Mesoderm-specific transcript is associated with fat mass expansion in response to a positive energy balance

Abstract

A 50-fold variation in mRNA and protein levels of the mesoderm-specific transcript gene (Mest) in white fat of C57BL/6J (B6) mice fed an obesogenic diet is positively correlated with expansion of fat mass. MEST protein was detected only in adipocytes, in which its induction occurred with both unsaturated and saturated dietary fat. To test the hypothesis that MEST modulates fat mass expansion, its expression was compared to that of stearoyl CoA desaturase (Scd1) in B6 mice exposed to diets and environmental temperatures that generated conditions separating the effects of food intake and adiposity. Under a range of conditions, Mest expression was always associated with variations in adiposity, whereas Scd1 expression was associated with the amount of saturated fat in the diet. Mest mRNA was expressed at its highest levels during early postnatal growth at the onset of the most rapid phase of fat mass expansion. MEST is localized to the endoplasmic reticulum/Golgi apparatus where its putative enzymatic properties as a lipase or acyltransferase, predicted from sequence homology with members of the α/β fold hydrolase superfamily, can enable it to function in lipid accumulation under conditions of positive energy balance. Variations in adiposity and Mest expression in genetically identical mice also provides a model of epigenetic regulation.—Nikonova, L., Koza, R. A., Mendoza, T., Chao, P.-M., Curley, J. P., Kozak, L. P. Mesoderm-specific transcript is associated with fat mass expansion in response to a positive energy balance.

Variability in diet-induced obesity among genetically identical C57BL/6J (B6) mice has characteristics suggesting that the phenotypic differences are determined by an underlying epigenetic mechanism (1,2,3,4). Not only is adiposity highly variable, but so is the expression of genes in adipose tissue for which plausible roles in the regulation of adiposity have been proposed (3). These genes include secreted frizzled related protein 5 (Sfrp5), an inhibitor of Wnt signaling whose other family members (i.e., Sfrp2) have been shown to repress the Wnt signaling pathway and permit the entrance of preadipocytes into the adipogenic pathway (5, 6); bone morphogenetic protein 3 (Bmp3), which represses bone formation when inactivated (7); and plasminogen activator inhibitor (Serpine1), with potential involvement in increasing the vasculature required for adipose tissue expansion (8). In contrast to these genes, the well-known transcription factors of adipogenesis are essentially invariant, as are other biomarkers of adipocytes, including enzymes of lipogenesis (3). However, the most variable expression, with mRNA levels ranging over 50-fold in adipose tissue, is found for the gene encoding mesoderm-specific transcript (Mest). In B6 mice with variable diet-induced obesity, Mest expression is positively correlated with both adiposity and genes of Wnt and Bmp3 signaling.

Mest has also been reported in other studies to be overexpressed in adipose tissue from obese mice (9). Transgenic mice in which Mest is overexpressed from the aP2 promoter show enlarged adipocytes; however, no increase in obesity was observed (10). Mice with gene-targeted inactivation of Mest exhibit growth retardation and behavioral defects, but no effects on adiposity were noted (11). These phenotypes of Mest knockout mice, together with the fact that Mest is an imprinted paternally expressed gene (12, 13), suggest that Mest has an important role in development; however, its function in adipose tissue expansion is unclear. Although its enzymatic/biochemical function has not been established, MEST is a member of a α/β fold hydrolase superfamily, whose members also include lipases, acyltransferases, and esterases (14, 15). MEST has the highly conserved catalytic triad, serine 145-histidine 146-aspartate 147, within the conserved sequence motif for lipases and serine proteases (accession number NP_032616); thus, it is possible that MEST has a role in lipid metabolism, even though its specific enzymatic activity has not yet been identified.

Our previous data showing that expression of Mest in fat biopsies correlated with increased adiposity, even before mice were switched to a high-fat diet, has led to the hypothesis that Mest plays an essential role in the expansion of adipose tissue during positive energy balance. The alternative hypothesis is that Mest is regulated by dietary fat. To test these hypotheses, we have dissociated changes in adipose tissue expansion from dietary input by manipulating diet and environmental temperature in B6 wild-type mice. Additional tests of the hypotheses were made through the analyses of Mest and stearoyl CoA desaturase (Scd1) knockout models. These experiments have led to a model for adipose tissue expansion in mice with a positive energy balance whereby MEST, located in the endoplasmic reticulum (ER), facilitates the uptake of lipid from the circulation through its enzymatic function on lipid metabolism and/or accumulation.

MATERIALS AND METHODS

Animals

B6 mice for experiments were bred at the Pennington Biomedical Research Center from breeding pairs obtained from The Jackson Laboratory (Bar Harbor, ME, USA). Animal rooms were maintained at 22–24°C with a 12-h light-dark cycle. Breeders were housed in plastic pens with corncob bedding and fed a breeder diet ad libitum (LabDiet 5015; PMI, St. Louis, MO, USA) From weaning until 8 wk of age, mice were fed a low-fat chow diet ad libitum. (LabDiet 5053; PMI). (Composition of diets 5015 and 5053 is available from http://www.labdiet.com.) At 8 wk of age mice were fed ad libitum a high-saturated fat diet (D12331; Research Diets, New Brunswick, NJ, USA) (3). Mice with a spontaneous mutation disrupting stearoyl CoA desaturase (Scd1; B6;D1Lac-Scd1^ab-2J/J) were obtained from The Jackson Laboratory. The Mest knockout (KO) mice on a mixed B6/129 genetic background were created and maintained at Cambridge University (Cambridge, UK) (11), and tissues were sent to the Pennington Biomedical Research Center for analyses.

Phenotyping

Adiposity was determined from body weights and measurements of body composition by nuclear magnetic resonance (NMR) (Bruker, The Woodlands, TX, USA). Gene expression analyses by quantitative reverse transcriptase-polymerase chain reaction (qRT-PCR) with TaqMan probes and primers were performed as described previously (3). Sequences of probes and primers for qRT-PCR are listed in Supplemental Table 1. In some experiments, because of variations in adipose tissue cyclophilin B that were observed during development, gene expression was normalized by the total amount of RNA in each reaction as measured by a NanoDrop analyzer (NanoDrop Technologies, Wilmington DE, USA). Recombinant MEST was made in Escherichia coli strain Rozetta 2 (DE3), purified, and used to prepare polyclonal antibody by immunizing New Zealand White rabbits (see Supplemental Data for experimental details).

Adipocyte cell size

Epididymal adipose tissue was fixed in formalin and embedded in paraffin, and 10-μm sections were stained with eosin/hematoxylin. Adipocyte cell size was estimated by counting the number of adipocytes within a microscopic field of known area. Data from this simple method agreed well with data based on analysis with Metamorph imaging software and were more rapidly obtained when tissues from a large number of animals were analyzed.

Preparation of tissue lysates for immunoblots

Frozen tissues were homogenized in 5 vol of RIPA/DOC lysis buffer [50 mM Tris (pH 7.4), 150 mM NaCl, 1 mM EDTA, 0.1% SDS, 1% Nonidet P-40, and 0.25% sodium deoxycholate, supplemented with 1 mM PMSF and protease inhibitor cocktail tablets (catalog no. 11 836 170 001; Roche Applied Science, Indianapolis, IN, USA)]. Laemmli sample buffer was added to the homogenate, and samples were heated at 95–100°C for 10 min (16). After centrifugation at 10,000 g for 15 min at room temperature, cleared supernatants were analyzed by SDS-PAGE followed by Coomassie blue staining. Equal amounts of protein (20 μg) were used for each sample analyzed.

Subcellular fractionation

Inguinal (ING) adipose tissue from mice fed a high-fat diet for 16 wk was homogenized in 3 vol of TES buffer (50 mM Tris-HCl, pH 7.5; 1 mM EDTA; and 0.25 M sucrose), supplemented with 1 mM PMSF in a Teflon-glass homogenizer, and centrifuged at 1300 g for 15 min. The infranatant was aspirated from the pellet and the overlaying fat layer and centrifuged at 14,000 g for 20 min to generate a pellet. The 14,000 g supernatant was centrifuged at 32,000 g for 20 min to recover high-density microsomes. The supernatant was again centrifuged at 100,000 g for 90 min to separate low-density microsomes from soluble proteins. All pellets after differential centrifugation were suspended in 0.1 vol of Tris/Mg buffer (50 mM Tris-HCl, pH 7.5, 3 mM MgCl₂, and 1 mM PMSF). The 1,300 g pellet (nuclear fraction) was transferred into a new tube and washed in Tris/Mg buffer. The nuclear proteins were extracted with 0.1 vol of 0.4 M NaCl and 50 mM Tris-HCl, pH 7.5, followed by an incubation on ice for 30 min. After centrifugation at 10,000 g at 4°C for 15 min, the supernatant (nuclear proteins) was analyzed by SDS-PAGE. The fat layer from the first centrifugation was washed 3 times in 10 vol of Tris/Mg buffer. The fat layer proteins were extracted with an equal volume of RIPA/DOC lysis buffer. The mixture was warmed to 42°C for 30 s, mixed vigorously for 1 min, and kept at room temperature for 30 min. After centrifugation at 10,000 g at 4°C for 15 min, the infranatant was aspirated (lipid droplet protein fraction 1), and the upper fat layer was extracted a second time with an equal volume of RIPA/DOC lysis buffer at 37°C for 30 min. After centrifugation at 10,000 g at 4°C for 15 min, the infranatant (lipid droplet protein fraction 2) was recovered. Protein concentrations were determined by the Bradford or DC protein method (Bio-Rad Laboratories, Hercules, CA, USA). Equal amounts of protein (10 μg) for each sample were mixed with 1/6 vol of 6× Laemmli sample buffer, heated at 95°C for 10 min, and analyzed on a 10% SDS-PAGE gel (16).

Immunoblots

The Odyssey Infrared Imaging System (Li-COR, Lincoln, NE, USA) was used for Western blot analysis. Approximately 20 μg of protein samples was separated by 10% SDS-PAGE and transferred to polyvinylidene difluoride Western blotting membranes. The blots were incubated in stripping solution (100 mM 2-mercaptoethanol, 2% SDS, and 62.5 mM Tris-HCl, pH 6.8) at 55°C for 30 min. After washing in water for 30 min and in Tris-buffered saline (10 mM Tris-HCl and 0.9% NaCl, pH 7.5) for three 10-min washes, the membranes were incubated for 1 h with blocking solution from Li-COR and then incubated overnight at room temperature with specific rabbit antiserum against MEST diluted 1:1000 in blocking solution. Rabbit antibody against recombinant MEST was prepared in our laboratory as described in the Supplemental Data. Blots were also analyzed for glyceraldehyde-3-phosphate dehydrogenase (mouse anti-GAPDH monoclonal antibody diluted 1:5000; Abcam, Cambridge, UK); perilipin (rabbit polyclonal antibody diluted 1:1000; Research Diagnostics, Concord, MA, USA) and calnexin (rabbit polyclonal antibody, diluted 1:2000; Stressgen, San Diego, CA, USA) followed by reactions with corresponding secondary antibodies according to the manufacturers’ protocols.

Statistics

Statistical analyses were performed by ANOVA for multiple groups and Student’s t test for paired comparisons using Statview (1999; SAS Institute, Cary, NC) and Microsoft Excel (2003).

RESULTS

Expression of adipose tissue Mest is associated with variable adiposity in B6 mice

Global analysis of RNA in adipose tissue from mice with low and high weight gain after being fed dietary fat showed that the largest difference in gene expression was for Mest, an imprinted gene also known as paternally expressed gene 1 (Peg1) (3). Figure 1A shows the growth curves of a cohort of 112 male B6 mice after being fed a high-fat diet containing 58 kcal % saturated fat for 4 wk starting at 8 wk of age. Under these conditions increased adiposity is readily detected, but it is not excessive, thereby reducing the effects of obesity on metabolic parameters such as insulin resistance. In a separate group of 28 mice fed a high-fat diet (58 kcal % in saturated fat; Research Diets 12331) for 4 wk starting at 8 wk of age, except for one animal with a blood glucose level of 130 mg/dl, levels in all others were <95 mg/dl (data not shown). Body weights of mice after being fed the high-fat diet ranged from 25.5 to 37.3 g, with the 15th percentile of mice with the highest final body weight consistently accumulating ∼2.4 g/wk, whereas the 15th percentile of mice with the lowest final body weight gained just a little more than 1.2 g/wk during the 4 wk they were fed the high-fat diet. Measurements of changes in body composition of the mice with final body weights in the highest and lowest 15th percentiles revealed significant differences in adiposity between groups (0.37±0.01 vs. 0.23±0.01, respectively; P<0.0001) as measured by fat mass/lean mass (FM/LM) ratios, indicating that differences in body weight are primarily due to greater variation in adipose tissue mass relative to lean mass. Mest mRNA levels measured in ING, retroperitoneal (RP), and epididymal (EPI) fat pads using qRT-PCR varied more than 50-fold, and expression in each fat depot correlated significantly with the adiposity index (FM/LM ratio) (Fig. 1B). However, hypothalamic (HYP) Mest expression was relatively invariant (3.7-fold range) and did not correlate with adiposity (Fig. 1B). A summary of body composition and Mest expression in Table 1 reveals the following points.

Figure 1.

Variation of body weight, adiposity, and Mest mRNA expression in male B6 mice fed a high-fat diet (HFD). A) Growth curves of 112 mice are shown from weaning (3 wk of age) until 12 wk of age. Mice were fed a high-fat diet (58 kcal % saturated fat) for 4 wk starting at 8 wk of age. B) Strong associations occur between adiposity and Mest expression measured in ING, EPI, and RP fat but not in HYP. C, D) Adiposity indices as measured by the FM/LM ratio of mice are associated with adipocyte size.

TABLE 1.

Summary of regression analyses comparing Mest expression in ING, RP, and EPI fat and HYP with measurements of body weight, LM, FM, and adiposity (FM/LM ratio)

Measurement	Regression analysis (R)
	ING fat Mest	RP fat Mest	EPI fat Mest	HYP Mest
Body weight (3 wk)	0.20*	0.19*	0.14	0.05
Body weight (8 wk)	0.07	0.07	0.11	0.09
Body weight (12 wk)	0.30#	0.24*	0.24*	0.02
LM (8 wk)	0.01	0.01	0.09	0.04
FM (8 wk)	0.21*	0.26#	0.20*	0.19*
FM/LM ratio (8 wk)	0.21*	0.27#	0.16	0.18
LM (12 wk)	0.01	0.09	0.12	0.02
FM (12 wk)	0.50$	0.48$	0.45$	0.01
FM/LM ratio (12 wk)	0.55$	0.52$	0.51$	0.01
ING fat Mest	1	0.78&	0.68&	0.06
RP fat Mest	0.78&	1	0.81&	0.16
EPI fat Mest	0.68&	0.81&	1	0.11
HYP Mest	0.06	0.16	0.11	1

Data were collected from 112 male B6 mice before, during, and after feeding a high-saturated fat diet for 4 wk starting at 8 wk of age. Age of mice at the time of each measurement is indicated in parentheses. Adipose tissue and hypothalamic Mest mRNA was measured at the completion of the dietary feeding regimen. 

^*P < 0.05; 

^#P < 0.005; 

^$P < 10⁻⁶; 

^&P < 10⁻¹⁵. 

1) Except for marginal significance between hypothalamic Mest and fat mass at 8 wk of age (r=0.19), no other correlations involving hypothalamic Mest were significant.

2) No significant correlations were found between Mest and lean body mass under any condition or at any time.

3) Weak, but significant, correlations were present between fat mass and Mest in all fat depots at 8 wk of age, before the mice were fed a high-fat diet and with body weight at 3 wk of age. The data suggest that the effects of Mest are not dependent on a high-fat diet but are accentuated by the increase in a positive energy balance, which is stimulated by a high-fat diet.

4) The strong R values (R ranging from 0.68 to 0.81) between Mest expression in different fat depots, as well as the high correlations of Mest with adiposity (FM/LM ratio) in mice fed a high-fat diet, indicate that the adiposity index of an animal is determined by a mechanism that affects adiposity systemically and is not restricted to a specific regional fat depot.

Microscopic examination of tissue sections of ING fat from mice fed a high-fat diet for 8 wk indicated that mice with a lower adiposity index have smaller adipocytes (Fig. 1C). Estimates of adipocyte size in a cohort of 50 B6 mice with a range of adiposity indices showed a very strong association (r=0.81, P<10⁻¹¹) between adiposity indices and adipocyte size (Fig. 1D). The data suggest that variation in adipocyte size is a major subphenotype determining variability to diet-induced obesity.

An antibody generated in rabbits against bacterially expressed recombinant murine MEST was used to determine whether MEST protein levels correlate with mRNA expression in adipose tissue (see Supplemental Data for a description of antibody production in rabbits from recombinant MEST generated in a bacterial expression system). Western blot analyses using ING fat protein extracts from B6 and Mest knockout mice demonstrated specificity of the antibody to MEST (Fig. 2A) and showed that MEST protein was predominantly expressed in adipose tissue, but not in other tissues, including brown adipose tissue (Fig. 2B). Several tissues show bands of higher and lower molecular weight, but only white fat gives a prominent protein band of ∼35 kDa that is not detected in Mest KO mice (Fig. 2A). Mest belongs to a large family of α/β hydrolases, and it is possible that some of these other proteins (15), detected with the polyclonal antibody, are other members of the family. The specific relationship between Mest and adiposity is also apparent at the protein level as shown by the immunoblot analysis (Figs. 3A, B), and similar to the mRNA data, a correlation was found between adiposity and MEST protein (Fig. 3C). The pattern of expression among the tissues of the adult mouse and the high association between adiposity and MEST protein levels has led us to hypothesize, in agreement with Takahashi et al. (10), that Mest has a role in the function of the adipocyte, specifically in its capacity to store lipid. Although perilipin is a component of the lipid body in adipocytes and has an important role in regulating hormone-sensitive lipolysis (17,18,19,20), the levels of perilipin protein show no correlation with levels of adiposity (Fig. 3A, B). Although studies of Mest function by gene targeting experiments suggest that it has an important role in early morphological development and maternal behavior (11), exactly what that function may be at the biochemical level is unknown.

Figure 2.

Western blot analysis confirms the presence of MEST protein in white adipose tissue of mice. A) Analysis of EPI fat protein in mice with a targeted deletion of Mest shows no detectable band for the MEST protein. The numbers refer to individual wild-type (WT) and MEST KO mice. B) MEST protein can be readily detected in EPI and ING fat but not in brown adipose tissue (BAT) or other tissues of mice. The numbers refer to expression in individual mice to emphasize that the variability among individuals occurs in each adipose tissue.

Figure 3.

Western blot analysis shows a strong association between ING fat MEST expression and adiposity among mice. A) Adiposity indices (FM/LM ratio) and ING fat Mest mRNA expression are compared with MEST protein levels in 8 mice fed a high-fat diet for 12 wk. B) Western blot comparing the association of ING fat MEST and perilipin in mice with a range of adiposity indices (FM/LM ratio). C) GAPDH-normalized optical densities for protein bands in B show strong associations between adiposity of individual mice to MEST but not to perilipin. Numbers at the top of each blot refer to expression in individual mice, and the data below each blot refer to the adiposity index (FM/LM ratio) and levels of MEST mRNA (A, B). AU, arbitrary units.

Mest induction: a high-fat diet or adipose tissue hypertrophy?

The correlations between adiposity and Mest (Table 1, Figs. 1and 3) can be interpreted to show that Mest gene expression is increased in association with increased fat cell mass or that Mest gene expression is induced as a consequence of an increased flux of fat through the animal on the introduction of a diet high in saturated fat. Because Scd1 is induced in the presence of a high-fat diet through removal of suppression of transcription by unsaturated fat and has been associated with the development of obesity in mice, we have carried out several experiments to assess the behavior of Mest gene expression in wild-type mice and mice carrying the mutant genes for Scd1. Young adult mice fed a chow diet have low levels of Mest expression that are induced within 2 days of being fed a high-fat diet (Fig. 4A) and this induction occurs with either saturated or unsaturated fat in the diet (Fig. 4B). In contrast, Scd1 expression is only induced in the presence of saturated fat at either a low or high concentration (Fig. 4C). This result is in contrast to previously published data suggesting that expression of Scd1 was not suppressed by unsaturated fat in adipose tissue (21). These acute affects of dietary fat on adipose tissue Mest are actually accompanied by increases in fat mass that can be measured by NMR (Fig. 4D). It is conceivable that microphysical changes, which could affect cellular functionality including gene transcription in vivo, may be in place only hours after feeding of a high-fat diet as illustrated by the fusion of nascent lipid droplets by the action of phospholipase D1 and Erk2 on dynein phosphorylation to form large lipid droplets in a cell-free system in vitro (22, 23). Because the B6 mice become obese on diets high in either saturated or unsaturated fat (Fig. 4E), the association between induction of Mest and increased adiposity still holds, whereas induction of Scd1 is dependent on the presence of saturated fat in the diet. Figure 4F shows that an accumulation of MEST protein as well as its mRNA occurs after acute feeding of a high-fat diet. Thus, excessive adiposity is not a requirement for elevated Mest expression.

Figure 4.

Adipose tissue Mest expression is rapidly induced by feeding mice diets high in saturated (SF) or unsaturated fat (UF). A) The range of Mest expression in ING fat is shown after mice were fed chow or a high-saturated fat diet (HFD) for 2 or 7 days. B) ING fat Mest is induced after mice were fed for 2 days with diets high in saturated or unsaturated fat. C) Scd1 expression is elevated only with diets containing saturated fat. D) Changes in adiposity observed in mice fed diets high in saturated or unsaturated fat for 2 days are associated with changes in ING fat Mest expression. E) Male B6 mice show similar rates of weight gain when fed diets containing high saturated or unsaturated fat. F) MEST protein in ING fat of mice is induced after mice were fed a high-fat diet for 7 days. A minimum of 10 mice were used for each group (A–E). Numbers at the top of the immunoblots refer to the individual mice fed either a chow diet or a high-fat diet (F). Data are presented as the mean ± se groups with the same letter above each column do not differ significantly; Student’s t test (two-sample, assuming unequal variances).

Scd1 KO mice

When Scd1 mRNA levels were measured in the 112 B6 mice after 4 wk of feeding with a high-saturated fat diet, there was no consistent relationship with adiposity; the correlation of RP fat Scd1 mRNA expression and adiposity (FM/LM ratio) was negative with an R² of 0.267 and that with ING Scd1 mRNA was positive but with a R² of 0.058 (data not shown). Furthermore, the overall variability of Scd1 expression (ING fat ∼5.3-fold; RP fat ∼2.2-fold) was far less than that observed for Mest. Thus, high Scd1 expression in this model does not correlate with the development of obesity and mechanisms mediating Mest expression via dietary fat are different from those regulating Scd1 expression (24). However, it has been shown that Scd1 KO mice are resistant to diet-induced obesity (25). Accordingly, we have analyzed the expression of Mest in Scd1 KO mice in the context of diet-induced obesity (Fig. 5A, B). These results confirm that Scd1 KO mice are resistant to diet-induced obesity (25); in addition, they also show that Mest expression in these mice, which are not obese, is only 17% of that observed in wild-type controls (Fig. 5C). Thus, in the absence of obesity Mest induction is attenuated, despite the consumption of a high-fat diet. The expression of Sfrp5 was attenuated similarly to that of Mest (Fig. 5D). The absence of Scd1 mRNA confirmed the genotype of the putative Scd1 KO (Fig. 5E).

Figure 5.

Resistance to diet-induced obesity in Scd1 knockout mice after consumption of a high-fat diet (HFD) is associated with repressed adipose tissue Mest and Sfrp5 expression. Body weights (A) and adiposity (B) of control B6 and Scd1 KO mice before and after consumption of a high-fat diet for 4 wk. EPI (C) and ING (D) fat Mest and Sfrp5 expression was significantly repressed in Scd1 KO mice after 4 wk of consuming a high-fat diet. E) Scd1 mRNA was not detected in adipose tissue of Scd1 KO mice, which confirms the genotype of the animals used in this experiment. *P < 0.05, and **P < 0.01; ≥8 mice/group; Student’s t test (two-sample, assuming unequal variances).

Mest KO mice

The original characterization of the Mest KO mouse showed that adult mice were smaller, but no data were presented on a phenotype related to adiposity (11). We have studied a group of 9-month-old Mest KO mice and wild-type littermate controls on a mixed 129/B6 genetic background and confirmed that these mice have reduced body weights, but they were not very different (+/+=32.52±1.10 g, n=10 and Mest KO=28.12±0.74 g, n=12; P = 0.002). The absence of Mest also reduced adiposity (+/+=0.78±0.15 g and Mest KO=0.37±0.03 g for weights of EPI plus ING fat depots in 9-month-old male mice) suggesting that Mest enhances the capacity for lipid storage in adipocytes. However, the significant levels of adipose tissue mass in the Mest KO mice indicate that Mest is not essential for adipogenesis.

Mest expression in cold-adapted mice fed a high-fat diet

The above experiments with wild-type B6 mice and those carrying targeted mutations for Mest and Scd1 indicate that the expression of Mest is more highly correlated with increased adiposity than with dietary fat. To test this hypothesis further, we designed an experiment in which adiposity was suppressed under conditions in which the intake of a high-fat diet was increased. Mice were weaned to a chow diet and housed at near thermoneutrality, i.e., 28°C, to suppress thermogenesis. At 8 wk of age, the mice were divided into the following four groups: group 1 was maintained on the chow diet and at 28°C; group 2 was fed the chow diet, but the ambient temperature was reduced to 4°C, resulting in an increase in energy intake from 10 to 25 kcal/day (Fig. 6A); group 3, fed a high-fat diet and housed at 28°C, showed a slight increase in energy intake to 13 kcal/day: group 4, fed a high-fat diet and housed at 4°C, increased energy intake to 23 kcal/day, which was not significantly different from the mice in group 2 (chow fed at 4°C). The largest change in body weight, due principally to increased fat mass, was observed for mice fed a high-fat diet and housed at 28°C (group 3), despite the very modest increase in energy intake (13 kcal/day). We predicted that those mice housed at 28°C and fed a high-fat diet would show an increase in adiposity because of lower substrate oxidation necessary to maintain body temperature, whereas those at 4°C would not increase their adiposity despite a greater intake of the high-fat diet. If Mest expression depended on the increased flow of saturated fatty acids through the system, then those mice maintained at 4°C should have induced levels of Mest expression, despite the suppression of adiposity by a reduced ambient temperature (group 4). On the other hand, if Mest expression is associated with higher adiposity, then the mice maintained at 28°C with lower food intake but with elevated adiposity will have higher levels of Mest expression (group 3). As predicted, the results in Fig. 6A show that mice fed a high-fat diet and maintained at 28°C have increased adiposity compared with those maintained at 4°C despite the greater food consumption in the latter. Mest expression in EPI and ING fat depots was also highest in the mice fed a high-fat diet and maintained at 28°C, despite lower intake of a high-fat diet (Fig. 6B, C). In contrast, adipose tissue Scd1 expression was high in mice fed a high-fat diet and maintained at either 28 or 4°C compared with mice fed chow and maintained at 28°C, indicating that Scd1 is induced by diets high in saturated diets and its expression is not related to the level of adiposity. An unexpected observation was the high expression of Scd1 in mice fed a chow diet and maintained at 4°C, suggesting that the Scd1 gene is induced by cold exposure. Although Scd1 KO mice are cold sensitive (26), the induction of Scd1 expression in white fat by cold exposure has not been observed previously to our knowledge.

Figure 6.

Cold exposure prevents dietary fat-induced increases in body weight and fat mass and inhibits dietary fat-mediated increases in adipose tissue Mest expression. Mice in this experiment were maintained at thermoneutrality (28°C) and fed a chow diet from weaning until 8 wk of age. Mice were then fed chow (groups 1 and 2) or a high-fat diet (HFD) (groups 3 and 4) for 1 wk; subsets of mice within each dietary regimen were maintained at 28°C (groups 1 and 3) or exposed to 4°C (groups 2 and 4). A) Effects of cold exposure and high-fat diet on food intake (kcal/day), body weight, lean mass, and fat mass. B, C) Differences in the expression of Mest, Scd1, Ucp1, and Sfrp5 in EPI and ING fat, respectively, for each of the groups. Groups with the same letter above each column do not differ significantly; *P ≤ 0.05, **P ≤ 0.001; ***P ≤ 0.0001; ≥9 mice/group; Student’s t test (two-sample, assuming unequal variances). NS, nonsignificant difference between groups.

Thus, Mest expression is linked principally to cellular mechanisms controlling expansion of the fat mass. The results also show that elevated Sfrp5 expression was consistently coordinated with mechanisms that control Mest expression and adipose tissue expansion (Fig. 6B, C). It is also noteworthy that Ucp1 expression, used as a marker to validate the thermogenic response of the animal, was several hundred-fold higher in ING fat than in EPI fat. The higher expression of Mest in EPI fat compared with that in ING fat, in which high brown fat induction occurs (27), further supports a role for Mest in lipid body expansion, a process that does not occur in brown adipocytes unless the animal is maintained for extended periods in ambient conditions that do not require activation of nonshivering thermogenesis. This evidence for the absence of the MEST protein in interscapular brown adipose tissue (Fig. 2B) indicates that Mest is one of the few genes known to be expressed at high levels in white adipocytes but not brown adipocytes.

Mest and fat mass expansion in developing mice

In this research we designed experiments to test the hypothesis that MEST is a component of a gateway mechanism that mediates the uptake and deposition of circulating fat into lipid droplets of the adipocyte during a positive energy balance. A requirement for this function is first encountered by the mouse with the onset of nursing on a high-fat diet provided by mother’s milk after birth. It has been estimated that the energy content from fat of mouse milk is ∼55 kcal % (28), a nutritional state capable of initiating the positive energy balance necessary to establish essential reserves of fat. After adipogenesis in the developing mouse, during which the fat anlage has been laid down, a process of fat accumulation must occur rapidly. The accumulation of total fat in the developing mouse showed that the highest accumulation of fat per day occurred between 5 and 10 days of age and then remained low, until the mice were fed a high-fat diet beginning at 56 days of age (Fig. 7). We predicted that variation in Mest expression would follow a similar pattern. Because the only white fat depot able to be dissected from the mouse between birth and 7 days of age is the ING fat depot, we have determined the expression profiles of Mest, together with Bmp3 and Sfrp5, from 2 days of age through to weaning, from weaning to 8 wk of age when mice are fed a low-fat chow diet, and then from 8 wk of age to 16 wk of age when they were fed a high-fat diet (Fig. 8). The highest levels of Mest expression were found in mice at 2 days of age after which they declined rapidly during lactation and remained at a very low level until the mice were again fed a high-fat diet at 8 wk of age. The overall expression profile of Bmp3 was similar to that of Mest, but levels were higher during periods of low fat chow feeding. However, in contrast to the high degree of coordinate expression between Mest and Sfrp5 during diet-induced obesity (3), Sfrp5 was essentially undetectable in the suckling mice and remained so until they were fed a high-fat diet as young adults (Fig. 8C). Surprisingly, compared with Mest, Bmp3, and Sfrp5, Pparg2 expression remained relatively invariant during the 112 days of the study (Fig. 8D).

Figure 7.

Fat deposition occurs rapidly in mice during early postnatal growth and after adult mice are fed a high-fat diet. Fat deposition (mg of fat/day/mouse) was measured by NMR during the time intervals (days) indicated on the figure. Mice were weaned and fed a chow diet at 21 days of age, then fed a high-fat diet from 56 to 112 days of age. The data for the 5–10 d, 10–21 d, 21–56 d, and 56–112 d groups were calculated from 19, 52, 24, and 12 male mice, respectively. Data are presented as the mean ± se. Groups with the same letter above each column are not significantly different; Student’s t test (two-sample, assuming unequal variances).

Figure 8.

Inguinal fat Mest and Bmp3 mRNA expression corresponds to fat mass expansion during early postnatal development and after mice were fed a high-fat diet. Male mice (≥12/group) were weaned and fed a chow diet at 21 days of age, then fed a high-fat diet from 56–112 days of age as indicated by gray bar. Mest (A), Bmp3 (B), Sfrp5 (C), and Pparg2 (D) expression was measured using qRT-PCR, and data were normalized by the total RNA in each reaction. Data are presented as the mean ± se for each time point; time points with the same letter do not differ significantly; Student’s t test (two-sample, assuming unequal variances).

Subcellular localization of MEST

The amino acid sequence of MEST places it in the α/β hydrolase superfamily whose members, such as lipases, acyltransferases, esterases, and epoxidases, have enzymatic activities (15). Although the exact enzymatic activity of MEST has not been determined, a biochemical function for MEST as a lipase or acyltransferase would be compatible with a function related to adipose tissue expansion. To explore the function of MEST further, its subcellular localization was determined in comparison with protein markers with known subcellular locations. The immunoblot in Fig. 9A shows that MEST is distributed in a manner indistinguishable from that of the ER marker, calnexin. For comparison, the distributions of GAPDH, a cytoplasmic marker, and perilipin, a lipid droplet marker, are provided. Although these results suggest that MEST is located in the ER; it is possible that MEST is located in the Golgi apparatus. Because such localization would be expected if MEST was a secreted protein, we tested for the presence of MEST in the plasma by immunoblot analysis. Although this is not a highly sensitive method, nevertheless, we easily detected adiponectin, a protein secreted by adipocytes via the Golgi, which is expressed at the mRNA level in inguinal fat at ∼2.5 times the level of Mest mRNA (Fig. 9B).

Figure 9.

A) MEST comigrates with molecular markers of the ER of adipocytes. Western blot analysis shows enrichment of MEST expression in subcellular fractions of ING containing the ER protein calnexin. The lipid droplet protein perilipin and the cytosolic protein GAPDH were predominantly detected in the lipid droplet fractions and the 100K cytosolic fractions (Cyt), respectively. The preparation of subcellular fractions of adipose tissue is described in the Materials and Methods section. B) Expression of adiponectin and MEST in plasma samples of mice at the indicated ages. A positive control for MEST was provided with recombinant MEST.

Perilipin has been used as a control in Western blots to illustrate variations of MEST protein as it was not associated with variations in adiposity (Fig. 3C). If Mest expression is occurring in response to the need to expand adipose tissue fat mass due to a positive energy balance, it is possible that other components of lipid body formation have elevated levels of gene expression to promote increased fat mass. Table 2 shows that the increase in Mest expression in high gainers compared with low gainers is much greater than that for any other genes implicated in formation of the lipid vesicle. This high sensitivity of MEST to the relative level of adiposity was also evident in the expression of the perilipin protein by immunoblot (Fig. 3B, C). These correlations between gene expression and adiposity do not address the importance of these genes to lipid body formation, only that variation in expression is independent of the variation in diet-induced obesity. It is interesting to point out that Vldr, which is involved in uptake of fatty acids from very low density lipoprotein (VLDL) and chylomicrons in the circulation, has significantly increased expression and that adipose tissue morphology for the Vldr KO mouse is very similar to that found for low gainers as shown in Fig. 1C (29).

TABLE 2.

Expression of genes of lipid body formation

Gene	HG/LG ratio	P
Mest (Peg1)	23.92	<0.0001
Sfrp5	3.96	<0.0001
Scd1	1.15	0.17
Dgat1	−1.20	0.01
Dgat2	−1.15	0.13
TIP47	1.15	0.08
adipophilin	1.00	0.49
perilipin	1.17	0.13
S3–12	1.60	0.0001
Erk2 (MapK1)	−1.10	0.08
phospholipase D1	1.14	0.23
caveolin 1	1.13	0.26
caveolin 2	1.09	0.18
vimentin	−1.12	0.23
Grp78 (Hspa5)	1.04	0.37
OXPAT/PAT1	1.00	0.50
Vldlr	1.55	0.0003
Ldlr	1.38	0.02

The expression of genes associated with lipid body formation with data obtained from the microarray experiment described in Koza et al. (3) Equal aliquots of RNA from 15 high gainers (HG) and 15 low gainers (LG) were pooled, and microarray analysis was performed on three replicates from the RNA pools. The HG/LG ratio represents the ratio of gene expression in the high gainers compared to the low gainers. P values indicate the significance of differences between the HG and LG on the basis of the technical replication. Excellent agreement was found between the quantification of microarray data and qRT-PCR data. (See Fig. 1 for MEST quantification by qRT-PCR.) 

DISCUSSION

The ability to expand adipose tissue mass in response to an obesogenic environment has been found to be a vital factor in the development of susceptibility to diseases of the metabolic syndrome. Based on the severe insulin resistance observed in lipodystrophic models (30,31,32), Danforth (33) was among the first to underscore the importance of efficiently storing fat in adipose tissue to prevent the insulin resistance that occurs when fat accumulates in liver, skeletal muscle, and pancreas. Although this phenotype suggested that accumulation of lipid in adipose tissue is a protective mechanism against insulin resistance, other observations on the accelerated production and secretion of the inflammatory cytokine tumor necrosis factor-α by adipocytes from obese individuals (34), together with the realization that inflammatory cytokines are involved in the development of insulin resistance (35, 36), indicated that excessive fat in adipose tissue also leads to insulin resistance and diseases of the metabolic syndrome (37). Therefore, mechanisms that control fat mass expansion are of considerable importance regarding the development of lipotoxicity in nonadipose tissues and the production of inflammatory cytokines and reactive oxygen species in fat depots. The results of this study indicate that MEST could have a unique role to play in the deposition of circulating fat into adipose tissue during a positive energy balance, initially during early development when the neonatal mouse first encounters mother’s milk as a neonate and then as an adult in an obesogenic environment.

Fat mass expansion depends on increases in adipocyte hypertrophy and/or proliferation in an obesogenic environment. The extensive studies on adipogenesis in 3T3-L1 preadipocytes (38, 39), together with the severe lipodystrophies found in gene knockout mice (40,41,42), have clearly established a key role for peroxisome proliferator-activated receptor γ (PPARγ) in adipose tissue expansion by increasing adipocyte cell number through de novo adipogenesis; however, it is less clear how PPARγ regulates adipocyte hypertrophy. Although Pparg2 and other transcription factors are certainly key players in a program that activates adipocyte-specific genes as part of the basic mechanisms to establish a functional adipocyte during development (43, 44), the mechanisms by which hypertrophy and hyperplasia are harnessed to store accumulating lipids in an obesogenic environment are still not well understood. A particularly insightful study recently showed that in the B6 model of diet-induced obesity, adipocyte hypertrophy seems to be the dominant mechanism for storing excess fat during the first 20 wk, and, thereafter, adipocyte hyperplasia under control of the F box Skp gene comes into play (45). Mice with an inactive Skp gene have normal adipocytes, similar to Mest^−/− mice but show resistance to diet-induced obesity. An important concept in this study is that a genetic system exists to control adipose tissue mass by increasing cell number after the expansion of fat mass by a hypertrophic mechanism appears to have been exhausted. We propose, from the studies described in this article and our previous study (3) as well as by the observations of Takahashi et al. (10) with Mest transgenic mice that MEST controls the initial phase of adipose tissue expansion at the onset of a positive energy balance by regulating adipocyte hypertrophy.

A high positive correlation between adiposity and Mest mRNA levels in diet-induced obesity in B6 male mice first suggested a role for MEST in adipocyte expansion; however, we were not able to determine whether Mest was induced by a diet high in saturated fat, whether its induction was secondary to the increase in adiposity, or whether it had a primary, causative function in the expansion of fat mass during a positive energy balance. In the experiments described in this article, we first tested the hypothesis that Mest was induced by the high-fat diet. Perturbations were produced in Mest expression by using a combination of wild-type mice and mice carrying inactivated genes for Scd1 and Mest together with variation in diets and environmental temperature to determine whether the induction of Mest is caused by a high-fat diet or whether it is primarily associated with increased fat mass expansion. These studies showed that without exception increases in adipose tissue expansion are accompanied by increased expression of Mest. On the other hand, increased Mest expression is not always associated with increased fat in the diet. In one experiment in which we manipulated the ambient temperature to reduce adiposity but increase dietary fat intake, we observed that Mest expression was higher in B6 mice fed a high-fat diet and maintained at thermoneutrality compared with mice fed the high-fat diet and maintained at 4°C. The latter resulted in increased consumption of a high-fat diet and a reduction of fat stores because of increased substrate oxidation to maintain body temperature. In contrast, if the ambient temperature is increased to 28°C, consumption of the high-fat diet is reduced, but both adiposity and Mest expression were increased. Thus, Mest expression is linked to adiposity, but Scd1 expression is linked to diets containing saturated fat. In another experiment, Scd1 KO mice were fed a high-saturated fat diet and shown to be resistant to the development of obesity as observed previously (25). In our experiment, reduced adiposity was associated with a drastic reduction in Mest expression (Fig. 5C), indicating that the high-fat diet is not sufficient for Mest induction. Rather the correlation is between the level of adiposity and Mest. Alternatively, it is possible that induction of Mest requires unsaturated fats that are normally produced by Scd1 but are missing in the Scd1 KO mice. However, adipose tissue of mice expresses high levels of Scd2 (21), which is fully capable of synthesizing unsaturated fatty acids. Therefore, reduced Mest expression is not likely to be determined by the inability of adipose tissue to synthesize unsaturated fatty acids in the absence of Scd1, and this finding once again underscores the strong correlation of Mest with fat mass.

Our experiments do not support the hypothesis that Mest expression is induced as a secondary consequence of an increase in adiposity as a result of a positive energy balance. Several experimental observations pertinent to this conclusion are as follows: 1) Mest expression is already higher in mice destined to be high gainers even before the introduction of a high-fat diet (3); 2) Mest becomes induced in adult mice within 2 days of being fed a high-fat diet, before significant changes in adiposity can be detected by NMR (Fig. 4); 3) the highest levels of Mest mRNA are found in the neonatal mouse at the onset of fat mass accretion before fat depots are established and then Mest mRNA levels gradually decline during the following days of development when fat mass is expanding at a slower rate (Fig. 8); and 4) adipose tissue in Mest KO mice has normal morphology, but at reduced levels. These experimental data suggest that MEST is a component of a mechanism in the ER facilitating the uptake of fat into the adipocyte for storage in lipid droplets.

In addition to its expression patterns under nutritional manipulations and during development, a role for MEST in fat mass expansion comes from its molecular characteristics. Foremost, MEST is localized to the ER, a site where the VLDL receptor and diacylglyceride acyl transferase 1 (DGAT1) function, together with lipoprotein lipase in the vascular endothelium, to take up fatty acids from the circulation and repackage them into triglycerides (TGs) for assembly into the lipid droplet. These functions of the ER in lipid storage suggest that the function of MEST in the ER may be similarly related to TG storage. Secondarily, the enzymatic role for MEST in the ER to facilitate storage of fat is suggested from putative enzymatic functions of α/β fold hydrolase proteins and, indeed, the presence of the catalytic triad at serine 145-histidine 146-aspartate 147 of MEST strongly suggest an enzymatic function for MEST as a lipase or acyltransferase (15, 47). Such an enzymatic function could, accordingly, supplement the capacity of DGAT1 or other lipase activities associated with TG metabolism in the ER to efficiently store excess calories in adipose tissue during a positive energy balance.

It is indicative of the potential importance of the α/β fold hydrolase family of proteins in lipid metabolism that the α/β fold CGI-58 protein has recently been found to an integral component of the lipid droplet where it mediates the accessibility of hormone-sensitive lipase (HSL) to the lipid droplet (48). Not only does CGI-58 regulate HSL, but also its binding affinity to adipose tissue triglyceride lipase (ATGL) is essential for the enzymatic activity of ATGL. However, unlike MEST, which has the active site structure necessary for an enzymatically functional protein, CGI-58 lacks the essential serine in the catalytic triad and therefore functions as an accessory/carrier protein.

In summary, the profound importance of imprinted genes for the normal development of mammalian organisms from the embryo to the mature fetus has been well established (http://www.mgu.har.mrc.ac.uk/research/imprinting/). On the other hand, a role for imprinted genes in the normal physiology of adult mammals is relatively unknown. In this article we have shown that the Mest gene is almost undetectable in somatic tissues of the adult mouse, until it enters a condition of positive energy balance after the feeding of a high-fat diet. Under these conditions induction of Mest occurs selectively in mature white adipocytes within a subpopulation of genetically identical mice; however, the induction of both adiposity and Mest is highly variable among individual mice. Microarray analyses of ∼33,000 targets showed that Mest overexpression in adipose tissue of high vs. low weight-gaining mice is greater than that of any other gene, including those associated with lipid body formation. Immunoblot experiments indicate that MEST is localized to the ER or Golgi apparatus where other proteins associated with expansion of the lipid body of the mature adipocyte are located. The fact that MEST is a member of the α/β fold hydrolase family with an intact catalytic triad in its active site suggests a function for MEST in adipose tissue expansion that is related to its enzymatic activity as a lipase or acyl transferase.