| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Endocrinology and Diabetes Unit (V.C.-M., M.M., M.R.C., A.M., C.G., J.V.), Research Department, University Hospital of Tarragona Joan XXIII, "Pere Virgili" Institute, 43007 Tarragona, Spain; Endocrinology and Diabetes Unit (N.V., J.M.G.), University Hospital of Bellvitge 08907, Barcelona, Spain; Diabetes, Endocrinology, and Territorial Nutrition Unit of Girona (J.M.F.-R.), University Hospital "Dr. Josep Trueta", and CIBER Fisiopathology of Obesity (CB06/03/010), Health Institute Carlos III, 17007 Girona, Spain; Surgery Service (E.C.), Hospital de Sta Tecla, 43003 Tarragona, Spain; and Metabolic Research Laboratory (G.F.), Department of Endocrinology, University Clinic of Navarra, University of Navarra, 31080 Pamplona, Spain
Address all correspondence and requests for reprints to: Joan Vendrell, Secció d’Endocrinologia. Hospital Universitari Joan XXIII de Tarragona, C/ Dr. Mallafré Guasch, 4, 43007 Tarragona, Spain. E-mail: jvo{at}comt.es.
 |
Abstract |
|---|
 Top Abstract IntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Objective: Our objective was to test the hypothesis that AQP7 gene expression levels may be affected by presence of obesity and type 2 diabetes in humans.
Design: The obesity study cohort consisted of 12 lean, 22 nonseverely obese, and 13 severely obese subjects. The type 2 diabetes study cohort consisted of 17 lean and 39 obese type 2 diabetic patients. Circulating levels of plasma soluble proteins monocyte chemoattractant protein-1, TNF receptors 1 and 2, and IL-6 and glycerol were measured. The sc adipose tissue gene expression of AQP7, MCP-1, IL-6, TNF
, PPAR
, and SREBP1c genes was measured by real-time PCR. AQP7 gene mutation analysis was performed.
Results: Severely obese women showed lower AQP7 expression levels compared with lean and nonseverely obese (P < 0.001). Moreover, circulating glycerol concentration was lower in severely obese subjects, but no correlation with AQP7 adipose tissue expression was observed. AQP7 expression was negatively related with proinflammatory genes (for monocyte chemoattractant protein-1, r = –0.203 and P = 0.044; for TNF, r = –0.209 and P = 0.036). Concerning adipogenic factors, AQP7 expression levels were found to be positively determined by PPAR mRNA expression levels (r = 0.265; P = 0.012). AQP7 expression did not show differences regarding the presence of type 2 diabetes.
Conclusion: Expression of AQP7 is down-regulated in women with severe obesity. The expression of this glycerol channel is not affected by type 2 diabetes.
|
Introduction |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
The molecular basis for the secretion and uptake of glycerol has been recently characterized with the discovery of the aquaporin (AQP) family of proteins. The AQP7 gene is a member of this family of membrane channels implicated in controlling fat and in turn glucose metabolism. The AQP7 gene was cloned from human adipose tissue and is responsible for glycerol permeability and other small solutes (5). The release of glycerol in mouse 3T3-L1 cells increases during differentiation to adipocytes in parallel with increasing AQP7 mRNA levels (6). The AQP7 gene contains a putative peroxisome proliferator response element and a putative insulin response element in the promoter region (5). Adipose tissue AQP7 expression is regulated by fasting-refeeding, insulin, dexamethasone, TNF, isoproterenol, and peroxisome proliferator-activated receptor (PPAR)- and - (6, 7, 8, 9).
AQP7-knockout mice models developed obesity and insulin resistance, had impaired plasma glycerol permeability (altered uptake or secretion rates) compared with wild-type mice, and exhibited accumulated excess glycerol and triglyceride concentrations in adipocytes (10, 11, 12).
In humans, AQP7 has been shown to be equally expressed in omental and sc adipose tissue, with higher expression in women than in men (13). Studies on AQP7 expression in human adipose tissue are scarce but suggest that the AQP7 gene is less expressed in sc adipose tissue of obese compared with lean male subjects (14). Three missense mutations (R12C, V59L, and G264V) have been identified in human subjects in a Japanese cohort. Nevertheless, the frequency of these mutations was not associated with obesity or type 2 diabetes. Functional analysis showed that the permeability of water and glycerol is disturbed in Xenopus oocytes expressing G264V mutant protein (5).
Taking into consideration the above-mentioned observations, we suggest that AQP7 gene expression could be altered by the presence of obesity and type 2 diabetes in humans, both conditions associated with insulin resistance, differential adipogenic transcription, and higher local and systemic inflammatory environment. Therefore, the aim of the present study was to analyze AQP7 gene expression in human sc adipose tissue biopsies in obesity and type 2 diabetes and to relate it to inflammatory and adipogenic markers. Likewise, a preliminary study of described gene mutation frequencies was also performed in our Caucasian cohort to analyze its frequency in obese and type 2 diabetic patients.
|
Subjects and Methods |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
A group of 103 subjects was recruited at the Hospital Universitari Joan XXIII (Tarragona, Spain), Hospital Sant Pau i Santa Tecla (Tarragona, Spain), and Hospital de Bellvitge (Barcelona, Spain). All subjects were of Caucasian origin and reported that their body weight had been stable for at least 3 months before the study. They had no systemic disease other than obesity or type 2 diabetes, and all were free of any infections in the previous month before the study. Liver and renal diseases were specifically excluded by biochemical work-up.
Obesity study cohort.
Overweight and nonsevere obesity were defined as a body mass index (BMI) between 25.0 and 39.9 kg/m2 and severe obesity being considered when the BMI was higher than 40 kg/m2 (15). Normal weight was considered in individuals with a BMI less than 25.0 kg/m2. According to these criteria, three groups were studied: 12 lean subjects, 22 nonseverely obese subjects, and 13 severely obese subjects. All subjects were nondiabetic (Table 1
).
|
).
|
Height was measured to the nearest 0.5 cm and body weight to the nearest 0.1 kg. BMI was calculated as weight (kilograms) divided by height (meters) squared. Waist circumference was measured midway between the lowest rib margin and the iliac crest. Hip circumference was determined as the widest circumference measured over the greater throcanter. Waist-to-hip ratio (WHR) was accordingly calculated.
Analytical methods
Plasma and serum samples were stored at –80 C until analytical measurements were performed, except for glucose, which was immediately determined after blood was drawn.
Serum glucose was measured using the glucose analyzer YSI 2300 STAT Plus (YSI Inc., Yellow Springs, OH). Total serum cholesterol was measured through the reaction of cholesterol esterase/cholesterol oxidase/peroxidase. High-density lipoprotein (HDL) cholesterol was quantified after precipitation with polyethylene glycol at room temperature. Total serum triglycerides were measured through the reaction of glycerol-phosphate-oxidase and peroxidase.
Plasma soluble monocyte chemoattractant protein-1 (sMCP-1) levels were measured by (h)MCP-1 Biotrak ELISA kit (Amersham Biosciences, Buckinghamshire, UK); the sensitivity of the assay was less than 10 pg/ml, and the coefficients of variation (CV) were less than 10%. Soluble TNF receptor 1 (sTNFR1) and 2 (sTNFR2) were determined by solid-phase enzyme immunoassay with amplified reactivity (BioSource Europe, Nivelles, Belgium). The limit of detection was 50 pg/ml for sTNFR1 and 0.1 ng/ml for sTNFR2, and the intra- and interassay CV were less than 7% and less than 9%, respectively. Circulating sIL-6 was determined by an ultrasensitive solid-phase enzyme immunoassay (BioSource). The mean of the minimum detectable concentration was 0.039 pg/ml. Intra- and interassay CV were less than 9.8% and less than 11.2%, respectively. Plasma high-sensitivity C-reactive protein (hsCRP) was determined by a highly sensitive immunonephelometry kit (Dade Behring, Marburg, Germany).
Plasma glycerol levels were analyzed by using a free glycerol determination kit, a quantitative enzymatic determination assay (Sigma-Aldrich Corp., St. Louis, MO). Intra- and interassay CV were less than 6% and less than 9.1%, respectively.
Serum insulin concentrations were measured by a monoclonal immunoradiometric assay (Coat-A-Count insulin; Diagnostic Products Corp., Los Angeles, CA). Intra- and inter-assay CV were of 6.6% and 7.1%.
Adipose tissue samples
All adipose tissue samples were obtained from sc abdominal depots during abdominal elective surgical procedures (gastric bypass operation, cholecystectomy, and surgery of abdominal hernia).
Preoperative anthropometric measurements were made and blood samples were collected before the surgical procedure. All patients had fasted overnight, and at the beginning of surgery, 2–4 g sc fat tissue was removed by scalpel from each proband and immediately introduced in RNALater (Sigma-Aldrich) and stored at –80 C until RNA extraction.
Informed written consent was obtained, and the purpose, nature, and potential risks of the study were explained to the subjects. The experimental protocol was approved by the ethics committee of the hospital.
Total RNA isolation and RT
Total RNA was extracted from 400–500 mg frozen sc adipose tissue by using RNeasy Lipid Tissue Midi Kit (QIAGEN Science, Hilden, Germany) following the manufacturer’s instructions. The RNA was treated with 55 U RNase-free DNase (QIAGEN) before column elution to avoid contamination with genomic DNA. The RNA integrity was electrophoretically verified by ethidium bromide staining and its purity by the OD260/OD280 absorption ratio.
One microgram of RNA was reverse transcribed to cDNA by using Promega RT system (Promega Corp., Madison, WI); 20 µl RT mixture contained 1X RT buffer [10 mM Tris-HCl (pH 9.0), 50 mM KCl, 0.1% Triton X-100], 1 mM of each dNTP, 1 U/µl recombinant RNAsin ribonuclease inhibitor, 15 U AMV reverse transcriptase, and 0.5 µg random primers. The cDNA reaction was incubated for 10 min at 25 C followed by 60 min at 42 C and then heated 5 min at 95 C.
Real-time PCR
Real-time quantitative PCR analyses were performed with 2 µl cDNA on LightCycler Instrument (Roche Diagnostics, Basel, Switzerland), using the SYBR green fluorescence method and in a final volume of 20 µl. The following primers were used: for specific AQP7, 5'-caaaatggtctcctggtccg-3' and 5'-gactccgaagccaaaaccc-3'; for MCP-1, 5'-tctgtgcctgctgctcatag-3' and 5'-cagatctccttggccacaat-3'; for PPAR, 5'-ctatggagttcatgcttgtg-3' and 5'-gtactgacatttattt-3'; for sterol-responsive element binding protein 1c (SREBP1c), 5'-aaggtgaagtcggcgcgg-3' and 5'-atcggggctggcaggg-3'; for TNF, 5'-gagcactgaaagcatgatcc-3' and 5'-gctggttatctctcagctcca-3'; and for IL-6, 5'-cggtacatcctcgacgg-3' and 5'-tgatgattttcaccaggc-3'. The housekeeping genes used to normalize gene expression were as follows: ß-actin, 5'-ggacttcgagcaagagatgg-3' and 5'-agcactgtgttggcgtacag-3', and cyclophilin A (CYPA), 5'-caaatgctggacccaacac-3' and 5'-gcctccacaatattcatgccttctt-3'.
The purity of each amplified product was confirmed by melting curve analysis and by adjusting the detection of the fluorescent signal to avoid primer-dimmer detection.
Expression data were calculated with external standard curve, created with serial dilutions of a cloned PCR fragment from the respective gene, using LightCycler Software version 3.5 (Roche Diagnostics). Values are expressed as a ratio to the ß-actin or CYPA expression.
Analysis of missense mutations
G to T substitution at exon 8 of the AQP7 gene led to the amino acid substitution from glycine to valine at position 264 (G264V) (5). We searched for this mutation in 178 Caucasian subjects. These included 127 nondiabetic subjects (37 lean and 90 obese) and 51 type 2 diabetic subjects (14 lean and 37 obese). The following primers were used: 5'-TGAACGCAGCTGTGACCTTTG-3' and 5'-TGGTCTTCATACGCCACAGA-3', and for sequencing, 5'-GTAGCCTGGGGATGACTCCT-3'.
The PCR products were directly sequenced on an ABI PRISM 310 automatic sequencer.
Statistical analysis
Statistical analysis was performed by using the SPSS/PC+ statistical package (version 13 for Windows; SPSS, Chicago, IL). For clinical and anthropometric variables, normal distributed data are expressed as mean value ± SD, and for variables with no Gaussian distribution, values are expressed as median (25–75th percentile). For statistical analysis of expression variables that did not have a Gaussian distribution, values were logarithmically transformed.
Differences in clinical or laboratory parameters between groups were compared by using ANOVA with a post hoc Scheffé correction or a Student’s t test. A univariate general lineal model was used to analyze differences between groups correcting for confusing variables. Differences in sex between studied groups were analyzed by Pearson’s 
2 test. Associations between quantitative variables were evaluated by Pearson/Spearman’s correlation analysis, and correction for confounding and interacting variables was performed using a stepwise multiple linear regression analysis. Results are expressed as unstandardized coefficient (B) and 95% confidence interval for B [95% CI(B)]. Statistical significance occurred if a computed two-tailed probability value was < 0.05.
|
Results |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
The main anthropometric and analytical characteristics of the studied population are shown in Table 1
. Our data showed that severely obese patients had significantly lower glycerol concentrations than nonseverely obese.
Relative mRNA expression in sc adipose tissue
Table 1 shows the relative expression levels of the different studied genes. Regarding AQP7, severely obese subjects showed significantly decreased expression levels when compared with lean and nonseverely obese. Given that the groups were mismatched for gender and taking into account that our group of severely obese subjects was composed mainly of women, an analysis was performed to verify whether the observed differences were due to BMI. A univariate general lineal model was constructed considering only women and the results showed that AQP7 expression had significant lower levels with increasing BMI (P = 0.001). Therefore, we can conclude that AQP7 gene expression is reduced in severely obese women. Reduction in AQP7 expression in the adipose tissue in severely obese men was not confirmed because of the absence of an adequate number of subjects in this group.
Additionally, severely obese patients had significantly higher mRNA levels of MCP-1, IL-6, and TNF and lower mRNA levels for SREBP1c. No significant differences were found in PPAR expression among the studied groups.
Type 2 diabetes study
The main anthropometric and analytical characteristics of the studied population are shown in Table 2. No significant differences were found in glycerol concentrations between the nondiabetic obese cohort and the obese type 2 diabetic subjects.
Relative mRNA expression in sc adipose tissue
AQP7 expression in adipose tissue was similar in nondiabetes and type 2 diabetes (Table 2). Obese patients with type 2 diabetes, showed significantly increased PPAR mRNA expression levels and significantly reduced SREBP1c mRNA levels when compared with nondiabetic obese subjects.
Correlation and regression analysis results
For bivariate correlation analysis, all of the studied population was included (n = 103). We found that AQP7 mRNA expression levels were positively associated with age (r = 0.197; P = 0.046) and PPAR (r = 0.265; P = 0.012) and negatively associated with BMI (r = –0.473; P < 0.001), hsCRP (r = –0.264; P = 0.019) and MCP-1 (r = –0.203; P = 0.044) and TNF (r = –0.209; P = 0.036) mRNA expression. The independence of the associations was evaluated by linear regression analysis, including sex and presence of type 2 diabetes as confounding variables. We found that AQP7 mRNA expression was positively associated with PPAR [B = 0.248, P = 0.002, and 95% CI(B) = 0.102/0.394 in women, and B = 0.321, P = 0.004, and 95% CI(B) = 0.114/0.529 in men] and negatively associated with TNF mRNA in men [B = –0.329, P = 0.005, and 95% CI(B) = –0.549/–0.109].
Determination of G264V mutation of the AQP7 gene in human Caucasian subjects
From the 178 subjects analyzed for G264V mutation, 14 presented the studied mutation (8%). The distribution of the mutation in the obese and diabetic population did not show differences (2 = 0.15 and P = 0.69 between lean and obese subjects, and 2 = 2.34 and P = 0.12 between nondiabetic and type 2 diabetic patients). Frequency distribution is shown in Table 3. Only one homozygous subject who had type 2 diabetes and was overweight (BMI = 28) was identified and had glycerol levels below the 10th percentile.
|
|
Discussion |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
The metabolic outcome of adipose tissue may be different according to the anatomical origin; however, no differences between omental and sc fat appear to exist regarding AQP7 expression (13). In the same study, sexual dimorphism with higher levels in women has been described, although only 10 men and four women were analyzed (13). We have not found gender differences in AQP7 expression in subjects without severe obesity (data not shown). However, we cannot assume this observation for severely obese subjects, because in our case, this group was mainly constituted by women.
The biological function of this gene as a glycerol channel has been studied in animal models deficient for AQP7 (10, 11). Only in one model were lower glycerol plasma circulating levels detected (12), whereas others (11) did not find serum glycerol parameters altered; however, they found a reduction in glycerol permeability in 3T3-L1 knockdown adipocytes (10). In our study in humans, plasma glycerol levels were found to be significantly lower in severely obese women, and despite lower AQP7 mRNA expression in these subjects, no correlation was detected between gene expression and glycerol circulating levels. A possible explanation for this lack of correlation could be due to the existence of an alternative glycerol secretion in human adipocytes, as has been suggested by other authors (3, 5, 10).
Promoter studies on the AQP7 human gene have identified a putative insulin response motif that mediates insulin suppression of the human AQP7 gene as well as one putative regulatory element for PPAR (5), similar to what happened in the genomic structure of the mouse gene (18). In animal studies, it is well documented that AQP7-deficient mice or 3T3-L1 knockdown adipocytes exhibit an altered insulin state (11). In humans, we have shown that mRNA AQP7 expression in sc adipose tissue is not affected in obese type 2 diabetic patients when compared with their nondiabetic counterparts. Likewise, glycerol circulating levels were similar in both cohorts with independence of AQP7 expression, albeit it is known that plasma glycerol levels depend on several factors within the adipocytes, including the rate of lipolysis, cytoplasmic glycerol concentration, and plasma-membrane glycerol permeability. Although these results may argue against a direct implication of AQP7 on the insulin resistance state, additional experiments need to be done to determine whether there are any alterations at the posttranscriptional level altering glycerol permeability. In fact, AQP7 sc expression was negatively related to local proinflammatory expression genes in adipose tissue, such as MCP-1 and TNF, which are associated with poor environmental insulin sensitivity. This observation is in agreement with in vitro studies with differentiated 3T3-L1 adipocytes where suppression of AQP7 expression was mediated by TNF stimulation (8), a cytokine involved in the pathogenesis of insulin resistance. Concerning adipogenic factors, AQP7 expression levels were found to be positively determined by PPAR mRNA expression, reinforcing the implication of the peroxisome proliferator response element site as a transcriptional regulatory element.
Genotypic characterization of our Caucasian population on the AQP7 gene showed up to 8% of the subjects with the G264V mutation. Despite that this frequency was higher than the one reported by Kondo et al. (5) in the Japanese population, we failed to find an association with obesity or type 2 diabetes. However, a more adequately powered statistical association analysis in Caucasian subjects should be performed before a definite conclusion in the genetic susceptibility to both comorbidities can be stated. We detected absence of R12C and V59L gene variants (data not shown). It is worth mentioning that 43% of the subjects that carried the G264V mutation were both obese and type 2 diabetic, in contrast to 0% in the Japanese population (5).
Additionally, we detected a homozygous subject for the G264V mutation, who had plasma glycerol levels within the normal values, in agreement with the one described by Kondo et al. (5), suggesting once more the hypothesis of an alternative glycerol channel in adipocytes.
Although our studied sample was insufficient for determining genetic associations, the observed frequencies are clearly different from those reported in the Japanese population. Other association studies have been published with a correct sample size, which report that an A–953G variant, which negatively modulates AQP7 expression, in the AQP7 promoter, is associated with higher risk of being obese or having type 2 diabetes in women (17). However, although their results are statistically significant, there are odds ratio values close to the null hypothesis when analyzing the confidence interval, which means that this will be not accomplished in all –953G subjects. However, we have not detected down-regulation of AQP7 in type 2 diabetic patients.
In conclusion, the present study shows that sc adipose tissue AQP7 expression in severely obese women is down-regulated and has lower plasma glycerol levels. Additional studies need to be done to confirm whether this statement is accomplished also in severely obese men. Additionally, we report that type 2 diabetes does not affect AQP7 expression. A clear relation between AQP7 and PPAR expression has been observed, and local proinflammatory gene expression shows a negative relationship with AQP7 expression in adipose tissue.
Further studies are needed on human AQP7 adipocyte physiology (glycerol secretion and its regulation) to clarify its role in glucose homeostasis, insulin resistance, and adipocyte biology.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
Disclosure Information: All authors have nothing to declare.
First Published Online June 12, 2007
Abbreviations: AQP, Aquaporin; B, unstandardized coefficient; BMI, body mass index; CI(B), confidence interval for B; CV, coefficient of variation; HDL, high-density lipoprotein; hsCRP, high-sensitivity C-reactive protein; PPAR, peroxisome proliferator-activated receptor; sMCP-1, soluble monocyte chemoattractant protein-1; TNFR, TNF receptor; WHR, waist-to-hip ratio.
Received March 8, 2007.
Accepted June 6, 2007.
|
References |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
. J Biol Chem 276:48572–48579, and dexamethasone. Horm Metab Res 35:222–227[CrossRef][Medline] deficiency. Int J Obes 10.1038/sj.ijo. 0803555This article has been cited by other articles:
 |
|
 |
|
  P. Keller, J. T. Petrie, P. De Rose, I. Gerin, W. S. Wright, S.-H. Chiang, A. R. Nielsen, C. P. Fischer, B. K. Pedersen, and O. A. MacDougald Fat-specific Protein 27 Regulates Storage of Triacylglycerol J. Biol. Chem., May 23, 2008; 283(21): 14355 - 14365. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Endocrinology | Endocrine Reviews | J. Clin. End. & Metab. |
| Molecular Endocrinology | Recent Prog. Horm. Res. | All Endocrine Journals |
