A mismatch between fetal and postnatal environment can permanently alter the body structure and physiology and therefore contribute later to obesity and related disorders, as revealed by epidemiological studies. Early programming of adipose tissue might be central in this observation. Moreover, adipose tissue secretes adipokines that provide a molecular link between obesity and its related disorders. Therefore, our aim was to investigate whether a protein restriction during fetal life, followed by catch-up growth could lead to obesity in 9-mo-old male mice and could alter the adipose tissue gene expression profile. Dams were fed a low-protein (LP) or an isocaloric control (C) diet during gestation. Postnatal catch-up growth was induced in LP offspring by feeding dams with control diet and by culling LP litters to four pups instead of eight in the C group. At weaning, male mice were fed by lab chow alone (C) or supplemented with a hypercaloric diet (HC), to induce obesity (C-C, C-HC, LP-C, and LP-HC groups). At 9 mo, LP offspring featured increased relative fat mass, hyperglycemia, hypercholesterolemia, and hyperleptinemia. Using a microarray designed to study the expression of 89 genes involved in adipose tissue differentiation/function, we demonstrated that the expression profile of several genes were dependent upon the maternal diet. Among the diverse genes showing altered expression, we could identify genes encoding several enzymes involved in lipid metabolism. These results indicated that offspring submitted to early mismatched nutrition exhibited alterations in adipose tissue gene expression that probably increases their susceptibility to overweight when challenged after weaning with a HC diet.
- fetal programming
it has been well established from epidemiological data and animal models that the early life environment plays an important role in determining long-term health of the adult. Early epidemiological studies in the United Kingdom showed that individuals with low birth weight were at increased risk of developing many components of the metabolic syndrome such as impaired glucose tolerance, type 2 diabetes, hypertension, and cardiovascular diseases (1). These observations founded the “thrifty phenotype hypothesis,” which proposes that in response to a deprived early environment, the fetus adopts two strategies to ensure survival: 1) it will favor the growth of certain organs, such as the brain, at the expense of others, like peripheral organs; and 2) it will adopt metabolic adaptations that should ensure a better use of nutrients. These adaptations may jeopardize the development of the fetus and will determine its future health (27). More recently, Gluckman et al. (22) stipulated that it is not only a suboptimal fetal environment that is detrimental, but most likely it is the mismatch between individual's adaptations to the environment predicted to be experienced and the environment really encountered. Most low-birth weight infants show rapid early postnatal growth. This compensatory mechanism for prenatal growth deficit might have long-term benefits regarding adult height (32) and neurological development (35). Although catch-up growth can have advantages during early life, it is also viewed as a risk factor for later health (59). Indeed, epidemiological data showed that catch-up growth subsequent to low weight at birth or in infancy increases the susceptibility for central obesity, type 2 diabetes, and cardiovascular diseases (7, 15, 38). Animal studies also confirmed these observations. Previous work carried out in our laboratory showed that in rats, catch-up growth during lactation after fetal protein restriction was required for the development of obesity later in life (3). Evidence from rats and mice in cross-fostering studies correlates these data, showing that the suckling period is critical for developmental programming of increased adipose tissue in rodents (42, 45, 58). Moreover, the programming of adipose tissue development and physiology might be particularly important because obesity, as well as a central pattern of fat distribution, are main risk factors for type 2 diabetes and cardiovascular diseases (49, 56). Indeed, an excess of adipose tissue leads to alterations in triglyceride and free fatty acid plasma levels but also to reproducible changes of the expression in adipocytes of specific genes encoding proteins such as leptin, resistin, adiponectin, or plasminogen activator inhibitor-1 (26, 28, 54). Evidence for programming of adipose tissue molecules has been provided earlier by Guan et al. (25) who demonstrated that maternal low protein diet during gestation and lactation permanently alters the gene expression of enzymes involved in lipogenesis in adipose tissue from adult rats. Furthermore, previous results obtained in our laboratory indicated that catch-up growth after a fetal low protein diet induced upregulation of mRNA expression of angiotensinogen and adiponectin in adipose tissue from adult male rats (3).
Therefore, our aim was to investigate whether a mismatch in early life nutrition, produced by forced catch-up growth after fetal protein restriction, would have long-term consequences on adipose tissue function and the development of obesity in mice and whether a postweaning hypercaloric diet would amplify the phenotype. For that purpose, we investigated body weight, plasma parameters, and gene expression profile in adipose tissue using the ADIPOCHIP, a low-density array dedicated to monitor the expression of genes involved in adipocyte differentiation/function, in male mice that were fed after weaning with a normal or hypercaloric diet.
MATERIALS AND METHODS
Animals and Dietary Manipulations
C57BL6/J mice (Janvier, Le Genest Saint Isle, France) were bred in our laboratory and maintained under controlled conditions (22°C, 12:12-h light-dark cycle). Nonprimiparous females were housed with males overnight and mating was confirmed by the presence of a vaginal plug. The dams were then housed individually during gestation and lactation. Experimental procedures in animals were approved by the Animal Care and Use Committee of the Université catholique de Louvain and were performed in accordance with the “Principles of Laboratory Animal Care” (NIH Publication 85-23). During gestation, dams were fed a control diet (C: 192 g protein/kg/20% of energy) or an isocaloric low-protein diet (LP: 81 g protein/kg-8% of energy) purchased from Hope Farm (Woerden, Netherlands). The composition and source of diets were the same as those used in our previous studies of the LP rat and are described elsewhere (4). At birth, pups were weighed and litter size was recorded. To induce a catch-up growth, the number of pups was reduced to four in the LP group instead of eight in C litters and during lactation, all dams were fed with control diet.
At weaning, two male pups/litter were randomly selected and divided into two subgroups which received either standard laboratory chow alone (Carfil Quality, Turnhout, Belgium) or the same diet supplemented with a hypercaloric, high-fat, high-sucrose complement, consisting in an emulsion of 10% sucrose solution together with a progressively increasing concentration of corn oil (4% oil for 10 days, 8% for 15 days, 16% for other 15 days, and 32% from day 70 until the end of the experiment). During 36 wk, animals were weighed, and food intake was measured every 5 days. At the end of the experiment (9-mo of age), animals received a lethal injection of Natrii pentobarbitalum, 60 mg/kg body wt (Nembutal; CEVA Santé Animale, Brussels, Belgium), and after dissection, the different organs were weighed and stored at −80°C until analysis.
Blood Sampling and Analysis
Blood was collected at death, by cardiac puncture, in EDTA tubes, and plasma was obtained by centrifugation in nonfasting conditions. For glucose concentration analysis, proteins were precipitated by the addition of 200 μl HClO4 (0.33 N) to 20 μl blood, and concentration was determined by the glucose oxidase method using Trinder's reagent [coefficient of variation, CV, intra-assay < 5% and interassay < 4% (Stanbio Laboratory, Boerne, Texas)]. Triglyceride and total cholesterol concentrations were determined by using, respectively, the TRF400CH and CTF400CH kits following the manufacturers instruction [CV intra-assay < 1.7%, 1.8% and interassay < 3.3%, and 2.1%, respectively (Chema Diagnostica, Jesi, Italy)]. HDL-cholesterol concentration was determined with the total cholesterol kit after precipitation of low-density and very low-density lipoproteins by HDL precipitating reagent CD0400CH [CV intra-assay < 2% and interassay < 2.2% (Chema Diagnostica)]. Plasma insulin was measured using ultra-sensitive insulin enzyme-linked assay [CV intra-assay < 2% and CV interassay < 5% (Mercodia, Uppsala, Sweden)]. Leptin concentration was determined by ELISA [CV intra-assay < 2% and CV interassay < 4.5% (Biovendor, Heidelberg, Germany)]. Resistin, IL-6, TNF-α, monocyte chemoattractant protein-1 (MCP-1), and plasminogen activator inhibitor-1 (PAI-1) were measured by ELISA using the Mouse Serum Adipokine Linco plex kit [CV intra-assay < 5% and interassay < 12% (Linco Research, St. Charles, MI)]. Metabolite and adipokine analysis were performed in maximum two plates per assay.
Adipocyte size was estimated in perigonadal fat tissue from male mice as described previously (4). Fat cells were fixed following Etherton et al. (16) with an osmium tetroxide fixation of adipocytes and digestion of connective tissue with 8 M urea. Cell size was estimated by using the ZEISS KS 400 3.0 software with an Axioskop 2 Mot Plus Zeiss microscope (Carl Zeiss, Hallbergmoos, Germany) in a minimum of 500 cells per animal within six animals per group.
RT Real-Time PCR Assays
Perigonadal fat pads were harvested in acid guanidium thiocyanate solution and total RNA was extracted with the addition of phenol/chloroform as described by Chomczynski and Sacchi (6). The quality of RNA was assessed by electrophoresis in 1% agarose gel, and samples were stored at −80°. For cDNA synthesis, 1 μg total RNA was reverse transcribed with the SuperScript III First Strand Synthesis System for RT-PCR (Invitrogen Life Technologies, Gaithersburg, MD). Quantitative real-time PCR was performed using the SYBR Green I master mix according to supplier protocol (Eurogentec, Seraing, Belgium). Gene-specific oligonucleotides were designed with Primer Express 2.0 software (Applied Biosystems, Foster City, CA) and are presented in Supplemental Table 1 (published with online version of this article). Real-time PCR was performed with the ABI Prism 7000 Sequence Detection System instrument and software under standard conditions following manufacturer's instructions (Applied Biosystems, Foster City, CA). Each sample was run in duplicate and the mean value was used to calculate the mRNA abundance according to the ΔCt method. Briefly, data obtained were normalized to the expression of Tata Box binding protein as housekeeping gene for each sample (ΔCt), and then each sample was expressed in relation to the reference sample (ΔΔCt = ΔCt target sample − ΔCt reference sample). The final values are expressed as 2−ΔΔCt for each sample, and to compare differences due to treatment results, values were expressed relatively to the C group where the relative mRNA abundance has been arbitrarily set to 1.
To assess gene expression analysis, a low-density DNA array was used. This array, ADIPOCHIP, was developed in collaboration between the URBC of Namur at the Facultés Universitaires Notre-Dame de la Paix, Namur, Belgium) and Eppendorf Array Technologies (Namur, Belgium) (for more details see Ref. 11) and allows the expression analysis of 89 murine genes related to adipocyte differentiation/function. The technique is based on a system with two arrays on a glass slide and three identical subarrays per array (triplicate spots). The reliability of hybridization and experimental data was evaluated using several controls, as previously described (11). The validation and the reliability of the results obtained with the low-density arrays developed by Eppendorf are reported elsewhere (10, 11, 55). For the profiles of gene expression, two independent experiments were performed with one or two samples per group, providing at the end, the hybridization of three individuals per group. A pool of total RNA extracted from adipose tissue of several nontreated animals was taken as a reference sample.
RNA reverse transcription and cDNA hybridization.
Total RNA was extracted from murine perigonadal adipose tissue as described earlier (6). Quality was checked with a Bioanalyzer (Agilent Technologies, Santa Clara, CA), and 15 μg of RNA were used for reverse transcription with Superscript Reverse Transcriptase (Invitrogen Life Technologies) in the presence of biotin-11-dCTP (Perkin-Elmer, Waltham, MA) as described previously (11). Detection was performed with a cyanine 3-conjugate anti-biotin IgG (Jackson ImmunoResearch Laboratories, Suffolk, UK). Fluorescence of hybridized arrays was scanned with the Packard ScanArray (Perkin Elmer) at a resolution of 10 μm. To maximize the dynamic range of detection, the same array was scanned with different photomultiplier gains to quantify high-copy and low-copy expressed genes. The scanned 16-bit images were imported to ImaGene 4.1 Software (Biodiscovery, El Segundo, CA) to quantify signal intensities. The fluorescence intensity of each DNA spot was calculated as the average intensity of each pixel present within the spot and by subtracting local mean background. Intensities of triplicate fluorescent signals were averaged and used to calculate the intensity ratio of the reference and the other samples.
Data obtained in different hybridizations was normalized as previously described in more detail (11). First, internal standards were used for calculating a correction factor, but also for estimating the local background and evaluating the array homogeneity. Second, to consider the quality and quantity of mRNA, a second normalization was performed on the expression levels of a set of two to five housekeeping genes, depending on the sample. The variance of the normalized set of housekeeping genes was used to generate an estimate of expected variance, leading to a predicted confidence interval for testing the significance of the ratios obtained. Finally, gene expression in each sample was expressed as a mean ratio where the ratio was set arbitrarily to 1 in the C group. Ratios outside the 95% confidence interval were considered to be statistically significant as determined by two-way ANOVA and are presented in results.
Western Blot Analysis
Expression of leptin was detected by immunoblotting using a rabbit polyclonal antibody. About 100 mg of perigonadal adipose tissue was harvested in lysis buffer (2% SDS, 1% Triton-X) containing protease inhibitors (Roche, Manheim, Germany). The supernatant protein concentration was determined with a BCA protein assay kit (Pierce, Rockford, IL). Thirty-five micrograms of protein per sample were denatured by boiling for 5 min in 2% SDS buffer, separated by electrophoresis on 12.5% (wt/vol) SDS-polyacrylamide gels and transferred to a nitrocellulose membrane. The blots were incubated for 1 h in a blocking solution containing 5% nonfat dry milk in Tris buffer saline with 0.1% Tween (pH 7.4) followed by an overnight incubation with the primary anti-actin [dilution 1/850, A2066; Sigma] and anti-leptin [dilution 1/250, Ob (A-20):sc-842; Santa Cruz Biotechnologies, Santa Cruz, CA] antibodies in TBS-Tween with 1% nonfat dry milk. The membrane was then incubated for 1 h with a peroxidase-labeled goat anti-rabbit IgG secondary antibody (dilution 1/2,000; Dako Cytomation, Glostrup, Denmark). Immunoreactive bands were detected with ECL reagent (Perkin Elmer). The density ratios of the leptin to actin bands were analyzed using Image Station 1D 3.5 (Kodak, Tilburg, Belgium).
Experimental results are reported as means ± SE. Student's t-test was used to compare mean birth weight per litter and mean weight per litter at weaning, after testing for normal distribution. The Kolmogorov- Smirnov two-sample test was performed to analyze data of adipocyte cell size. Two-way ANOVA was used to test for effects of maternal and postweaning diets and interaction between these variables using the Prism Software (GraphPad Software, San Diego, CA). When specified, a Bonferroni post hoc test was performed to look at statistical differences within the variables. Differences with P < 0.05 were considered significant.
Fetal Growth Retardation, Catch-Up Growth, and Subsequent Induction of Obesity
As a consequence of maternal protein restriction during gestation, pups from LP litters presented significant growth retardation at birth (1.25 ± 0.03 vs. 1.41 ± 0.03 g, n = 12 litters/group, P < 0.01). During lactation, feeding dams with a control diet containing 20% proteins and reducing the number of pups to four, induced a catch-up growth in the protein-restricted group. At the end of lactation (day 30), the LP group presented a significant increase in body weight compared with the C group (12.88 ± 0.55 vs. 11.13 ± 0.81 g, n = 12 litters/group P < 0.05). After weaning, the male offspring were divided into two subgroups and fed by lab chow alone or supplemented with the hypercaloric diet that induced an increase of body weight gain in the C-HC and LP-HC animals. However, weight increase was greater for animals that were submitted to early mismatched nutrition (see Fig. 1). At the end of the experiment, animals from the LP-HC group age of 9 mo, featured statistically increased body weight compared with the C-HC animals, whereas there was no difference in body weight between the C-C and LP-C mice (see Table 1).
Body Composition, Fat Cell Diameter, and Daily Food Intake of 9-Mo-Old Male Mice
As summarized in Table 1, the absolute weight of the different organs showed no difference between groups except for the absolute weight of the liver that was increased in LP-HC mice. Relatively to their body weight, kidneys from these animals were smaller, whereas LP offspring had a smaller pancreas relative to their body weight, indicating an influence of the maternal diet. The various fat pads showed, as expected, a significant increase in relative weight due to the hypercaloric diet in both C-HC and LP-HC groups. In addition, all fat pads were affected significantly by the maternal diet, meaning that the hypercaloric supply had even more deleterious effects on fat deposition when animals were submitted to early mismatched nutrition. Moreover, the total fat pad relative weight was also increased in LP-C offspring reflecting an altered distribution of fat and lean mass in this group.
Fat cell diameter analysis showed a different distribution of diameters in LP animals compared with C animals, with larger diameters observed in LP cells (Fig. 2). C animals supplemented with the HC diet showed a population of larger cells indicating hypertrophy. In cells from LP-HC animals, we can observe in parallel to a population of larger fat cells with a diameter ranging from 90 to 180 μm, smaller adipocytes with a mean diameter ∼40 μm.
Food intake, measured throughout the experiment, revealed a significant increase in energy intake for the HC groups compared with the nonsupplemented animals (P < 0.0001). Nevertheless, no difference was noticed according to maternal nutrition (Fig. 3).
Plasma Parameters of 9-Mo-Old Male Mice
Plasma glucose concentration was higher in LP-C and LP-HC animals showing an early diet effect on this parameter (Table 2). No significant difference according to early diet was observed for plasma insulin. Plasma triglycerides were decreased by the HC diet compared with standard chow fed animals, whereas no significant difference in plasma free fatty acids was observed between groups. Total cholesterol concentration was raised in the C-HC and LP-HC groups and also in LP-C mice showing an influence of the maternal diet on the offspring's plasma cholesterol. However, analysis of HDL-cholesterol showed an increased concentration due to HC supplementation without influence of the maternal diet. Plasma leptin levels were increased in HC supplemented mice compared with standard fed ones. We also observed an increase of leptin plasma levels in LP-C and LP-HC mice (P < 0.05). No difference due to early or adult diet was observed for PAI-1 or resistin concentrations, whereas the concentration of MCP-1 was higher in HC animals (P < 0.001).
Gene Expression Profiling by ADIPOCHIP Microarray in Adipose Tissue of 9-Mo-Old Mice
Gene expression was analyzed in perigonadal adipose tissue from three individuals per group by a DNA microarray, bearing 89 murine probes. Among these, 18 mRNA were not detected in any of the three groups. As shown in Table 3, 35 genes presented a statistically significant fold change. Of these, 12 presented only a maternal diet influence (10 were downregulated, while 2 were upregulated), 18 genes were only influenced by the HC diet (15 were upregulated and 3 were downregulated), and six showed a significant interaction, meaning that the HC supply affected, in a different way, gene expression in the offspring from C and LP groups (see Fig. 4). Interestingly, among the genes that presented an early diet effect, eight are involved in carbohydrate and lipid metabolism (CPT-2, DHAPAT, FABP4, FAS, Glut4, GPAT, GPDcyto), while seven are considered as adipose tissue-expressed molecules (Adipsin, AGT, Cav-1, Resistin, SPARC, SREBP-1c, TF). Among the genes influenced by the HC diet, many were upregulated by the supplementation (ASP, Cav-1, Cav-2, C/EBPβ, Clic4, α2coll VI, Cst C, FABP4, Hp, leptin, LPL1, MCAD, PEDF, RAB3D, SDF2, SPARC, Stat6, VEGF A) and only three were downregulated (AGT, FAS, Resistin).
Quantitative Real-Time RT-PCR and Western Blot Analysis
To confirm the data obtained by microarray, expression of some target genes that exhibited increased or decreased expression due to early diet or HC supplementation (Leptin, FAS, SREBP-1c, LPL) was determined using quantitative real-time RT-PCR (Fig. 5). The results showed the same tendencies in gene expression alterations as observed with the ADIPOCHIP. Two-way ANOVA indicated a significant effect of early diet on SREBP-1c quantification of LP offspring, whereas leptin mRNA quantification presented an early diet as well as an HC diet effect reflected by significantly increased expression in LP-HC group. Leptin in the perigonadal adipose tissue was also analyzed at the protein level by Western blot analysis. As shown in Fig. 6, the quantification of the blots for three individuals per group showed an increase in leptin protein content relative to actin due to HC feeding but also due to early mismatched diet correlating the measured plasma concentrations.
Fetal protein restriction resulted in growth retardation at birth. The phenomenon of low birth weight as a result of nutritional perturbations in utero has been described several times in animal models and in humans (36, 40). In the present study, we showed that offspring smaller at birth caught up during lactation and exceeded the weight of controls by day 15 and then remained heavier than controls until weaning. Thereafter, mice were weaned onto standard lab chow alone or supplemented with a hypercaloric diet. At the end of the experiment, whereas LP-C animals exhibited the same weight as C-C animals, the body weight in the LP-HC group was increased compared with that one of C-HC group. In addition, animals that were protein restricted during fetal life had larger fat pads, reflecting an increase in adiposity. These results corroborate a previous study, which showed that protein-restricted mice during fetal life, but normally nourished during lactation, gained more weight than controls when given free access to a highly palatable diet (41). In the same way, a previous study carried out in our laboratory, showed that protein as well as calorie-restricted rats during fetal life gained more weight than controls after postnatal catch-up growth, when fed a cafeteria diet (3). Human data also suggest that small newborns who presented rapid postnatal growth may have increased risk of metabolic disease in adulthood, such as obesity (34, 39). In a systematic review, authors pointed toward an association between rapid postnatal growth and obesity later in life (37). These results indicated that more than fetal growth retardation, the adverse effect of a rapid postnatal catch-up growth subsequent to it, might lead to the development of overweight and obesity later in life. Consistent with this, manipulations during the suckling period might also contribute to increased risk of developing obesity at adult age. Overfed rodents during lactation, reared in small litters, presented at adult age with hyperphagia and obesity (50). In humans, it has been shown that breastfeeding is preferable to formula, because of the lower growth rate associated with breastfeeding that appeared to protect against obesity and cardiovascular disease (52). In accordance with these observations, we can conclude that postnatal overnutrition, especially during the suckling period, is able to trigger programming of adiposity, as a consequence of an adverse nutritional environment, although distinction from the effect of fetal LP diet or overfeeding during the suckling period could not be provided in our model. However, our experiments were carried out in male mice, and evidence in the literature indicates gender differences in programming mechanisms (23, 60). Moreover, obesity is a complex failure of normal metabolism where sex hormones can interact (44). We might therefore acknowledge that programming of obesity by early mismatched nutrition, as shown here on male mice, might not be applicable in females.
Some studies provided evidence that changes in perinatal nutrition might program the development of the hypothalamic neural network that regulates appetite in adult life (20, 46). It has also been shown that in rodents, contrary to humans, the development of the hypothalamic appetite regulatory system occurs predominantly after birth (5). In the present study, there was no effect of fetal LP diet and postnatal catch-up growth on food intake either on normal or hypercaloric diet. Therefore, the higher body weight and adiposity observed in LP offspring is probably not due to a difference in energy intake, although an alteration in appetite and energy expenditure network due to early nutrition could not be excluded without further investigation. Effectively, higher leptin plasma levels were measured in LP offspring, and leptin plays an essential role in the regulation of energy homeostasis, food intake, and body composition (30).
We and others (48) have shown that fetal dietary restriction, depending on the type and timing of malnutrition, altered the endocrine pancreas and may lead to permanent reduction of β-cell mass and impairment of insulin secretion. It was also shown that protein-restricted rats (in utero and postnatal) displayed better glucose tolerance than control rats at young age. In parallel, these animals showed increased sensitivity to insulin, since plasma insulin concentrations were reduced (51). However, with age, the glucose tolerance of LP rats deteriorated more rapidly and by the age of 17 mo in males and 21 mo in females, their plasma insulin concentration doubled compared with the control (18, 43). In the present study, offspring from dams that were protein restricted only during fetal life, showed an increase in plasma glucose concentration at the age of 9 mo, but their insulin concentration was similar to that observed in the C group. In previous results obtained in rats submitted to catch-up growth after a fetal protein restriction, no changes in plasma glucose concentration were observed, while supplementation after weaning with a cafeteria diet increased plasma glucose levels. In addition, insulin concentration was also increased by the cafeteria diet as in the nonsupplemented LP rats. Nevertheless, the intraperitoneal glucose tolerance test in rats indicated no difference due to early nutrition in insulin response after glucose load (3). However, the discrepancy in metabolic circulating parameters observed between this study and results obtained by Bieswal et al. (3) could be explained by the nutritional status of animals when taking the blood samples. Circulating parameters have been shown to change according to fasting or postprandial state, and glucose and insulin concentrations are susceptible to variations according to the nutritional status (9).
Surprisingly, despite the supplementation with the hypercaloric diet, C-HC and LP-HC animals have reduced plasma triglyceride concentration. This might reflect a higher triglyceride uptake from the circulation. Consistent with this, a higher lipid content was observed in livers from HC-supplemented animals (data not shown). While the liver plays a major role in rodent lipid metabolism, a fraction of circulating triglycerides might be taken up by adipose tissue, and it could be expected that LPL play an important function in the assimilation of plasma triglycerides into fat cells. In our study, as shown by the ADIPOCHIP analysis, there is an increase in the abundance of LPL mRNA in the adipose tissue of C-HC mice and this increase is even greater for LP-HC mice. In the study by Guan et al. (24), it was shown that maternal protein restriction during pregnancy raises LPL mRNA in the offspring adipose tissue. However, those researchers did not report any change due to maternal malnutrition in plasma triglycerides of adult rats.
Total cholesterol concentration was higher in LP animals, like in hypercaloric-supplemented groups. However, the analysis of HDL fraction of plasma cholesterol indicated an increased concentration due to HC supplementation, while no influence was observed due to maternal diet. We might point out here that the hypercaloric diet, certainly provides more fat (maximum 32%) than the lab chow but the HC diet is derived from corn oil without any contribution of exogenous cholesterol. Therefore, we might expect an influence of early nutritional environment on cholesterol biosynthesis and transport, but further work should be carried out to verify this hypothesis.
A novel observation of the current work is that early nutritional mismatch downregulated the expression at the mRNA level of several genes involved in lipid metabolism in adipose tissue (Glut4, GPDcyto, GPAT, DHAPAT, CPT-2, FAS) (see online Supplemental Fig. 1 for illustration of metabolic pathway). In rodents as in humans, the adipose tissue is a secondary site for de novo lipid synthesis, and lipogenesis occurs mainly in the liver (2). In both sites, the process uses glucose as the principal carbon source, and in adipose tissue, Glut4 mainly regulates its transport. Gardner et al. (21) showed a decrease in Glut4 expression in adipose tissue of adult sheep after maternal undernutrition and associated it to insulin resistance. GPD activates glycerol into glycerol-3P to enter the adipocyte triacylglycerol (TAG) synthesis. Then, glycerol-3p acyltransferase esterifies the alcohol functions converting glycerol-3P into lysophosphatidic acid. Finally, dihydroxy oxyacetone-P acyltransferase participates to the transformation of dihydroxyacetone phosphate also into lysophosphatidic acid (8, 33). On the other hand, FAS generates long-chain fatty acids by the condensation of acetyl CoA and malonyl CoA, while the oxidation of fatty acids occurs in mitochondria and is regulated by the carnitine palmitoyl transferase system (CPT-1 and CPT-2) (19). The downregulation of these six enzymes suggests that the LP offspring has an altered lipid metabolism reducing the de novo synthesis of TAG. Furthermore, the microarray analysis also showed an increase of FABP (aP2) and Cav-1 expression in the LP offspring. The first one is involved in fatty acid uptake and intracellular transport, and the second one, a caveola protein, plays a role in plasma membrane binding of fatty acids and in their transport to lipid droplets (13, 47). Whereas a decrease in enzymes involved in lipogenesis is not consistent with the increased fat mass observed in the LP offspring, an increase in adipocyte proteins involved in fatty acid uptake may reflect a programming of lipid transport and might contribute to hypertrophy of fat cells observed in the LP offspring. Furthermore, these animals presented a second population of smaller adipocytes suggesting the recruitment of fat cell precursors present in adipose tissue. During obesity development, the adipose tissue expansion is initially marked by fat cell hypertrophy. Nevertheless, observations in genetic and diet-induced obesity also showed an increase in fat cell number. Moreover, hyperplasia in the adipose tissue is associated with the more severe forms of obesity (12, 29, 53).
Finally, among genes presenting a decreased expression due to fetal diet, we could also find the sterol response element binding protein (SREBP-1c). This transcription factor is involved in preadipocyte differentiation and regulates the expression of several enzymes involved in lipogenesis, such as FAS (31). In a recent publication, authors report a decrease in SREBP-1c hepatic expression due to LP diet, highlighting the effect of fetal protein restriction on the programming of lipid metabolism (14).
We have also shown that the plasma adipokine's levels were raised when animals were supplemented with a hypercaloric diet, suggesting a relationship between high body weight and adipokine plasma concentrations. However, we also observed an influence of the maternal diet on the leptin plasma concentration. Very few data are available on the possible induction of different plasma adipokines by fetal or early programming. Nevertheless, it has been shown that protein restriction during gestation results in a significant maternal, but not fetal, decrease in leptin plasma concentration during late gestation (17). In addition, exposure of rats to maternal severe energy restriction during gestation leads to an increase in relative adiposity and high circulating leptin concentrations in the adult offspring (57). It has been demonstrated that leptin mRNA levels were increased in adipose tissue of protein-restricted offspring (24). In our experiments, microarray analysis showed an increase in leptin mRNA level in LP-HC adipose tissue but not in LP-C animals. Nevertheless, the analysis by Western blot of leptin protein in adipose tissue confirmed that the perinatal malnutrition has an influence on the abundance of this adipokine.
In conclusion, this work emphasizes the effect of an early mismatched environment induced by forced catch-up growth subsequent to fetal protein restriction on the development of obesity in mice. We demonstrated significant alterations in the expression of genes playing a major role in the differentiation and/or function of the adipose tissue that are involved mainly in lipid metabolism. Of course, it would be worthwhile to validate these changes at the level of protein and or enzymatic activity. We might postulate that these changes could operate as part of combined processes that lead to programming of adipose tissue function and reflect an adaptive response of the fetus that is jeopardized during lactation by the adverse dietary context of overnutrition and amplified in adult life by hypercaloric diet.
- Copyright © 2009 the American Physiological Society