at increased risk of becoming overweight or obese in later life (21, 26). In pregnancies complicated by maternal diabetes or glucose intolerance, maternal hyperglycemia results in increased fetal plasma glucose and insulin concentrations, and offspring from these pregnancies are at increased risk of obesity and type 2 diabetes in later life (25, 26). It has, therefore, been proposed that exposure to an excessive nutrient supply during critical windows of fetal development may result in permanent changes within tissues involved in peripheral insulin sensitivity to program glucose intolerance, insulin resistance, and an increase in body fat mass in later life (14, 22). We have previously demonstrated that maternal overnutrition in the sheep results in elevated plasma glucose concentrations and increased fat deposition in the offspring in early postnatal life (16). The mechanisms through which exposure to prenatal overnutrition can alter the regulation of glucose metabolism after birth are, however, unknown.
AMP-activated protein kinase (AMPK) is a metabolic master switch that senses changes in energy stores within cells and acts to maintain intracellular energy homeostasis (9). AMPK is activated by the upstream kinase LKB-1 in response to an increase in AMP concentrations within the cell (cellular energy depletion). Phosphorylated AMPK then coordinately regulates specific enzymes within key metabolic pathways in the cell to inhibit energy-consuming pathways, promote energy-generating pathways, and thereby restore cellular energy stores (9, 28). In hepatocytes, the activation of AMPK results in an increase in fatty acid oxidation and decreased activity of gluconeogenic enzymes. In the skeletal muscle, the activation of AMPK results in an increase in fatty acid oxidation and increased expression of the insulin-dependent glucose transporter, GLUT4, resulting in increased uptake of glucose by muscle cells. In adult humans with type 2 diabetes, the activation of AMPK during exercise is blunted compared with the response in healthy controls (27), and it has been suggested by a number of experts in the field that dysregulation of AMPK may contribute to the fasting hyperglycemia and reduced insulin sensitivity in these patients (7, 33). It is unknown, however, whether reduced phosphorylation of AMPK contributes to the development of type 2 diabetes in individuals exposed to an excessive supply of nutrients before birth. Two separate isoforms of the AMPK catalytic subunit have been described: AMPKα1 and AMPKα2. The AMPKα2 isoform predominates in skeletal muscle and liver in the adult and appears to assume a principal role in regulating whole body glucose metabolism (32). The respective role of these two isoforms in liver and skeletal muscle during prenatal and perinatal development are, however, unknown.
In the fetus, glucose utilization rate (GUR) by fetal tissues is directly related to fetal plasma glucose and insulin concentrations. Thus, the rate of glucose utilization increases with increasing fetal glucose supply (10, 11). It is unknown, however, whether AMPK plays a role in the regulation of cellular energy status before birth and whether the expression of AMPK mRNA and the proportion of the total AMPK pool that is phosphorylated is altered following exposure to hyperglycemia in utero. In the present study, we have, therefore, investigated the hypothesis that maternal overnutrition results in altered expression of AMPK mRNA and a reduced proportion of the total AMPK pool that is phosphorylated in liver and skeletal muscle before birth and that these changes are also present in the lambs of overnourished ewes in early postnatal life.
Animals and feeding.
All procedures were approved by the Adelaide University Animal Ethics Committee. Thirty Pregnant Merino ewes were used in this study.
From 90 days of gestation, ewes were acclimatized to a control diet, comprising 300 g of concentrated pellets [89% dry matter, metabolizable energy (ME) content 11.6 MJ/kg; Ridley Agriproducts Sheep Nut Ration, Murray Bridge, SA, Australia] and 1 kg Lucerne chaff (85% dry matter, ME content = 8.3 MJ/kg), designed to meet 100% energy requirements (MER) for maintaining pregnant ewes with singleton or twin fetuses, as appropriate (1). Ewes were randomly assigned to the control or Overfed group, and from 115 to 124 days gestation, they were provided with 6.5 ± 0.4 g pelleted concentrate and 14.0 ± 0.4 g chaff per kg body wt or 10.4 ± 0.7 g pelleted concentrate and 22.1 ± 0.8 g chaff/kg body wt, respectively. Every 10 days, feed allowances were increased (10%) to meet fetal nutrient needs during late gestation. Feed refusals were weighed daily for calculating the ME intake of ewes as a percentage of their MER. The ME intake of ewes in the Overfed group was significantly higher than the control group in both the fetal (control, 100% MER; Overfed, 155 ± 1.89% MER P < 0.01) and lamb (control, 91.9 ± 2.2% MER, Overfed, 133.4 ± 3.5% MER, P < 0.01) studies. (16, 18).
Feeding regime—lamb study.
Before lambing, food was provided to each ewe in two equal portions each day, one at 0900 and one at 1500, with water provided ad libitum. After lambing, all ewes were provided with the same diet, which consisted of 1 kg Lucerne chaff and 500 g pelleted concentrate once daily. If all feed was consumed before 1500, an additional 1 kg of chaff was provided. After birth, each ewe and her lamb(s) were housed in an individual pen in an indoor housing facility, which was maintained at a constant ambient temperature of between 20 and 22°C and a 12:12-h light-dark cycle.
Fetal surgery and blood sampling regimen.
In 14 pregnant ewes, surgery was performed between 103 and 113 days gestation (term = 147 ± 3 days) under aseptic conditions as previously described (4). General anesthesia was induced by intravenous injection of thiopental sodium (1.25 g iv, Pentothal, Rhone Merieux, Pinkenba, Queensland, Australia), maintained with 2.5–4% halothane (Fluothane, ICI, Melbourne, VIC, Australia) in oxygen. Vascular catheters were implanted in a jugular vein and carotid artery of the ewe and fetus, as well as in the amniotic cavity (4). During surgery, intramuscular injections of antibiotics (2 ml: 250 mg/ml procaine penicillin, 250 mg/ml dihydrostreptomycin, 20 mg/ml procaine hydrochloride; Lyppards, Adelaide, SA, Australia or 0.1 ml/kg liveweight terramycin 100, 100 mg/ml oxytetracycline hydrochloride; Pfizer, West Ryde, NSW, Australia) were administered to each ewe and fetus. All catheters were filled with heparinized saline, and the fetal catheters were exteriorized through an incision in the ewe's flank.
Before and after surgery, the ewes were housed in individual pens in animal holding rooms with a 12:12-h light-dark cycle. Ewes were allowed at least 3 days to recover from surgery prior to experimentation.
Between 116 and 139 days gestation, maternal (5.0 ml) and fetal (3.0 ml) arterial blood samples were collected three times per week prior to morning feeding at 0900. Blood samples were centrifuged at 1,500 g for 10 min at 4°C, and plasma was stored at −20°C for subsequent glucose and hormone assays. Fetal arterial blood (0.5 ml) was also collected three times per week for determination of fetal blood gases [Po2, Pco2, oxygen saturation (So2), pH, hematocrit, and hemoglobin] using an ABL 520 analyzer (Radiometer, Copenhagen, Denmark).
The remaining animals (n = 16 ewes) were allowed to lamb. Lambs (control, n = 6 male and n = 6 female; Overfed, n = 3 Male and n = 6 female) were born spontaneously at term (150 ± 3 days gestation) and were housed together with their mother throughout the duration of the experiment. The day of birth was designated as day 1. Birth weight (kg), crown rump length (cm), and abdominal circumference (cm) were recorded within 6 h of birth, for the first 5 days after birth, and every 3 days thereafter for the first 30 days of postnatal life. Blood samples were collected by subcutaneous venapuncture each day for the first 5 days of age, and every 3 days thereafter for the subsequent determination of plasma glucose, insulin, leptin, and free fatty acid concentrations.
Postmortem and tissue collection.
In the fetal study, control and Overfed ewes (control, n = 6; Overfed, n = 8) were killed between 138 and 141 days gestation by an overdose of pentobarbital sodium (Virbac, Peakhurst, NSW, Australia). Fetal sheep were delivered by hysterectomy and killed by decapitation.
Lambs (control, n = 9, Overfed, n = 12) were killed by overdose with sodium pentobarbital at 30 days of postnatal age. At postmortem, samples of fetal and lamb liver and skeletal muscle (quadriceps) were collected, snap frozen in liquid nitrogen, and stored at −80°C.
Plasma glucose, nonesterified fatty acid, insulin, and leptin assays.
Plasma glucose concentrations were measured by enzymatic analysis using hexokinase and glucose-6-phosphate dehydrogenase to measure the formation of NADH photometrically at 340 nm (COBAS MIRA automated analysis system, Roche Diagnostica, Basel, Switzerland). The sensitivity of the assay was 0.5 mmol/l and the intra-assay and interassay coefficients of variation were both <5%. Plasma nonesterified fatty acids (NEFAs) were measured by an in vitro enzymatic colorimetric method (Wako Pure Chemicals Industries, Osaka, Japan). The sensitivity of the assay was 0.25 meq/l and the intra-assay and interassay coefficients of variation were both <10%. Both assays have been previously validated for use in sheep plasma (4, 16).
Plasma insulin concentrations were measured using a radioimmunoassay (Rat insulin kit; Linco Research, St. Charles, MO), which was validated for use with sheep plasma (16). The sensitivity of the assay was 0.01 ng/ml, and the intra-assay and interassay coefficients of variance were both <10%.
Plasma leptin concentrations were measured using a competitive bovine leptin ELISA, which has previously been validated for sheep plasma (12). The sensitivity of the assay was 0.5 ng/ml and the intra-assay and inter-assay CVs were <16%.
Total RNA was extracted from fetal and lamb livers and skeletal muscle (quadriceps) in control and Overfed groups. Tissue samples (liver ∼30 mg; quadriceps ∼100 mg) were homogenized in 500 μl TRIzol reagent (Invitrogen Life Technologies, Carlsbad, CA), chloroform (100 μl) was added, and homogenates were shaken vigorously and incubated at room temperature (15 min). Homogenates were centrifuged (15 min) at 9,600 rpm at 4°C. The upper aqueous phase was collected, and 50% ethanol (∼400 μl) was added. Samples were added to RNeasy Mini Spin Columns (Qiagen, Doncaster, VIC, Australia) and centrifuged (15 s) at 10,500 rpm at room temperature, discarding flowthrough. Subsequent stages of extraction followed the standard Qiagen protocol. RNA concentration and quality were determined by spectrometry, and RNA integrity was confirmed by agarose gel electrophoresis.
RNA (5 μg) was mixed with dNTPs (Invitrogen Life Technologies) and random hexamers; then it was heated to 65°C (5 min). 5× First Strand Buffer (Invitrogen Life Technologies), 0.1 mol/l DTT (Invitrogen Life Technologies), and SuperScript III RT (Invitrogen Life Technologies) were added, and samples were reverse transcribed using PCR Express (Hybaid, Ashford, UK) with cycles programmed at 25°C (5 min), 50°C (50 min), and 85°C (15 min).
Quantitative real-time PCR.
The relative expression of AMPK α1, AMPK α2, and acidic ribosomal protein P0 (ARP-P0) transcripts was measured by quantitative real-time RT-PCR (qRT-PCR) using the SYBR Green system in an ABI Prism 7300 Sequence Detection System (PE Applied Biosystems, Foster City, CA). Primers used for analysis of gene expression are presented in Table 1. Primers for AMPKα1 and AMPKα2 were based on the human or rat sequence of the AMPK gene, which shares a 93% homology with the ovine sequence. The ARP-P0 primers have been validated previously for use in sheep tissues (17). All products were run on an ethidium bromide gel to confirm correct size of the amplicon, and all products were sequenced to confirm their identity. Reaction mixtures were prepared comprising 7 μl water, 3.5 μl forward primer, 3.5 μl reverse primer, 3.5 μl cDNA (1 nmol/l), and 17.5 μl 1× Power SYBR Green (Applied Biosystems). Triplicate aliquots (3 × 10 μl) of each sample were pipetted into 96-well optical plates (Applied Biosystems). At least six technical replicates were performed for each sample. The thermal cycle protocol used for amplification by the 7300 real-time PCR System (Applied Biosystems) included steps of 50°C (2 min), 95°C (10 min), 40 cycles of 95°C (15 s), and 60°C (1 min), followed by steps of 95°C (15 s), 60°C (30 s), and 95°C (15 s). Sample fluorescence was measured each cycle during the 60°C (1 min) step. At the end of each run, dissociation melt curves were obtained to confirm the presence of a single amplicon.
The abundance of each mRNA transcript was measured, and the expression relative to that of the reference gene, (ARP-P0), was calculated using Q-gene qRT-PCR analysis software, which provides a quantitative measure of the relative abundance of a specific transcript in different tissues by the comparative threshold cycle (Ct) method. The Ct value was taken as the lowest statistically significant [>10 standard deviation (SD)] increase in fluorescence above the background signal in an amplification reaction. This procedure takes into account any differences in the amplification efficiencies of the target and reference genes.
Quantification of AMPKα and P-AMPKα protein abundance.
Protein abundance of AMPKα and Phospho-AMPKα (P-AMPKα in liver and quadriceps before and after birth) was determined by Western blot analysis. Protein was extracted from samples (∼50–80 mg) of liver and quadriceps muscle. Tissue was homogenized in 500 μl of ice-cold homogenizing buffer (8). Homogenates were centrifuged (30 min) at 13,200 rpm at 4°C. Supernatant was collected, and protein content was determined using Bradford microassays [standard, bovine-γ-globulin (Bio-Rad Laboratories, Hercules, CA)].
Each liver and muscle homogenate were mixed with 17.5 μl 1× SDS sample loading buffer (Invitrogen Life Technologies) and 7 μl reducing agent (Invitrogen Life Technologies). Protein from liver (50 μg) and quadriceps (80 μg) samples from control and Overfed groups were separated by gel electrophoresis utilizing 10% NuPAGE bis-Tris Gels (Invitrogen Life Technologies). Proteins were transferred to nitrocellulose membranes that were probed using anti-sera raised against total AMPKα and P-AMPKα (Cell Signaling Technology, Boston, MA). Using 5% BSA-TBS/T [BSA, Tris buffered saline (TBS)/0.1% Tween-20 (T)]; membranes were blocked for 70 min (room temperature) with gentle agitation. Membranes were washed in TBS/T with gentle agitation. Primary antibodies (Cell Signaling Technology) were diluted (1:1,000) in either TBS/T (monoclonal anti-P-AMPKα) or 5% BSA-TBS/T (polyclonal anti-AMPKα). Membranes were incubated with primary antibodies overnight (4°C), then for 60 min (room temperature), washed in TBS/T, and incubated 90 min (room temperature) with secondary antibody (goat anti-rabbit alkaline phosphatase-conjugated) (Cell Signaling Technology) diluted (1:2,000) in 5% BSA-TBS/T, and washed in TBS/T.
Immunoreactive proteins were detected by electrochemifluorescence (ECF; Amersham, Piscataway, NJ) and were scanned using Typhoon (Amersham). Densities of bands were quantified using ImageQuant (ver. 5.2, Molecular Dynamics, San Diego, CA) and were normalized to a positive control sample (lamb liver protein; 40 μg). Normalized data were expressed as the relative amounts of total AMPKα and P-AMPKα. The relative proportion of total AMPKα in the phosphorylated form was calculated as Phospho-AMPKα/total AMPKα (%P-AMPKα). This ratio provided a measure of the proportion of AMPKα in the phosphorylated form in the basal (fasted) state for each experimental animal and, therefore, a measure of baseline AMPKα activity. The phosphorylation of AMPK has previously been shown to provide a measure of the extent of AMPK activation (20, 29).
All data are presented as means ± SE. Multifactorial ANOVA with repeated measures were used to determine the effect of maternal overnutrition on plasma hormone and metabolite concentrations. Two-way ANOVA was used to determine the effect of maternal nutrition (control or Overfed), development (fetal vs. postnatal), and their interaction on AMPKα1 mRNA and AMPKα2 mRNA expression, the abundance of total AMPKα, P-AMPKα, and the proportion of phosphorylated AMPKα in liver and skeletal muscle samples. Relationships between variables were determined by simple linear regression analyses. All statistics were performed using the Statistical Package for Social Scientists (SPSS) (ver. 13.01). A probability of 5% (P < 0.05) was considered statistically significant in all analyses.
Plasma glucose, insulin, nonesterified fatty acid, and leptin concentrations.
Maternal glucose (3.53 ± 0.15 vs. 2.98 ± 0.04 mmol/l, P < 0.05) and leptin (8.8 ± 1.4 vs. 5.7 ± 0.8 ng/ml, P < 0.05) concentrations were higher in overfed ewes compared with controls. Mean fetal plasma concentrations of glucose and insulin in late gestation were significantly higher in the Overfed group compared with controls (Table 2). There was no effect of maternal overnutrition on the mean concentrations of leptin in the fetal circulation in late gestation (Table 2). The mean plasma glucose concentration during the first 4 wk of postnatal life was significantly greater in the lambs of Overfed ewes compared with controls. There was, however, no effect of maternal overnutrition in late gestation on the mean concentration of insulin, NEFA, or leptin in the lamb in early postnatal life (Table 2) (16, 18).
There was no effect of maternal nutritional treatment on fetal weight at 139–141 days gestation (4.83 ± 0.24 vs. 4.49 ± 0.27 kg; P > 0.05). Birth weight (5.08 ± 0.20 vs. 5.10 ± 0.23 kg; P > 0.05) and weight at 30 days of postnatal life (13.49 ± 0.77 vs. 13.61 ± 0.49 kg; P > 0.05) were also not different between lambs of control and Overfed ewes.
The impact of prenatal overnutrition on hepatic AMPKα expression before and after birth.
There was no effect of maternal overnutrition on the total or relative weight of the liver before birth (Total liver weight, 125.3 ± 8.64 vs. 126.3 ± 13.64 g; relative liver weight, 26.0 ± 1.4 vs. 28.0 ± 2.3 g/kg). The relative weight of the liver was higher in lambs of Overfed compared with control ewes at 30 days of postnatal age (21.5 ± 0.6 vs. 19.4 ± 0.6 g/kg, P < 0.05).
The relative expression of the α1 isoform of AMPK was lower (P < 0.05), and the expression of the α2 isoform was higher (P < 0.05), in the postnatal lamb compared with the fetus, independent of the level of maternal nutrition in late gestation (Fig. 1). There was no effect of maternal overnutrition on the expression of AMPKα1 or AMPKα2.
In the postnatal lamb, but not in the fetus, there was a direct relationship between the relative mRNA expression of the AMPK α1 isoform and mean plasma insulin concentrations in the week preceding tissue collection when data from the control and Overfed groups were combined (r = 0.73, P < 0.001).
The impact of prenatal overnutrition on hepatic AMPKα abundance and phosphorylation before and after birth.
There was no effect of maternal nutrition or developmental age (fetus vs. postnatal) on the abundance of either total AMPKα or P-AMPKα in the liver (Fig. 2, A and B). There was, however, a significant effect of maternal overnutrition on the proportion of Phospho-AMPKα (%P-AMPKα; an index of AMPK activation), which was lower in the Overfed compared with the control group both before and after birth (Fig. 2C). There was no difference between the fetus and the postnatal lamb in the proportion of the total AMPK pool, which was phosphorylated.
In the Overfed group, there was a significant inverse relationship between plasma glucose concentrations present in the first 24 h after birth and hepatic %P-AMPKα on postnatal day 30 (r = 0.74, P < 0.05; Fig. 3). This relationship was not present in the control group.
The impact of prenatal overnutrition on AMPKα expression in skeletal muscle before and after birth.
There was no significant effect of maternal overnutrition on the expression of either the AMPKα1 or AMPKα2 isoform in the skeletal muscle (Fig. 4, A and B). As in the liver, the expression of the α1 isoform was higher in fetal compared with postnatal skeletal muscle, while the expression of the α2 isoform was higher in the muscle of the postnatal lamb compared with the fetal sheep (Fig. 4, A and B), and this effect was independent of maternal nutrition.
The impact of prenatal overnutrition on AMPKα abundance and phosphorylation in skeletal muscle before and after birth.
There was no effect of either maternal nutrition or developmental age (fetus vs. postnatal) on the abundance of AMPKα or P-AMPKα (Fig. 5B) in skeletal muscle. The %P-AMPKα in skeletal muscle was higher in the fetus compared with the postnatal lamb independent of maternal nutrition (Fig. 5C).
In the postnatal lamb, but not in the fetus, the total AMPKα abundance was inversely related to plasma insulin concentrations on the day of tissue collection (r = 0.43, P < 0.001). There was also a direct relationship between the %P-AMPK and plasma leptin concentrations in postnatal week 3 in the Overfed group, but not in the controls (Fig. 6).
To our knowledge, this is the first study to demonstrate the presence of AMPK mRNA and protein in fetal tissues in a species in which the systems regulating glucose homeostasis develop before birth, as in the human. We have shown that there is a significant increase in the ratio of AMPKα2 to AMPKα1 mRNA from fetal to postnatal life in both the liver and skeletal muscle, as a consequence of a decrease in AMPKα1 and an increase in AMPKα2 mRNA expression between fetal and postnatal life. We have also reported that the proportion of AMPKα in the phosphorylated form in the liver is decreased in response to maternal overnutrition before birth and that this suppression is also present in the lamb of overfed ewes at 30 days of postnatal age. Thus, these findings suggest that increased availability of glucose in utero may have the capacity to regulate the phosphorylation of AMPK in the liver from before birth.
Developmental regulation of AMPK α1 and α2 expression.
The relative abundance of mRNA for the two isoforms of the AMPK catalytic subunit (α1 and α2) in liver and skeletal muscle was subject to developmental regulation, as represented by the significant increase in the abundance of AMPKα2 relative to that of AMPKα1 between fetal and early postnatal life in both tissues. This was a consequence of an increase in the abundance of AMPKα2 mRNA and a decrease in the abundance of AMPKα1 mRNA during this period, and this suggests that there may be a reciprocal relationship between the relative expression of the two isoforms of the catalytic subunit over the course of development. This is consistent with the results of studies in adult rodents, in which AMPKα2 null mutant mice have been shown to have an increased abundance of AMPKα1 in skeletal muscle compared with wild-type mice (32).
In adults, the α2 isoform is predominant in both skeletal muscle and liver, and global and tissue-specific AMPKα2 null mutant mice exhibit a phenotype consistent with severe dysregulation of metabolic function, including fasting hyperglycemia, basal increases in gluconeogenesis, glucose intolerance, insulin resistance, and elevated free fatty acid concentrations in both the fed and fasted states (31, 32). Knockout of the α2 isoform is also associated with a defect in glycogen synthesis in skeletal muscle and altered function of the autonomic nervous system (19). In contrast, AMPKα1 null mutant mice are essentially normal in their metabolic phenotype (31). This has led to the suggestion that the α2 catalytic subunit has a more important role in the regulation of glucose metabolism in liver and skeletal muscle, compared with AMPKα1 (32).
It is, therefore, possible that the increase in the relative abundance of AMPKα2 may reflect a developmental maturation of the systems regulating glucose homeostasis and metabolism between fetal and postnatal life. The basis of this developmental switch in the abundance of AMPK is not known but may be related to the prepartum increase in circulating cortisol in the fetus, which acts to stimulate the maturation of physiological and metabolic systems required for a successful transition to extrauterine life, including the upregulation of hepatic gluconeogenic enzymes (6). An alternate possibility may be that the higher plasma glucose and insulin concentrations present in postnatal compared with prenatal life (18) may act to increase the expression of AMPKα2 mRNA.
Maternal overnutrition and the suppression of hepatic AMPKα activation.
Maternal overnutrition resulted in a decrease in the proportion of AMPK in the liver present in the phosphorylated form both before and after birth. Kraegen et al. (13) have demonstrated that AMPK activity in vivo was significantly suppressed in response to an increase in plasma glucose concentrations. In the fetus, GUR in fetal tissues is directly related to fetal plasma glucose and insulin concentrations, and GUR therefore increases with increasing fetal glucose supply (10, 11). It is possible that the increased glucose utilization in the presence of sustained hyperglycemia is not counteracted by a similar increase in the metabolic rate of the hepatocyte and that this results in increased ATP production, a reduced cellular AMP:ATP ratio, and a consequent reduction in the total hepatic AMPK pool that is phosphorylated. This is supported, in part, by the demonstration of an inverse relationship between the proportion of AMPK in the phosphorylated form in the liver and plasma glucose concentrations in the lambs in the Overfed group. The reason for this decrease in AMPK phosphorylation is not completely clear. Kraegen et al. (13) have reported that glucose infusion decreased AMPK activity in the liver of the adult rat, in the absence of any change to the cellular AMP:ATP ratio, which suggested an alternative pathway may be involved in the regulation of AMPK phosphorylation in this physiological state. One possibility is that higher insulin concentrations, which were present in the fetuses of overfed ewes, may also contribute to the suppression of AMPK phosphorylation and that an increase in cellular glucose uptake, mediated by an increased availability of insulin, may result in an in increase in cellular energy stores and enhanced turnover of AMPK protein. It is also possible that insulin may directly contribute to the suppression of AMPKα phosphorylation, since insulin has been shown to inhibit AMPK phosphorylation in vitro in the heart (5). An alternate possibility is that exposure to increased glucose results in decreased activity of upstream regulators of AMPK phosphorylation, such as LKB-1 (24), which thereby reduces the phosphorylation of the AMPK molecule.
Interestingly, we have demonstrated that hepatic AMPK phosphorylation was also lower in lambs of overfed ewes at 1 mo of postnatal age. While the inverse relationship between the proportion of phosphorylated AMPK in the lamb liver and plasma glucose concentrations was strongest for glucose concentrations present in the first 24 h after birth, it is not possible to conclude whether there is a stronger influence of prenatal, compared with postnatal glucose on hepatic AMPK activation, given that lambs of overfed ewes have higher glucose concentrations than control counterparts for the first month after delivery (16). It is the case, however, that a reduction in the proportion of the total AMPK pool that is phosphorylated would potentially block insulin suppression of glucose release from the liver, thereby increasing hepatic glucose output and contributing to the elevated basal glucose levels during the early postnatal period. The regulation of hepatic glucose production by AMPKα is mediated via the activation of downstream targets of kinase activity, including glucose-6-phosphatase and phosphoenolpyruvate (23), and further studies that explore the relationship between baseline AMPKα phosphorylation and the expression and activity of these gluconeogenic enzymes in offspring exposed to prenatal overnutrition are clearly warranted.
Maternal overnutrition and the expression and activation of AMPKα in skeletal muscle.
In this study, maternal overnutrition did not alter the abundance of AMPK or the proportion of phosphorylated AMPK in the skeletal muscle before or after birth. The total abundance of AMPK in the skeletal muscle was, however, inversely related to plasma insulin concentrations on the day of tissue collection, which may suggest that an increase in insulin-mediated glucose uptake by the muscle results in an enhanced turnover of AMPK protein. Interestingly, the proportion of phosphorylated AMPK in the skeletal muscle of lambs exposed to maternal overnutrition in utero was directly related to plasma leptin concentrations in early postnatal life, which suggests that leptin may have a role in inducing phosphorylation of AMPK in the muscle of this group in early postnatal life, as in the adult (15, 30). The different response of AMPK in liver and muscle to prenatal overnutrition may suggest that there is a difference in nutrient portioning to these two tissues during fetal life. The liver is the first organ to receive nutrient-rich blood from the umbilical vein during fetal life (2), and it may be that the fetal liver is exposed to higher concentrations of glucose from the nutrient-rich blood of the overnourished pregnant ewe compared with more peripheral organs, such as the skeletal muscle, in the same fetuses. This could, therefore, result in more pronounced increases in cellular energy content in the liver compared with the muscle, and the precipitation of greater changes in the activity of cellular fuel sensors, such as AMPK and its upstream regulators.
We have, therefore, demonstrated that both mRNA and protein for the metabolic master switch, AMPK, are present in liver and skeletal muscle before birth in the sheep. In the present study, we found that the proportion of phosphorylated AMPK protein (an indicator of hepatic AMPK activation) was reduced both before and after birth in response to prenatal exposure to maternal overnutrition in late gestation. This provides evidence that signals of increased nutrient availability can regulate AMPK phosphorylation by at least 141 days gestation in the sheep and that a reduction in the proportion of the total hepatic phosphorylated AMPK pool may result in reduced AMPK activity within this tissue, which, in turn, may contribute to increased hepatic glucose production and basal hyperglycemia, present in lambs of overfed ewes in early postnatal life.
This work was supported by an National Health and Medical Research Council of Australia Program Grant to I. C. McMillen.
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.
- Copyright © 2008 the American Physiological Society