The nutritional environment encountered during fetal life is strongly implicated as a determinant of lifelong metabolic capacity and risk of disease. Pregnant rats were fed a control or low-protein (LP) diet, targeted to early (LPE), mid-(LPM), or late (LPL) pregnancy, or throughout gestation (LPA). The offspring were studied at 1, 9, and 18 mo of age. All LP-exposed groups had similar plasma triglyceride, cholesterol, glucose, and insulin concentrations to those of controls at 1 and 9 mo of age, but by 18 mo there was evidence of LP-programmed hypertriglyceridemia and insulin resistance. All LP-exposed groups exhibited histological evidence of hepatic steatosis and were found to have two- to threefold more hepatic triglyceride than control animals. These phenotypic changes were accompanied by age-related changes in mRNA and protein expression of the transcription factors SREBP-1c, ChREBP, PPARγ, and PPARα and their respective downstream target genes ACC1, FAS, L-PK, and MCAD. At 9 mo of age, the LP groups exhibited suppression of the SREBP-1c-related lipogenic pathway but between 9 and 18 mo underwent a switch to increased lipogenic capacity with a lower expression of PPARγ and MCAD, consistent with reduced lipid oxidation. The findings indicate that prenatal protein restriction programs development of a metabolic syndrome-like phenotype that develops only with senescence. The data implicate altered expression of SREBP-1c and ChREBP as key mediators of the programmed phenotype, but the basis of the switch in metabolic status that occurred between 9 and 18 mo of age is, as yet, unidentified.
- transcription factors
- insulin resistance
- metabolic syndrome
the environment encountered in fetal life is an important determinant of disease risk in adult life (25). Exposure to less than optimal nutrition in utero modifies long-term gene expression and the nature of interactions between the genotype and postnatal environment (16). Epidemiological studies show that, in humans, impaired growth in fetal life, followed by rapid catch-up growth in infancy, is a risk factor for non-insulin-dependent diabetes and cardiovascular disease (5, 6). Such findings suggest that, while the origins of the metabolic syndrome in humans are complex and multifactorial, nutrition in early life may be a contributing factor. The etiology of all of the main components of the syndrome (obesity, hyperinsulinemia, dyslipidemia, and cardiovascular and renal disease) is likely to involve a variety of influences across the lifespan (9). The expression of genes that predispose to, or protect against, any of these conditions will be modified through interactions with the postnatal lifestyle and environment (16).
Studies of rodents and sheep are consistent with the assertion that undernutrition in pregnancy is able to program raised blood pressure, glucose intolerance, insulin resistance, and obesity (16). The feeding of low-protein diets during rat pregnancy exerts strong programming effects on major organs and tissues that are, at least in part, mediated through remodeling of cell type and numbers (16). Maternal low-protein diets are clearly established as early-life inducers of hypertension (18), renal injury (20), and glucose intolerance (7, 15). A number of mechanisms have been suggested to explain the long-term programming effects of nutrition during development. These include epigenetic modification of gene expression and disturbance of the maternal-fetal endocrine axis (9, 16, 21). A relatively small number of “gatekeeper” genes, for example, transcription factors that regulate early cell differentiation, may play a critical role in mediating programming effects.
Previous studies in our laboratory have identified programmed changes in the expression of transcription factors involved in regulation of lipid metabolism as possible drivers of disordered metabolism in rats exposed to low-protein diets in fetal life (4). In young rats subjected to protein restriction in utero, lipogenesis appears to be suppressed, as expression of sterol response element-binding protein (SREBP)-1c and carbohydrate response element-binding protein (ChREBP), along with their downstream target genes, is significantly reduced (4). A similar phenomenon is noted in the young adult offspring of rats fed an iron-deficient diet (40). Expression of peroxisome proliferator-activated receptor-α (PPARα) was noted to be upregulated at weaning in rats exposed to low-protein diets in utero, suggesting enhanced lipid oxidation (4). Early life programming of the expression of transcription factors such as SREBP-1c may therefore play a critical role in determining the metabolic consequences of prenatal undernutrition.
Most observations of programmed obesity and related metabolic disorders in animals have been made in relatively young adult rats. In terms of understanding how prenatal factors contribute to long-term disease processes, this represents a major weakness. The aim of the present study was therefore to characterize the development of an insulin-resistant phenotype and the associated underlying patterns of gene expression in an aging population of rats subjected to intrauterine protein restriction. Our working hypothesis was that nutritionally programmed, long-term modifications of expression of the transcription factors governing lipid metabolism would impact on the metabolic profile of the aging animal.
MATERIALS AND METHODS
The experiments in this paper were performed in accordance with the Animals (Scientific Procedures) Act, 1986, and were licensed by the Home Office. Animals were held under temperature-controlled conditions on a 12:12-h light-dark cycle. The animals had ad libitum access to food and water at all times. Fifty-seven virgin female Wistar rats (Harlan, Belton, UK) were mated at weights between 200 and 250 g. Upon confirmation of mating by the appearance of a semen plug on the cage floor, the rats were allocated to be fed either a control diet (18% casein) or a low-protein diet (LP, 9% casein), as described previously (1). The full composition of the diets is published elsewhere (18). The diets were isocaloric, the difference in energy between the control and MLP diets being made up with additional carbohydrate in a ratio of 2:1 starch/sucrose (wt/wt).
LP feeding was targeted at single weeks in gestation days 0–7 (LPE), days 8–14 (LPM), and days 15–22 (LPL) and also fed throughout gestation (days 0–22, LPA). Eight dams (1 control, 1 LPA, 3 LPE, and 3 LPM) failed to deliver at the end of gestation, and the final numbers of litters were Control n = 10, LPA n = 11, LPE n = 10, LPM n = 8, and LPL n = 10. The early period (days 0–7) corresponds to the embryonic phase of development in the rat, and in fact embryos implant only at around day 4.5 (6). The midgestation period (days 8–14) largely corresponds to the period of organogenesis, and late gestation (days 15–22) is the period of most rapid growth and differentiation of key structures. By feeding at these targeted periods it is possible to identify when nutritional programming occurs, and this can provide important indicators of potential mechanisms.
At delivery of litters, all dams were transferred to a standard laboratory chow diet (B&K Universal rat and mouse diet, 20% protein, 3% fat), and the litters were culled to a maximum of eight pups to minimize variation in the nutrition of the pups during suckling. The offspring of the control and LP-fed dams thus differed only in terms of their prenatal nutritional exposures. At 4 wk of age, the offspring were weaned onto a chow diet. One randomly selected male and one randomly selected female from each litter were killed using a rising concentration of carbon dioxide at ages 1, 9, and 18 mo. Animals were not fasted prior to culling. Whole blood was collected into Vacutainers by heart puncture and plasma prepared by centrifugation at 13,000 rpm at 4°C for 10 min. The liver, adipose tissue (perirenal depot), skeletal muscle (gastrocnemius), heart, brain, lungs, and kidneys were dissected from each animal, weighed to the nearest 0.1 mg, and then snap-frozen in liquid nitrogen. Organs and plasma were stored at −80°C until used for further analyses. A standard portion of liver taken from the left lobe was fixed in buffered formalin for histological analysis. Data from other aspects of this trial have been published elsewhere (1, 19).
Histological examination of the liver.
Liver specimens were fixed in 10% formalin for ≥24 h before processing by use of an automated system (Spencers Automatic Tissue Processor) for paraffin embedding in preparation for sectioning. Serial sections of 7-μm thickness were taken using a microtome, and these were then stained using hemotoxylin and eosin.
Determination of mRNA expression.
RNA was isolated using the phenol-chloroform extraction method described by Chomczynski and Sacchi (2). cDNA was synthesized using the Taqman Reverse Transcription Reagents kit, and then quantitative RT-PCR was performed using the ABI Prism 7700 Sequence Detection System (Applied Biosystems). Fluorogenic probes were labeled with FAM (6-carboxyfluorescein) at the 5′ end and with TAMRA (6 carboxytetramethylrhodamine) at the 3′ end. A negative template control and relative standard curve were included on every PCR run. The standard curve was prepared from a pool of sample cDNA over a fourfold range of dilutions. Relative target quantity was calculated from the standard curve, and all samples were normalized against appropriate housekeeping genes. Hepatic and muscle mRNA expression was normalized relative to eukaryotic 18S rRNA (Applied Biosystems), which did not vary significantly between the different experimental groups. Adipose tissue expression was normalized relative to β-actin, as prenatal diet influenced 18S rRNA expression. 18S expression did not vary between groups in other tissues, and, similarly, β-actin expression in adipose tissue did not vary significantly between the different experimental groups. Sequences of primers and probes used for the PCR studies are shown in Table 1.
Determination of protein expression.
The expression of proteins in the liver was quantified using Western blotting, as described previously (12). Briefly, hepatic protein was extracted using the method of Chomczynski and Sacchi (2). Isolated protein concentrations were quantified using the method of Lowry (22), adapted for use in a 96-well microassay plate. μg Protein sample (50–75) was denatured by boiling for 5 min, and sample was then loaded onto an SDS polyacrylamide gel for separation (40 mA for 2 h). Separated proteins were transferred on to Hybond ECL (enhanced chemiluminescence; Amersham Pharmacia Biotech) nitrocellulose membrane, and, after blocking of nonspecific binding sites, membranes were incubated overnight with primary antibody solutions at 4°C. For SREBP-1c the primary antibody was SREBP-1 specific, developed in mice (ATCC, Middlesex, UK). Primary antibody for ChREBP was a specific anti-ChREBP raised in rabbit (Novus Biologicals), for PPARα, a rabbit polyclonal (Cayman Chemical) raised against rat/human PPARα (amino acids 22–36) and for PPARγ was a rabbit polyclonal IgG (Santa Cruz Biotechnology). Expression of α-actin protein was used to normalize specific protein expression. For this, a rabbit anti-α-actin was used as the primary antibody (Sigma-Aldrich).
After this incubation, further washing and blocking steps were followed by incubation with a horseradish peroxidase-labeled secondary antibody to allow imaging of antibody binding using AIDA Image Analyzer software (Raytest). For SREBP-1c the secondary antibody was peroxidase-conjugated rabbit anti-mouse (Dako, Cambridgeshire, UK). For α-actin, ChREBP, PPARα, and PPARγ the secondary antibody was peroxidase-conjugated swine anti-rabbit (Dako).
Determination of plasma insulin, glucose, lipid profiles, and homeostasis model assessment of insulin resistance index.
Total plasma cholesterol and plasma triglyceride concentrations were determined using commercially available kits, following the manufacturers’ instructions (Wako). Insulin was determined using a radioimmunoassay kit (Linco Research, Charles, MO). Plasma glucose was determined using an automated analyzer (2300 STAT Plus, glucose and lactate analyzer). The homeostasis model assessment of insulin resistance index (HOMA-IR) was calculated as ([plasma glucose] × [plasma insulin]/22.5.
Extraction and quantification of hepatic triglycerides.
Lipid content of liver tissue was extracted using hexane and isopropanol. Tissue (200 mg) was homogenized in 1 ml of hexane-isopropanol (3:2 vol/vol) and centrifuged at 3,500 rpm for 5 min at room temperature. The extract was transferred to a clean glass vial and the extraction repeated. Extracts were dried at 56°C for 15 min and then dissolved in 450 μl of acetone.
Lipid extracts were applied to thin-layer chromatography plates and developed in 90:30:1 (vol/vol/vol) hexane-diethyl ether-glacial acetic acid. Standards of cholesterol, cholesterol ester, and triglyceride were run with each sample plate. After drying, the plates were sprayed with 10% sulfuric acid in methanol and heated gently. Plates were scanned, and band intensities were measured using an AIDA image analyzer. To quantify the lipids in each band, the stained bands were cut out of each plate and extracted again with hexane-isopropanol mixture. Extracted lipids were then assayed using triglyceride and total cholesterol kits (Wako).
All data are presented as means ± SE. Unless stated otherwise in the text, data were analyzed using one- or two-way ANOVA using SPSS version 11.5. Where ANOVA indicated a significant effect of treatments, post hoc tests were performed using a Bonferroni test. Post hoc testing was performed only for univariate effects, and it was not possible to identify differences between specific groups where the difference might have arisen through and interaction of two or three factors.
Body and liver weights.
As reported previously (1), body weights of the animals exposed to a maternal LP diet at any stage of development did not differ significantly from the control animals (Table 2). There was no evidence that LP exposure programmed obesity. Liver weight and liver weight as a proportion of body weight varied between males and females. Liver weight in relation to body weight provides a measure of organ size that corrects for variation in body weight with aging. This variable tended to decline between 1 and 9 mo of age.
Metabolic indexes in plasma.
Concentrations of insulin in nonfasted plasma samples showed evidence of fetal programming effects (Table 3). Within each group of male offspring from dams exposed to LP diet, apart from the LPM group, insulin concentrations were higher than in controls at 18 mo of age. Among female offspring, hyperinsulinemia relative to control animals was noted only in the LPL group. Plasma glucose concentrations increased between 1 and 9 mo of age in most groups of animals, but no significant influence of the maternal diet was noted. The HOMA-IR index did not vary significantly with maternal dietary group or age but tended to be higher in all LP-exposed offspring than in controls at 18 mo of age.
Profound changes in plasma lipid profiles were noted as the animals aged (Table 4). Triglyceride concentrations were significantly elevated in males and females at 9 and 18 mo compared with 1-mo-old rats. Males exhibited higher total cholesterol at both 9 and 18 mo, but in females cholesterol was elevated only at 18 mo compared with 1 mo of age. Maternal diet also influenced later triglyceride profiles. All males exposed to LP in utero tended to have lower triglyceride concentrations at 1 mo, but this was significant only in the LPE group. By 18 mo of age, males of the LPE and LPL groups had significantly elevated plasma triglycerides (Table 4). Females of the LPA, LPE, and LPM groups exhibited vastly elevated plasma triglyceride concentrations at 18 mo of age, being 3- to 4.5-fold higher than in control animals. Total cholesterol concentrations tended to be elevated in all 18-mo-old males exposed to LP diets in utero, but the effect was significant only in the LPL group. Eighteen-month-old females of the LPA and LPE groups also had elevated total plasma cholesterol.
Hepatic lipid content.
Hepatic triglyceride concentrations were determined in a subset of the samples obtained at all three ages (Table 4). There were no differences between maternal dietary groups at 1 or 9 mo of age, although hepatic triglyceride concentrations were markedly higher at 9 mo than at 1 mo. By 18 mo of age, all LP-exposed groups exhibited massive elevation of hepatic triglyceride concentrations, with an almost threefold difference noted between the LPA and control groups. As shown in Fig. 1, there was evidence of hepatic steatosis in all of the LP-exposed groups (both male and female animals). At 1 mo of age (Fig. 1, A–E), the livers of LP-exposed animals appeared histologically normal, and this remained the case at 9 mo (data not shown). By 18 mo of age, however, there was increased lipid deposition evident in all LP groups. LPA, LPE, and LPL groups all exhibited microvesicular steatosis, where lipid was deposited in small cytoplasmic vesicles, and macrovesicular steatosis, where large vesicles displace the cell nuclei to the periphery of the hepatocytes. Sections of liver from LPM rats (Fig. 1I) showed particularly severe steatosis with many macrovesicular deposits.
mRNA expression in liver, adipose tissue, and skeletal muscle.
Quantitative RT-PCR demonstrated large changes in the mRNA expression of genes and transcription factors regulating and controlling lipid metabolism in the liver at both 9 and 18 mo of age. As shown in Fig. 2A, at 9 mo of age, hepatic expression of SREBP-1c was suppressed by LP exposure at any stage of fetal rat development. The downstream targets of SREBP-1c, fatty acid synthase (FAS), and acetyl-CoA carboxylase-1 (ACC1) were similarly suppressed at the same age (Fig. 2, B and C). By 18 mo of age, however, the expression of all three genes was significantly elevated in LP-exposed groups relative to controls. ChREBP (Fig. 2D) and L-type pyruvate kinase (L-PK; Fig. 2E) showed a similar pattern of expression, both being suppressed in LP-exposed groups at 9 mo but being significantly overexpressed relative to controls by 18 mo of age. In contrast, expression of IRS-2 was unchanged by prenatal protein restriction among 9-mo-old rats. Controls exhibited an increase in expression (Fig. 2F) between 9 and 18 mo, which did not occur in the LP-exposed groups. Hepatic expression of PPARα mRNA was similar in all groups at 9 mo (Fig. 3A). As with IRS-2, expression increased between 9 and 18 mo in controls but not in LP-exposed groups, such that the latter showed expression levels significantly below control at the later age point. Medium-chain acyl-CoA dehydrogenase (MCAD) expression (Fig. 3C) mirrored that of PPARα. Expression of PPARγ was similar in all groups at 9 mo (Fig. 3B) but increased significantly in all LP-exposed groups at 18 mo.
Although dramatic changes in hepatic gene expression were evident as a result of fetal programming, gene expression in skeletal muscle was unaffected by the maternal diet (Fig. 4). Age-related declines in expression of SREBP-1c (Fig. 4A), FAS (Fig. 4B), and IRS-2 (Fig. 4E) were noted in all groups in this tissue. Expression of muscle ACC1 (Fig. 4C) and ChREBP (Fig. 4D) increased between 9 and 18 mo of age. In adipose tissue (Fig. 5), expression of SREBP-1c, FAS, ACC1, and ChREBP (Fig. 5, A–D) declined between 9 and 18 mo of age. Expression of IRS-2 (Fig. 5E) in this tissue increased dramatically over this period. Rats of the LPE group exhibited increased adipose tissue expression of SREBP-1c and its downstream targets (Fig. 5, A–C), but this effect was noted only at 9 mo of age.
Protein expression in liver.
Protein expression was determined for the four transcription factors SREBP-1c, ChREBP, PPARα, and PPARγ in liver tissue. In general terms, protein expression complemented that of the mRNA. For SREBP-1c there was a clear programming effect of the maternal diet on expression of both immature and mature forms of the protein (data for mature form only shown; Fig. 6A). The protein was expressed at a lower level in the LP-exposed groups relative to the controls at 1 mo of age. Expression increased with age in all groups but to a much greater extent in the LP-exposed groups, producing a significant overexpression in all except LPL, by 18 mo. Expression of the mature (r = 0.692, P < 0.001) and immature (r = 0.735, P < 0.001) proteins was significantly correlated with mRNA expression. Similarly, the expression of ChREBP protein was related to mRNA expression (r = 0.397, P = 0.008). Expression was suppressed in 1-mo-old LP-exposed animals. In controls, expression declined with age, but in all LP groups an increase was noted, producing a significantly higher expression at 18 mo of age (Fig. 6B).
PPARα protein expression was greater in all LP-exposed groups at 1 mo of age (Fig. 7A). An increase in expression in control animals between 1 and 9 mo abolished this difference. PPARα expression declined dramatically in all groups between 9 and 18 mo, but this reduction was greater in the LP-exposed animals, with all groups showing significantly lower expression than controls. Protein expression was correlated (r = 0.584, P = 0.01) with mRNA expression. The relationship between protein and mRNA expression was much weaker for PPARγ but was still significant (r = 0.31, P = 0.028). However, as shown in Fig. 7B, there were no significant effects of age or maternal diet on expression of this protein.
This paper reports an animal model displaying aspects of the metabolic syndrome programmed by maternal dietary manipulation during pregnancy. The experiment has clearly demonstrated the power of nutritional factors during development to permanently alter expression of the fetal genome and the lifelong response of the animal to environmental stimuli. The measurements from plasma samples in this study represented the fed state. Although all blood samples were collected at the same time of day, the study has a more limited scope for the assessment of metabolic parameters. Despite this, there were some variations in the programmed phenotype that related to the sex of the offspring and timing of protein restriction; in general, LP exposure promoted clear elevations of plasma triglycerides and cholesterol, hepatic steatosis, and insulin resistance. Importantly, these features emerged only with aging. Although the animals were not obese, it is has been previously demonstrated that the LP diet programs other features of metabolic syndrome, including hypertension (18) and renal dysfunction (20). Blood pressure was not determined in this study, but our previous studies (18, 20) demonstrated raised pressure to be a consistent outcome following maternal protein restriction.
The metabolic syndrome-like phenotype clearly emerged as a consequence of altered expression of key genes involved in the regulation of fat metabolism and insulin signaling. Hepatic expressions of SREBP-1c, ACC1, FAS, and ChREBP have been previously reported to be suppressed in LP-exposed offspring at 1 mo of age (4), and this finding was replicated at the protein level in the present study. With aging, all of these genes were upregulated. Expression of PPARγ was also increased in the liver with aging. In contrast, the expression of PPARα and its downstream target MCAD declined with aging and was significantly lower in the LP-exposed animals than in controls at 18 mo of age. Collectively, these changes in gene expression are consistent with the development of a phenotype that promotes lipogenesis and lipid storage (11). Despite the fact that PPARγ protein expression did not rise significantly alongside the mRNA expression changes in the aging animals, the trend for the change was observed, and there was a significant association between protein and mRNA expression measurements (r = 0.31, P = 0.028). We are therefore confident that the gene expression changes are drivers of increased hepatic lipogenesis and storage. In the LP-exposed animals, much of this lipid clearly remained within the liver, which may reflect the fact that expression changes were noted only in this tissue. Induction of fat synthesis in the offspring of protein-restricted rats has also been reported by Maloney et al. (24).
Although we report changes in the expression of PPARα protein between 1 and 18 mo of age in the present study, data regarding mRNA expression at 1 mo was not shown. Data published elsewhere shows that hepatic PPARα mRNA expression was elevated in all LP groups relative to control animals (4), which is consistent with reports from other workers (21).
Downregulation of IRS-2 in the liver at 18 mo of age provides the strongest evidence that insulin resistance was present in the LP-exposed groups at that age (39). Other studies have demonstrated that an LP diet in rat pregnancy can program glucose intolerance and insulin resistance in the offspring (7). These studies are also suggestive of an age-related component to the development of the insulin resistance. Rats subjected to intrauterine protein restriction generally show enhanced insulin sensitivity in the early postnatal period and exhibit insulin resistance only beyond the age of 1 yr. These studies also provide evidence of sex-specific programming in that, although males exposed to LP diets in fetal life exhibited frank diabetes by 17 mo (32), females developed impaired glucose intolerance only at 21 mo (8). Programmed reductions of GLUT4 and PKCζ in skeletal muscle accompanied these effects in males (29).
The present study is important in that it followed a large colony of aging animals from birth to 18 mo. Most studies of fetal programming focus on early adulthood and therefore overlook age-related changes in phenotype. Although Lucas et al. (23) previously reported on plasma triglyceride and cholesterol profiles of LP-exposed offspring at 6 mo of age, the present work provides a broader profile of metabolic data and was specifically designed to assess how the programmed phenotype changes with senescence. The observed changes in plasma cholesterol and plasma and hepatic triglycerides suggest that there are profound changes at the metabolic level between 9 and 18 mo of age. One of the strengths of our findings is that the changes in metabolic parameters very closely parallel the changes in gene and protein expression within the liver. This strongly implicates hepatic SREBP-1c and ChREBP and their associated lipogenic targets in the development of programmed metabolic syndrome.
Upregulation of SREBP-1c between 9 and 18 mo of age clearly had an impact on expression of ACC1 and FAS in the livers of the LP-exposed offspring, as all three rose in a coordinated fashion. We propose that the increase in circulating and hepatic triglycerides observed in these groups is largely attributable to this change. Decreased lipid oxidation associated with suppression of PPARγ and MCAD would also contribute to the observed phenotype (13). The lack of differences in expression of these genes between control and LP-exposed animals at 9 mo of age, however, may indicate that the lipogenic pathways are more important in determining the development of the programmed phenotype.
The experimental design generated four different LP groups based on different periods of maternal protein restriction. The intention of the study was to explore the possibility that programming of lipid metabolism occurs during a specific critical window during fetal life. The finding that protein restriction generated a broadly similar phenotype regardless of the timing of the insult was surprising but is an important observation. Several inferences might be drawn from the study. First, it is possible that the pathways involved in lipid metabolism are extremely sensitive to undernutrition at any stage of fetal life and that the time frame in which programming might occur is very broad. This would be an important contrast to other aspects of physiological function and health markers, where periods of greater sensitivity in later gestation have been identified (16, 20). Alternatively, it may be argued that the sensitive period for programming of lipid metabolism does lie in mid- to late gestation. The metabolic consequences for the mother of experiencing protein undernutrition in early pregnancy and then changing back to a control diet could provide a stressor to the fetus that produces programmed metabolic sequelae in the LPE and LPM groups. Metabolic markers in the dams in this study were not evaluated, so this latter possibility cannot be assessed.
The LP-exposed animals in this study all exhibited hepatic steatosis regardless of the timing of protein restriction. This occurred in the absence of obesity. Fatty liver, caused by chronic hepatocyte accumulation of lipid, can ultimately lead to inflammation and scarring, with the potential to progress to cirrhosis and liver failure (31). Insulin resistance is a strong driver of fat accumulation in the liver, and it is now recognized that the liver is a major target of injury in patients with metabolic syndrome (27). Fatty liver has also been related to overexpression of SREBP-1c, which is elevated in response to hyperinsulinemia (34, 35). ob/ob Mice with steatotic livers exhibit selective upregulation of SREBP-1c, ACC1, and FAS (33), all of which were elevated in the present study. Fatty liver is often associated with obesity but has been reported in nonobese human subjects with β-cell secretory impairments preceding glucose intolerance (26). Nonalcoholic fatty liver disease may be viewed as an early predictor of metabolic disorders in subjects of normal weight (14). On the basis of the histological examination, the LPM group was most severely affected by hepatic steatosis, suggesting that the midperiod of rat gestation is a particularly sensitive period for programming of this phenotype.
The lack of obesity in the LP-exposed animals has been discussed at length elsewhere (1) but is surprising, given reports that prenatal undernutrition in humans, rats, and sheep can elicit pronounced deposition of abdominal fat later in life (16, 38). To a large extent, this may depend on the nature of the postnatal diet (17), and in this case the standard laboratory chow diet was a low-fat diet designed to prevent development of obesity. It would be of interest to assess the metabolic responses of the different groups of LP-exposed offspring to a high-fat diet in adult life. Induction of obesity through high-fat feeding may be expected to bring forward the age at which the metabolic profile of the liver switches to a lipogenic, lipid-storing phenotype.
It might be argued that the development of the insulin-resistant phenotype with age is a product of a postnatal stressor that overcomes a compensatory insulin sensitivity noted earlier in life (15, 30). However, the nature of that stressor and the basis of a differential metabolic response to aging between groups is currently unknown. The patterns of gene expression noted in control animals in the present study were consistent with other reports that in aging animals expression of SREBP-1c and its downstream target lipogenic enzymes is lower than in younger littermates (28). Aged rats switch fuel utilization in peripheral tissues from glucose to fatty acids. This is accompanied by an increase in hepatic expression of genes that promote both glucose and lipid oxidation (37). Thus, LP-exposed offspring appear to have specific defects in this normal response to aging, which lead to accumulation of triacylglycerol in the liver and plasma. Hepatic IRS-2 may be a key element in the switch from suppressed lipogenic pathways observed at 9 mo to the overexpressed lipogenic components noted at 18 mo. IRS-2 expression was similar in the livers of LP-exposed animals at 1 and 9 mo but was suppressed, relative to controls, by 18 mo. This could be a consequence of SREBP-1c inhibition, an effect that can be mediated by interference of the binding of FoxO1 to the IRS-2 promoter (3) but could also be a cause of SREBP-1c overexpression (10, 36).
The current study confirms the importance of early-life nutrition on the development of metabolic competence later in life. Importantly, the greatest impact of maternal diet on the phenotype of the offspring was manifested in the oldest animals. This highlights the need for future studies that specifically allow for consideration of the impact of senescence on phenotypes determined by metabolic programming in early life. Clearly, maternal and early-life diet may be important factors in the development of the metabolic syndrome through reprogramming of lifelong expression of key transcription factors involved in the regulation of both carbohydrate and lipid metabolism. The mechanisms underlying these fundamental changes remain to be elucidated.
The animal studies described in this paper were funded by a project grant from the British Heart Foundation. A. Erhuma was supported by a scholarship from the Libyan government.
The technical assistance of Mr. R. Plant is gratefully acknowledged.
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 © 2007 by American Physiological Society