Objective: Investigation of the expression of sterol regulatory element binding protein-1c (SREBP-1c) in different adipose tissue depots in morbidly obese subjects before and after 1 year of weight loss induced by gastric banding operation.
Research Methods and Procedures: SREBP-1c expression was studied in 20 massively obese subjects (6 men and 14 women; age: 41 ± 9 years; weight: 148 ± 34 kg; percentage of body fat: 42 ± 4; mean ± SD) using reverse transcription competitive polymerase chain reaction. Adipose tissue biopsies were taken from omental, subcutaneous abdominal, and femoral depots before weight loss, and from subcutaneous depots after weight loss. Subcutaneous samples were taken also from 6 normal weight subjects.
Results: The level of SREBP-1c mRNA was significantly lower in omental (1.8 ± 0.2 amol/μg of total RNA) than in subcutaneous abdominal (3.7 ± 0.4 amol/μg of total RNA) or femoral (3.9 ± 0.4 amol/μg of total RNA; p < 0.001, mean ± SEM) depots. The values in subcutaneous depots were about twice as high in normal weight (7.4 ± 2.5 for abdominal and 6.5 ± 1.5 for femoral, p < 0.01) as in obese subjects. After weight loss, the mRNA levels of SREBP-1c increased in obese subjects, both in subcutaneous abdominal (5.3 ± 0.7, p < 0.01) and in femoral (4.8 ± 0.8, p < 0.05) tissue.
Discussion: SREBP-1c mRNA expression was lower in omental adipose tissue than in subcutaneous depots in obese subjects before weight loss. Furthermore, the expression of SREBP-1c in obese subjects was clearly lower than in normal weight subjects, but mRNA levels increased along with weight reduction. Weight reduction was associated with increased mRNA levels of SREBP-1c in obese subjects. The reduced expression of SREBP-1c in obesity could be ascribed to lowered action or concentration of insulin, changeable along with weight reduction. However, changes in SREBP-1c expression after weight reduction could also be ascribed to the changes in calorie intake or nutritional habits after gastric banding operation.
Sterol regulatory element binding proteins (SREBPs) are members of the basic helix-loop-helix family of transcription factors that have been associated with lipogenesis, adipocyte development (1), and cholesterol homeostasis (2). Two distinct genes encode SREBP-1 and SREBP-2 (3). SREBP-1 gene encodes two proteins, SREBP-1a and SREBP-1c that differ in the length of amino-terminal transactivation domain. SREBP-1a bears 29 additional amino acids and seems to be a more potent transcriptional activator than SREBP-1c, at least in cultured cells (4). However, SREBP-1c has been shown to be the major form of mRNA expressed in white adipose tissue in humans and mice as well as in cultured adipocyte cell lines (5). Independently cloned rat homolog of SREBP-1c is named as adipocyte determination and differentiation factor-1 (6). In mouse adipose tissue, SREBP-1c mRNA expression was dramatically reduced by fasting and elevated on refeeding (7). Furthermore, in cultured cells, SREBP-1c expression and activity have been found to be regulated by insulin (7) (8), leading to the hypothesis that it may be a major component of the transcriptional effects of insulin in target cells (9) (10). Regulation of the expression of fatty acid synthase, lipoprotein lipase, and leptin genes have been demonstrated to be mediated, at least partly, by SREBP-1c (7) (11) (12) (13). In addition to the regulation of these important genes of lipid metabolism, SREBP-1c has been shown to activate another important adipogenic transcription factor, peroxisome proliferator-activated receptor-γ by stimulating the production of its ligands (14) or by increasing its expression in adipocyte cell lines (15). Finally, in two recent studies using oligonucleotide microarrays (16) (17), it was shown, quite paradoxically, that the mRNA expression of SREBP-1c was decreased in adipose tissue of obese mice. Most of these results about SREBP-1c are based on studies of primarily rodent models and cultured cell lines, and, thus, little convincing data regarding the expression levels of SREBP-1c in human adipose tissue have been reported. Because SREBP-1c seems to be a highly important transcriptional factor in adipocytes, participating in lipogenesis, adipogenesis, and in the action of insulin, it is tempting to speculate that it may also play a role in the development or maintenance of human obesity. To reinforce this hypothesis, the aim of the present study was to investigate the expression of SREBP-1c mRNA in human morbid obesity in different depots of adipose tissue before and after long-term weight reduction.
Research Methods and Procedures
Twenty-eight massively obese subjects (10 men and 18 women) participated in the study. The subjects went through medical examinations before a gastric banding operation for weight reduction. Twenty subjects (6 men and 14 women) had normal fasting glucose concentrations, whereas 8 had significantly elevated fasting glycemia and had been previously diagnosed as type 2 diabetic patients. Among the 28 obese subjects, 20 (8 men and 12 women), of whom 4 men and 10 women were normoglycemic, participated in the 1-year follow-up study. Two of the subjects had a cardioselective β-blocking medication (atenolol) not known to affect adipose tissue metabolism. A 2-hour oral glucose tolerance test (OGTT) with a glucose load of 75 g was performed only in the normoglycemic obese subjects 7 to 56 days before and 1 year after gastric banding operation. In addition, six healthy normal weight subjects took part in the study as a control group. Normal weight subjects were healthy volunteers, who went through the same laboratory examinations as obese subjects did, except the OGTT. All subjects received both oral and written information of the study and they gave written consent. The Ethics Committee of the Kuopio University Hospital and the University of Kuopio approved the study.
The subjects visited the outpatient clinic of the Kuopio University Hospital before gastric banding operation (84 to 168 days before) and 1 year after it. At both visits, body composition was determined by a bioelectrical impedance method (Bioelectrical Impedance, Body comp II, version 1.5; RJL Systems Inc., Detroit, MI) in fasting condition. Adipose tissue samples for the determination of SREBP-1c mRNA concentrations were taken during the gastric banding operation from abdominal subcutaneous, omental, and femoral subcutaneous adipose tissue under general anesthesia, in the fasting state (after 12 to 14 hours of fasting). One year after surgery, abdominal and femoral subcutaneous adipose tissue biopsies were taken under local anesthesia at the outpatient clinic. Furthermore, abdominal and femoral subcutaneous adipose tissue biopsies were taken from normal weight subjects under the same procedures as for obese subjects. Both the obese and normal weight subjects were advised to fast at least 12 hours before the biopsy at the outpatient clinic, so that all biopsies (before and after weight reduction) were taken after a similar period of fasting. The samples were immediately frozen in liquid nitrogen and stored at −70 °C for later analyses.
Fasting blood samples for the analyses of serum glucose, leptin, free fatty acid, and plasma insulin concentrations were obtained in the morning after 12 hours of fasting. Serum glucose was analyzed by a glucose dehydrogenase method (Merck, Darmstadt, Germany). Serum free fatty acids were measured by a turbidometric analyser (Kone Ltd., Espoo, Finland). Plasma glucose and insulin were determined from fasting blood samples taken before OGTT and 120 minutes after glucose load. Plasma glucose concentration was analyzed using the glucose oxidase method (Daiichi Co., Kyoto, Japan). Commercial radioimmunoassay kits were used for the analysis of plasma insulin (Phasedeph insulin RIA 100; Pharmacia Diagnostics, Uppsala, Sweden) and leptin (Linco Research Inc., St. Louis, MO).
Preparation of Total RNA and Quantitation of Target mRNAs
For total RNA preparation, adipose tissue samples were pulverized in liquid nitrogen. Total RNA from the frozen powder was prepared using RNeasy total RNA kit (Qiagen, Hilden, Germany). The amount of total RNA was quantified spectrophotometrically at 260 nm. The ratio of absorption (260/280 nm) of all preparations was between 1.8 and 2.0. Total RNA was suspended into water and stored at −80 °C.
The levels of SREBP-1c mRNA was quantified by reverse transcription reaction followed by competitive polymerase chain reaction (RT-cPCR). Detailed description and validations of the method has been published previously (18). For the construction of a SREBP-1c-specific competitor DNA molecule, a 311 nucleotide-long cDNA fragment was synthesized by RT-PCR from human adipose tissue total RNA using 5′-8GCGGAGCCATGGATTGCAC11–3′ as sense primer (specific of the exon 1c of the SREBP-1 gene) and 5′-311CTCTTCCTTGATACCAGGCCC291–3′ as antisense primer. The competitor was obtained by adding 20 bp in the SREBP-1c cDNA fragment by PCR-mediated mutagenesis and subcloned in the pGEM phagemid (Promega, Charbonnières, France). To validate the RT-cPCR assays, known amounts of in vitro synthesized RNA (Riboprobe System; Promega) corresponding to the SREBP-1c cDNA fragment, were quantified by RT-cPCR, as previously recommended (18).
For the assay of SREBP-1c mRNA, the specific first-strand cDNA was synthesized from 0.1 μg of total RNA with 2.5 U of thermo stable reverse transcriptase (Tth DNA polymerase; Promega) and with the specific antisense primer indicated above in conditions that warrant optimal efficiency of the reverse transcription reaction (18). To improve the analysis of the PCR products, the sense primer (sequence indicated above) used during the cPCR was 5′-end labeled with the CY-5 fluorescent probe. The PCR products were analyzed in denaturating 4% acrylamide-gel (Ready-mix; Pharmacia Diagnostics) using an ALF Express DNA sequencer (Pharmacia Diagnostics). The fluorescence of the target and competitor cDNA pics were evaluated using the Fragment Manager software (Pharmacia Diagnostics), and the initial target mRNA concentration was calculated at the competition equivalence point (18).
All calculations were performed using the SPSS/WIN program, version 9.0 (SPSS Inc., Chicago, IL). Results are given as means and SDs unless otherwise stated. Wilcoxon nonparametric paired t test was used for studying differences among different adipose tissue depots at baseline as well as the differences within the adipose tissue depot before and after weight reduction. The Mann–Whitney nonparametric t test was used for studying the difference in mRNA level between the genders. The difference in the change of mRNA levels along with weight loss between genders was studied using general linear model for repeated measures. Fat mass (in kilograms) was used as covariate when necessary. Spearman's nonparametric correlation analysis was used for studying the associations between factors. When the correlation was corrected with fat mass (in kilograms) or the percentage of body fat, the partial correlation analysis was used. Bonferroni's correction was used because of multiple correlation analysis.
Anthropometric and metabolic characteristics of the nondiabetic obese subjects before (n = 20) and after (n = 14) 1-year weight reduction are shown in Table 1. The mean weight reduction was ∼16% (−21% for men and −14% for women), but the subjects were still obese after the 1-year weight reduction. The percentage of fat mass was significantly greater in women before weight reduction compared with men. As expected, serum leptin concentrations were lower in men, both before and after weight reduction.
Table 1. Anthropometric and biochemical measurements in nondiabetic and type 2 diabetic subjects before and after 1-year weight reduction (mean ± SD)
Type 2 diabetic subjects
p < 0.01 difference from the value before weight reduction.
p < 0.05 difference from the value before weight reduction.
p < 0.01 difference between genders.
p < 0.05 difference between genders.
Bioelectrical impedance data for one subjects could not be obtained.
The obese diabetic subjects (before n = 8, after n = 6) included in the study did not differ in their anthropometric or biochemical data from other obese subjects (Table 1). However, as expected, the fasting serum glucose concentrations were significantly higher in diabetic subjects both before (p = 0.001) and after (p = 0.026) weight reduction (Table 1), whereas fasting plasma insulin concentration did not differ between nondiabetic and diabetic subjects.
Normal weight subjects (1 man and 5 women; age, 39.5 ± 11.2 years; body weight, 61.6 ± 15.0 kg; BMI, 23.3 ± 3.6 kg/m2; percentage of fat mass, 25 ± 4%; fasting serum free fatty acids, 0.66 ± 0.33 mM; fasting serum glucose, 4.4 ± 0.5 mM; fasting plasma insulin, 40.6 ± 9.4 pM; fasting serum leptin, 11.6 ± 4.4 ng/mL; mean ± SD) had normal values for anthropometric and biochemical measurements.
SREBP-1c mRNA Concentration in Different Adipose Tissue Depots
The SREBP-1c mRNA concentrations in different adipose tissue depots before and after weight reduction are shown in Table 2. There were no gender differences in SREBP-1c mRNA concentrations in any of the depots studied. The concentration in omental adipose tissue was significantly lower than in abdominal or femoral adipose tissue in nondiabetic women (p = 0.001 and p = 0.002, respectively). The trend was similar in nondiabetic men but did not reach statistical significance due to smaller sample size (p = 0.068 for the difference between abdominal subcutaneous and omental depot, and p = 0.109 for the difference between femoral subcutaneous and omental depot). In contrast, SREBP-1c mRNA levels were similar in abdominal and femoral subcutaneous adipose tissue in both genders.
Table 2. SREBP-1c concentration (amol/μg total RNA) in different adipose tissue depots of nondiabetic and type 2 diabetic subjects before and after 1-year weight reduction (mean ± SEM)*
Type 2 diabetic subjects
Results on the entire study population combining nondiabetic and type 2 diabetic subjects by gender are given in the text.
p < 0.01, difference from the OM depot before weight reduction.
p < 0.01, difference from the value before weight reduction.
Omental adipose tissue samples were taken from obese subjects only before weight reduction.
The obese diabetic subjects had similar concentrations of SREBP-1c in all of the depots at baseline to that of the other obese subjects(Table 2). As in nondiabetic obese subjects, SREBP-1c mRNA concentration was lower in omental adipose tissue compared with abdominal (p = 0.050, n = 8) or femoral (p = 0.018, n = 8) subcutaneous adipose tissue.
When the nondiabetic and diabetic subjects were analyzed together, the differences between the depots were significant also in men (omental depot: 1.8 ± 0.4 amol/μg of total RNA vs. abdominal depot: 3.2 ± 0.3, p = 0.050 or vs. femoral depot: 3.9 ± 0.4, p = 0.028). Similarly, these differences were more pronounced in women (omental: 1.8 ± 0.2 amol/μg of total RNA vs. abdominal depot: 3.7 ± 0.5, p = 0.001 or vs. femoral depot: 3.7 ± 0.4, p = 0.001). The depot-related differences detected in SREBP-1c mRNA expression were not observed when measuring hormone-sensitive lipase mRNA as a control by RT-cPCR in the same subjects (omental depot: 120.4 ± 22.4 amol/μg of total RNA vs. abdominal depot: 134.9 ± 26.2, p = 0.135 or vs. femoral depot: 135.6 ± 24.1, p = 0.249, n = 14).
SREBP-1c mRNA concentrations were also determined in abdominal and femoral subcutaneous depots in a group of normal weight subjects. Interestingly, SREBP-1c mRNA levels were markedly higher in abdominal (7.4 ± 2.5 amol/μg of total RNA, p = 0.001) and femoral (6.5 ± 1.5, p = 0.006) subcutaneous depots of the normal weight subjects than in the obese subjects, analyzed either with or without the diabetic subjects.
SREBP-1c mRNA concentration increased along with weight reduction in both abdominal and femoral subcutaneous depots in nondiabetic obese subjects (Table 2). However, the increase in abdominal depot was significant only in women. A clear trend for an increased expression was also seen in femoral adipose tissue in both genders (p = 0.109 for men, p = 0.093 for women).
In the obese diabetic subjects the change in SREBP-1c mRNA level was similar to that in nondiabetic obese subjects in both subcutaneous depots. When the nondiabetic and diabetic subjects were analyzed together, the increase was significant in both depots for men (p = 0.025 in abdominal and p = 0.043 in femoral depot) and in abdominal subcutaneous depot for women (p = 0.008). Figure 1 shows the individual changes in SREBP-1c mRNA concentrations along with weight reduction in both subcutaneous adipose tissue depots.
Correlation between SREBP-1c mRNA and Anthropometric and Biochemical Measures
There were no significant correlations between anthropometric data presented in Table 1 and SREBP-1c mRNA concentration in any depot, before or after weight reduction, in obese subjects. Furthermore, there were no significant correlations between SREBP-1c mRNA concentrations in any of the depots and fasting or 120-minute plasma insulin concentrations before or after weight loss. The changes in plasma insulin values or the area under the curve for insulin values during the OGTT and the changes in SREBP-1c mRNA during the weight reduction did not correlate with each other. Similarly, there was no relationship between SREBP-1c mRNA expression and the data on glucose levels, in the fasting state or during the OGTT.
We found that SREBP-1c mRNA concentrations in subcutaneous adipose tissue in obese subjects were markedly lower than that found in normal weight subjects. Furthermore, a depot-specific difference occurred in obese subjects showing lower SREBP-1c gene expression in omental adipose tissue than in subcutaneous regions. Interestingly, weight loss induced an increase in SREBP-1c mRNA levels in subcutaneous depots. Our study is the first to investigate the expression of SREBP-1c in morbidly obese human subjects both before and after long-term weight reduction in different adipose tissue depots.
SREBP-1c mRNA levels in abdominal and femoral subcutaneous adipose tissue were clearly lower both in massively obese nondiabetic subjects and obese diabetic subjects than in normal weight control subjects. Similarly, in a recent study on different patient material, we observed a decreased expression of SREBP-1c mRNA in abdominal subcutaneous adipose tissue of moderately obese patients compared with lean subjects (19). The results of these two studies indicate that human obesity is characterized by a reduction in SREBP-1c expression in subcutaneous adipose tissue. Similar to our findings, two recent rodent studies have reported that the mRNA expression of SREBP-1c is decreased in obese mice, in addition to the expression of number of genes important for lipogenesis (16) (17). However, in view of the role of SREBP-1c in adipose tissue, these results seem to be quite paradoxical, considering increased fat mass in obese subjects. It might be possible, as also speculated in studies conducted with rodents (16) (17), that the reduction in the expression of SREBP-1c in obesity reflects down-regulation of the pathway leading to lipogenesis to prevent further fat accumulation and weight gain. However, we did not find a significant relationship between SREBP-1c mRNA concentrations and fat mass or the percentage of fat either before or after weight reduction. An alternate possibility is that a reduced SREBP-1c expression is a consequence of an insulin-resistant state commonly found in obese subjects (20). SREBP-1c expression has been shown to be directly regulated by insulin in adipose tissue cell lines (7) (9) and in other cell models (8). We have recently reported that this regulation also occurs in human subcutaneous adipose tissue in vivo (19). Therefore, one can assume that in obese subjects, the effects of insulin are attenuated due to the insulin resistant state. In keeping with this hypothesis in the present study, weight reduction was associated with an increase in SREBP-1c mRNA concentration in subcutaneous adipose tissue along with an improvement of insulin sensitivity, as indicated by a decrease in the fasting concentration of plasma insulin and lower levels of insulinemia at time 120 minutes in the OGTT. It seemed, thus, that the expression of SREBP-1c may be controlled by the ability of insulin to exert its effects on target genes, and a low expression of SREBP-1c could be a marker of insulin resistance. However, we did not find any correlations between SREBP-1c mRNA expression and fasting plasma insulin values or the estimates of insulin sensitivity during the OGTT. The link between insulin sensitivity and SREBP-1c gene expression in adipose tissue seems, therefore, not so evident. In addition, it should be emphasized that the expression level of SREBP-1c in adipose tissue can affect insulin action, as it was nicely demonstrated in mice with over-expression of SREBP-1c that developed lipoatrophy and severe insulin resistance (21). Thus, additional studies are required to define the association between SREBP-1c expression in adipose tissue and insulin sensitivity in humans. To explain the changes in SREBP-1c expression during weight reduction, it should also be noted that we cannot rule out the contribution of the other factors, e.g., changes in nutritional habits and caloric intake of the subjects after gastric banding operation. However, the impact of these factors are quite difficult to study separately in a subject population participating in this kind of clinical study aimed at substantial weight loss, and, thus, they were not investigated in the present study.
Surprisingly, we observed that, although the subjects were still largely obese after 1 year of the weight reduction, SREBP-1c mRNA expression increased in subcutaneous adipose tissue to reach the values close to that determined in normal weight subjects. In a previous study (19), SREBP-1c mRNA levels were significantly lower in moderately obese subjects with stable body weight than in lean subjects. These results may suggest that the observed modification in SREBP-1c mRNA reflects a dynamic change during the process of weight reduction. A longer follow-up with analysis of SREBP-1c expression after several months of stable body weight could be of interest to verify whether the mRNA levels of SREBP-1c remain close to the values of lean subjects or return toward lower values seen in obesity.
The concentration of SREBP-1c mRNA was markedly lower in omental adipose tissue compared with either abdominal or femoral subcutaneous adipose tissue depots in obese subjects of both genders. In agreement with previous studies (22) (23), this depot-related difference was not observed for hormone-sensitive lipase mRNA expression, another important gene of adipocyte metabolism, indicating that the difference observed with SREBP-1c mRNA was not artifactually related to the experimental procedures and, thus, corresponds to a specific regulation of the expression of this gene. Omental adipose tissue has been shown to be metabolically more active than the other fat depots (24), especially when concerning lipolytic activity. It is also known that insulin-mediated effects on adipose tissue metabolism are much weaker in omental adipose tissue than in subcutaneous adipose tissue (25). Because SREBP-1c is suggested to induce the transcription of lipogenic genes and to mediate the effects of insulin (1) (7) (12), the observed low expression of SREBP-1c in lipolytically active adipose tissue region is in agreement with its possible role in mediating insulin action. Alternatively, a low insulin response of this depot may participate in the reduced expression level of the SREBP-1c in omental fat.
In conclusion, the level of SREBP-1c mRNA expression in subcutaneous adipose tissue is markedly lower in obese than in normal weight subjects, which might be due, at least in part, to the insulin-resistant state of the obese subjects. Based on the present results, weight reduction results in a significant increase in SREBP-1c expression in abdominal and femoral adipose tissue along with lowered plasma insulin values. However, changes in calorie intake or other nutritional factors after gastric banding operation might also contribute to the regulation of SREBP-1c expression. These data demonstrate that SREBP-1c expression is controlled in human adipose tissue in a similar way to that found in rodents. In addition, we found that SREBP-1c mRNA levels are significantly lower in omental adipose tissue than in subcutaneous fat depots both from abdominal and femoral regions in obese subjects. Our results strengthen the concept that the regulation of SREBP-1c in adipose tissue is involved in the metabolic adaptations related to weight changes in humans and, thus, may be an interesting target for the development of strategies for the treatment of obesity.
This work was supported by grants from the Academy of Finland, Research Council for Health, Jenny and Antti Wihuri Foundation, Finland, and Saastamoinen Foundation, Kuopio, Finland. We thank Paulette Vallier, Natalie Vega, Erja Kinnunen, Kaija Kettunen, and Irja Kanniainen for skillful technical assistance.