|Home | About | Journals | Submit | Contact Us | Français|
The impact of maternal nutrient restriction during early-to-mid gestation, a period coinciding with early fetal brain development, on appetite regulation and energy balance in the offspring following juvenile obesity was examined. Pregnant sheep were either fed to fully meet their nutritional requirements throughout gestation or 50% of this amount between 30-80 days gestation. Following weaning, offspring were either made obese through exposure to a sedentary obesogenic environment or remained lean. Maternal nutrient restriction had no effect on birth weight or subsequent growth. At one week of age, only, gene expression for neuropeptide Y in the hypothalamus was reduced in nutrient restricted offspring. By 1 year of age, all obese animals had raised plasma leptin, non-esterified fatty acids and insulin, with the latter effect amplified in nutrient restricted offspring. Fasting plasma glucose, triglycerides and cortisol were unaffected by obesity. The entrained reduction in physical activity that led to obesity persisted when all animals were maintained within individual pens. Obese nutrient restricted offspring, however, exhibited reduced daily food intake and were, therefore, no longer in positive “energy balance”. This adaptation was accompanied by elevated hypothalamic gene expression for the melanocortin-4 and insulin receptors, AMP-activated kinase and acetyl CoA carboxylase α. In conclusion, nutrient restriction specifically targeted over the period of early fetal brain development, contributes to a profoundly different adaptation in energy balance following juvenile obesity. The extent to which this adaptive response may benefit the offspring or result in an exacerbated risk for type II diabetes remains to be established.
The incidence of obesity (BMI >30 kg/m2) in the UK population and worldwide has increased markedly over the past decade (1, 2). Many epidemiological and prospective clinical studies in humans, together with experimental findings in animals, have demonstrated that modifications of the nutritional environment during fetal development can have a substantial impact on later health and disease, including the onset of obesity (3, 4). The extent to which, modifications in the maternal diet during early fetal development may contribute to changes in energy balance so as to modulate the effect of obesity remains unknown. Increasing maternal food intake to ad libitum during late pregnancy in sheep results in an upregulation in gene expression of pro-opiomelanocortin (POMC) at 30 days of postnatal age (5) but this adaptation is not accompanied by any persistent change in food intake. Indeed, it has no effect on the orexigenic neuropeptides neuropeptide Y (NPY) and agouti related peptide (AgRP) nor on the leptin receptor (ObR) (5) that normally mediates the effects of leptin on food intake (6). An increase in maternal food intake in late gestation was, therefore, unlikely to be associated with additional stimulation of energy expenditure through regulation of melanocortin receptors type 3 and 4 (MC3/4 R) (5). Conversely, severe maternal undernutrition in rats that is sufficient to induce intra-uterine growth retardation results in a reduction in physical activity in the resulting offspring, although they remain smaller than controls (7). To date, however, there are no studies that have examined the extent by which appetite regulation may be reset in the adult following exposure to nutrient restriction in utero, when this is followed by exposure to an obesogenic environment after birth.
Insulin has an important role in appetite regulation as it crosses the blood brain barrier in direct proportion to its circulating concentration (8) where it has a direct effect on a range of appetite pathways. These are mediated through the insulin receptor (IR) and the IR substrate type 2 (9) resulting in impaired gene expression and activity of NPY within the arcuate (ARC) and the paraventricular (PVN) nuclei of the hypothalamus (10). In the adult, plasma insulin concentration can be programmed by changes in the maternal diet during pregnancy (11) although it is has yet to be determined whether this impacts on appetite regulation or energy balance.
Other factors that are important in the regulation of energy balance include AMP-activated kinase (AMPK) and its co-factor acetyl CoA carboxylase (ACC)α (12, 13) that are both responsive to leptin (12). In addition, glucose transporter (GLUT)-1 regulates glucose transport in appetite responsive neurons (14) and the peroxisome proliferator-activated receptor gamma co-activator (PGC1)α determines fatty acid metabolism in the brain and may control energy expenditure through promoting physical activity (15). It remains unknown, however, whether the abundance of these key hypothalamic genes are modulated following obesity or whether the magnitude of response is reset in offspring exposed to maternal nutrient restriction during early development.
We have previously established in the sheep that maternal nutrient restriction, targeted over the period of the maximal placental growth (i.e. 30-80 days; term ~145 days), results in offspring with increased adiposity at birth and elevated tissue glucocorticoid sensitivity (16). In this regard, enhanced mRNA abundance, for both the glucocorticoid receptor (GR) and 11ß-hydroxysteroid dehydrogenase type 1, persists for up to 6 months of age (17), although fat mass is unaffected when the offspring are reared in an outdoor “free-living” environment. The period between early (30 days) to mid (80 days) gestation in the sheep represents a critical time for fetal organogenesis including brain development. This is likely to include the development of the appetite control network and energy sensing pathways in the hypothalamus (18, 19), although the extent to which such development may extend into, and beyond, the postnatal period is currently unknown. Maternal nutrient restriction between early-to-mid gestation reduces the size of the fetal brain (20) and is associated with behavioral adaptations in the resulting adult (21). In contrast to the effects on glucocorticoid action in peripheral tissues, no effects are seen in the hypothalamus or pituitary in the newborn (22).
Dysregulation of energy balance is one of the main features of obesity (23). No study to date however, has assessed its long term outcomes in adults born to nutrient restricted mothers following adolescent-onset obesity. We therefore examined the hypothalamic consequences in the resulting offspring both in the immediate postnatal period as well as at 1 year of age following exposure to obesogenic environment after weaning. In particular, we aimed to examine whether the previously established beneficial effects on the kidney with regard to protection against glomerulerosclerosis that occur when previously nutrient restricted offspring become obese (24) could extend to the regulation of energy balance.
All procedures were performed in accordance with the UK Animals (Scientific Procedures) Act, 1986 and approved by the local ethics committee of the University of Nottingham. At day 30 of gestation, 22 twin-bearing Welsh Mountain sheep were randomly allocated to receive either a control (C; 7MJ/day: n=14) or nutrient restricted (NR; 50% of C, n=8) diet until day 80 of gestation. Thereafter, all sheep were fed 100% of their calculated metabolisable energy requirements to term (12-13 MJ/day near term) (25). Offspring were delivered spontaneously at term (~147 days). The ratio of females to males was (1:1) in the control and (2:1) in the NR group and all animals remained intact. Seven control and 8 of the nutrient restricted offspring were tissue sampled at 7 days of age. The remaining offspring (C n=14; NR, n=8) were reared by their mothers up to weaning (~10 weeks of age). During lactation all mothers were fed a diet of hay ad libitum together with a fixed amount of concentrate pellets sufficient to fully meet their own metabolisable requirements, together with those required to maintain lactation. From weaning to 12 months of age, 7 of the gestational control fed sheep and all (n=8) nutrient restricted offspring were exposed to an ‘obesogenic’ environment: i.e. group-housed in a barn at a stocking rate of 17 animals per 50m2 with ad libitum access to hay and concentrate pellets, as described previously (24). The remaining 7 control offspring were reared in a non-obesogenic ‘free-living’ environment: i.e. out to pasture at a stocking rate of 17 animals per 3000 m2 with free grazing allowance and a fixed amount of concentrate pellets offered (400 g/day). Seven day old offspring are thus classified with regard to the maternal environment: i.e. control or nutrient restricted. The remaining post weaning “free-living” offspring were defined as lean (L, n=7). Offspring raised in the obesogenic environment were defined as obese (O, n=7) or nutrient restricted obese (NRO, n=8) as previously described (24). The number of animals analyzed in this study differed from the total number of offspring born because some of the controls were used in an independent study.
The level of spontaneous physical activity was determined by an uniaxial accelerometer (Actiwatch; Linton Instrumentation, Diss, UK). These probes were placed on collars around the neck of the animal for a 24 h period, with a measurement of activity being recorded 32 times per second. This resulted in 960 measurements per 30 seconds that were automatically summated into 30 second time bins. The resulting 2880 readings for the 24 h period (arbitrary accelerometry units) were then downloaded to specialist software (Actigraph) for imaging and analysis. Spontaneous physical activity was recorded in all offspring at 12 months of age when housed individually in similar UK Home Office designated floor pens (3m2). This allows an estimation and comparison of energy expenditure related to physical activity between experimental groups (26).
Also at 12 months of age when all offspring were housed individually daily food intake was recorded in order to assess appetite regulation. For each animal, daily energy intake was assessed through weighed intake and food refusal i.e. sheep were offered sufficient energy for the 24 h period based on a mix of low (hay; 8.9 MJ/kg) and high (concentrate pellets; 12.6 MJ/kg)) energy dense food.
From the all the individual values of food intake and physical activity we calculated relative values taking the average lean group as a reference (100%):
Then, we determined a ratio of food intake to physical activity:
Therefore: A relative ratio can be determined:
Thus, the average relative ration for lean offspring is considered to be equal to 0. A positive value represents net gain in energy, whereas a negative value represents a net loss.
Morning (~ 08.00 h) fasted blood samples (5 ml) were collected from the jugular vein into lithium heparin and K+EDTA tubes at 12 months of age. Plasma was recovered by centrifugation (3000 rpm × 10 min at 4°C) and kept frozen at -20°C until analysis. Plasma leptin, ACTH (Diasorin, Slough, UK) and cortisol (Coat-a-Count, Euro DPC, Caernarfon, UK) were measured by radioimmunoassay as previously described (27, 28). Plasma insulin was assessed by ELISA assay (Mercodia, Sheep insulin ELISA kit, UK). Plasma glucose (Randox GPO-PAP, Randox, Crumlin UK), non-esterified fatty acid (NEFA) (wako NEFA-c Kit, Wako Chemicals Gmbh, Neuss, Germany) and triglyceride (TG) (Randox GOD-PAP, Randox, Crumlin UK) were assessed by spectrophotometic methods.
An intravenous insulin tolerance test was used to assess insulin-dependent glucose uptake in adult sheep at 1 year of age. Briefly, after an overnight fast, 0.75 IU.kg-1 human insulin (Novorapid®, Novo Nordisk, Crawley, UK) was administrated as a rapid intravenous (jugular vein) bolus at time 0. Venous blood samples were collected into K+-EDTA treated blood tubes every 2 minutes for 16 minutes. The percentage (%) of decline per minute (Kitt, in % glucose decline/kg/min) in plasma glucose from time 0 min was taken to represent peripheral tissue insulin sensitivity.
All the animals were euthanized in the morning between 09.00 and 11.00 h. The 1 year old offspring were euthanized as previously described after an overnight fast (24). With regard to the sampling of 7 day old offspring they could not be subjected to a similar period of food withdrawal as this would have meant prolonged physical separation from their mothers and would have thus been unethical. To avoid any additional handling stress in these animals prior to euthanasia blood samples were not taken from these offspring. Each animal’s brain was immediately recovered and the entire hypothalamus dissected and snap frozen in liquid nitrogen, before being stored at -80°C. The hypothalamus was dissected using the rostral border of the optic chiasm and the caudal border of the mamillary bodies as landmarks on the ventral surface and the lateral hypothalamic sulci as the lateral boundaries. In addition, all visceral fat depots were dissected and weighed.
Total RNA from the entire hypothalamus was extracted using Qiazol® lysis reagent and RNeasy® extraction kit (Qiagen Ltd., Crawley, UK). Total RNA samples were DNAse treated (Promega RQ1 endonuclease, Promega Ltd, Southampton, UK) and reverse transcribed (Superscript® II reverse transcriptase, Invitrogen Ltd., Pasley, UK). cDNA was amplified using a real-time thermocycler (Quantica®, Techne incorporated, Barloword Scientific Ltd, Stone, UK) using Quantitect® SYBR green PCR kit (Qiagen Ltd., Crawley, UK) (31). Gene expression data (mRNA) were determined for: ACCα, AgRP, AMPK, GR, GLUT-1, INS-R, MC3/4-R, NPY and its type 1 receptor (NPY1R), PGC1α, POMC and the Ob-R. Primer sequences and reaction conditions for each gene examined are detailed in Table 1. Prior to real-time PCR, the PCR product of each primer set was gel extracted (QIAquick Gel Extraction Kit, Qiagen Ltd., Crawley, UK) and sequenced in order to confirm the specificity of the product and then to develop the necessary standards for real-time PCR analysis. 18S rRNA was used as a housekeeping gene and results calculated using the 2-ΔCT method (32). Gene expressions from 7 days old control offspring were used as a relative reference for all analyses.
Data were analysed using SPSS v14.0 for Windows (SPSS Inc., Chicago, USA). Data were normally distributed and were, therefore, analysed by using one-way ANOVA with Bonferroni Post-Hoc test. The effect of postnatal age on hypothalamic gene expression between C (7 days old) and L (1 year old) offspring were assessed by unpaired Student t test. Data are reported as mean ± S.E.M, a P value < 0.05 was taken to represent a statistically significant difference.
At birth, and at 7 days of age, no morphometric differences were observed between control and nutrient restricted offspring (Table 2A). There were no differences in growth rates between groups until after 3 months of age and exposure to the post-weaning obesogenic environment. By 1 year of age, all obese animals were heavier and fatter than those maintained in a “free-living” environment (Table 2B). In addition, the obese group were significantly longer (i.e. increased crown-to-rump length) and taller (i.e. increased shoulder height) than lean animals.
The average daily caloric intake for obese and lean animals was similar at 1 year of age, but this was reduced in the nutrient restricted obese offspring (Figure 1). Total physical activity in all obese animals was ~50% lower than in the lean group. This reflected the difference in activity between those animals kept in a “free-living” environment as opposed to being maintained within an obesogenic environment (24). As a consequence relative ratio of daily energy intake to physical activity was significantly greater in obese compared to lean animals. However, the positive effect of obesity on this ratio was no longer apparent in obese offspring born to nutrient restricted mothers (Table 3).
At 1 year of age, after 9 months exposure to an obesogenic environment, basal plasma glucose, TG, cortisol and ACTH were all similar in obese and lean animals (Table 3). Plasma leptin and NEFA were both significantly raised with obesity, irrespective of maternal diet. In addition, although plasma insulin was increased in all obese animals, this adaptation was amplified in offspring born to nutrient restricted mothers.
The decline of plasma glucose (Kitt, %/min/kg) following IVTT was lower in obese relative to lean animals (Table 3) but prenatal nutrient restriction did not exacerbate this effect
Seven days after birth, the only hypothalamic gene examined that was decreased by maternal nutrition was NPY (Table 4). Between 7 days and 1 year of age, gene expression for all orexigenic factors increased as did mRNA abundance for the anorexigenic factors POMC and MC3R, although none were affected by the prenatal nutritional environment. In contrast, gene expression for NPY1R, GR and MC4R remained unchanged with age, but the latter was significantly upregulated in obese offspring born to nutrient restricted mothers.
Gene expression for the metabolic and energy sensing factors IR, PGC1α and ACCα all increased with age, whereas GLUT-1 and AMPK were unchanged. Obesity resulted in upregulation of IR and PGC1α gene expression, adaptations that were accompanied with increased mRNA abundance for both AMPK and ACCα, but only in the nutrient restricted group.
To further evaluate the magnitude of leptin resistance within the hypothalamus for each animal, plasma leptin concentration was expressed relative to mRNA abundance of AgRP and NPY. Following this analysis it was apparent that both were significantly elevated by obesity (Figure 2).
We have shown that, in sheep, exposure to an obesogenic environment i.e. a simple reduction in the available living-space results in pronounced obesity, by one year of age. Importantly this reduction in activity persists when the animal is moved to a more confined living environment although food intake does not necessarily adapt accordingly. The obese phenotype we observed in this study is in accord with that seen in humans namely, raised plasma leptin and NEFA (33, 34). Furthermore, the lack of any change in gene expression for NPY or AgRP within the hypothalamus emphasises that all obese animals were leptin resistant (35). Interestingly, offspring born to nutrient restricted mothers exhibited an appreciable reduction in daily food intake when individually housed that was not seen in those obese animals born to control fed mothers. This adaptation was accompanied by raised plasma insulin that may have been the trigger for the adaptive response seen in the hypothalamus of those offspring born to nutrient restricted mothers.
In an environment with restrained physical activity, offspring born to nutrient restricted mothers maintained their mean daily food intake in accord with estimated metabolic demands with the result that they were in neutral energy balance. Obese offspring born to control fed mothers, however, failed to adapt in the same way. Unexpectedly, this period of energy restriction of the mother and fetus was not related to any major differences in gene expression within the hypothalamus as measured soon after birth. At this age, the only significant difference related to the prenatal environment was a downregulation in NPY in the nutrient restricted offspring. This transient adaptation was not, however, associated with any difference in early postnatal growth rate. The only other study during fetal development in which maternal nutrition had an effect on NPY was following a 50% reduction in food intake during the final month of gestation where fetal plasma glucose and insulin decreased (29). In accord with our findings, changes in NPY mRNA abundance around the time of birth were not accompanied by any major changes in postnatal and juvenile growth (36, 37). In the present study, the materno-fetal endocrine environment may have been modified after 80 days of gestation following restoration of the maternal diet to the same amount as controls. For example, this results in a rise in maternal plasma cortisol, thus restoring its concentration to the same as in control fed mothers (16) with no effect, however, on maternal plasma glucose concentrations (16). Taken together, these findings support the hypothesis that the NPY/AgRP system is sensitive to changes in energy supply (5) although the primary factors regulating fetal NPY gene expression between mid and late gestation remain to be determined.
It should be noted that the expression of a majority of hypothalamic genes examined were much lower at 1 week compared with 1 year of age. Notable exceptions were the anorexogenic factors MC4 and GR. Interestingly, MC3R was found to be 100 times lower at 1 week, than at 1 year of age. These findings are in accord with those in the rat in which MC4R increases in the prenatal period, whereas MC3R expression, only appears near to the time of weaning (38). The ontogeny of GR gene expression we observed in the hypothalamus is in accord with that seen in the liver between birth and early adulthood (39). Relatively low expression of genes involved in the regulation of the appetite within the hypothalamus of the newborn suggests that the central control of appetite is much less important in the pre- and early postnatal period when the drive to feed is crucial to survival. In this regard, during early lactation, the interaction between the offspring and its mother and/or the competition for food with its siblings (40) may well have a much greater impact on feeding than hypothalamic signaling alone. This would explain why differences seen in hypothalamic gene expression of appetite controllers in the early postnatal period are not related to persistent differences in milk intake (5).
One particularly interesting outcome of our study was that offspring born to nutrient restricted mothers were better adapted to maintain a neutral “energy balance” after they had become obese. This was despite the fact that they were markedly leptin resistant and would not, therefore, be predicted to show any difference in food intake compared to the obese offspring born to control fed mothers. Nevertheless, the control of food intake is a complex process based on the integration of a range of signals (41). It has been suggested that the energy sensing pathway and insulin responsiveness coordinate with leptin to regulate appetite and energy expenditure. For example, AMPK and ACCα have been found to directly interact with leptin (12) and adiponectin (42) to control hypothalamic NPY and POMC neuronal activity, in conjunction with the regulation of energy sensing (43). In addition, insulin acting through its hypothalamic receptor induces an anorexigenic signal (8, 10). It could be hypothesized that a combination of elevated plasma insulin in conjunction with up-regulation of the IR and of the energy sensing system (i.e. AMPK and ACCα) act synergistically to promote satiation. As a consequence, offspring born to nutrient restricted mothers showed a pronounced reduction in food intake as measured when individually housed and were able to maintain a neutral “energy balance”. A response of this type would be further amplified by the increased gene expression of MC4R that would be predicted to reduce food intake and stimulate energy expenditure (41).
To date, the mechanisms underlying developmental programming of appetite and insulin homeostasis remain unclear. In our study we observed a long term effect of maternal nutrient restriction, resulting in elevated plasma insulin concentrations. Interestingly, a similar up-regulation of resting plasma insulin has been found in nutrient restricted rats when habituated onto a high energy diet (44). Surprisingly, this hyperinsulinemia was abolished when offspring were treated with leptin in the early post-natal period (44) that in the rat coincides with maturation of the hypothalamic-pituitary axis. One question that arises is whether the hyperinsulinemia observed in our offspring born to mothers nutrient restricted at a much earlier time of brain development could be related to low leptin concentrations in these offspring. Unfortunately, we could not measure plasma leptin in the neonatal period, although we found no difference in fat mass at one week of age and would thus not expect any difference in plasma leptin as this is the main determinant at this age of life (45). Nevertheless, one of the main functions of leptin is to reduce NPY gene transcription (46, 47). Thus, the down-regulation of NPY mRNA abundance we observed in the NR at 1 week of age could be mediated by a reduction in plasma leptin.
The thrifty phenotype hypothesis supports a link between an adverse fetal environment and the risk for adult metabolic diseases (48, 49). We nonetheless shown that, in our nutrient restricted obese offspring, maternal nutrient restriction protected against the development of glomerulosclerosis associated with obesity (24). Similarly, in this present study we demonstrate an adaptation that may promote improved regulation of food intake with obesity. However, the absolute level of obesity and insulin resistance remained similar between all animals. Furthermore, more importantly, the obese offspring born to nutrient restricted mothers developed hyperinsulinemia a well recognized major risk for type II diabetes (50).
One of the limitations our study is the use of real-time RT PCR technology on the entire hypothalamus in order to assess the expression of a large number of genes involved in appetite regulation. The advantage of this procedure is that it enabled us to obtain quantitative information with regard to the effect of age, maternal diet and obesity. It is important, however, to note that the hypothalamus is not a homogeneous entity but includes many nuclei that may exhibit differential regulation within some of the genes we examined (5). Gene expression on the entire hypothalamus has previously been used to study appetite regulation in both rats (30) and sheep (29) but of course cannot provide detailed information on site specific regulations. So although our results substantiate current knowledge in appetite regulation, future work of this type needs to look at changes within specific nuclei of the hypothalamus.
In conclusion, we have shown that, although maternal nutrient restriction targeted over the period of early fetal brain development has very little, if any, direct effect on appetite regulation early in postnatal life, it can have a significant impact in the obese adult. The potential mechanisms allowing better control of food intake with obesity remain to be fully determined as to whether it highlights a protective adaptation or an indirect consequence of hyperinsulinemia, but our model may enable the design of new therapeutic strategies against the major dysfunctional outcomes of obesity.
This work was supported by the British Heart Foundation (BS/03/01), the European Union Sixth Framework Programme for Research and Technical Development of the European Community – The Early Nutrition Programming Project (FOOD-CT-2005-007036), the Medical Research Council and the Universities of Nottingham and Alberta.
DISCLOSURE STATEMENT: The authors have nothing to disclose