SEARCH

SEARCH BY CITATION

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods and Procedures
  5. Results
  6. Discussion
  7. SUPPLEMENTARY MATERIAL
  8. Acknowledgments
  9. Disclosure
  10. REFERENCES
  11. Supporting Information

Obesity can be considered as a low-grade inflammatory condition, strongly linked to adverse metabolic outcomes. Obesity-associated adipose tissue inflammation is characterized by infiltration of macrophages and increased cytokine and chemokine production. The distribution of adipose tissue impacts the outcomes of obesity, with the accumulation of fat in visceral adipose tissue (VAT) and deep subcutaneous adipose tissue (SAT), but not superficial SAT, being linked to insulin resistance. We hypothesized that the inflammatory gene expression in deep SAT and VAT is higher than in superficial SAT. A total of 17 apparently healthy women (BMI: 29.3±5.5 kg/m2) were included in the study. Body fat (dual-energy X-ray absorptiometry) and distribution (computed tomography) were measured, and insulin sensitivity, blood lipids, and blood pressure were determined. Inflammation-related differences in gene expression (real-time PCR) from VAT, superficial and deep SAT biopsies were analyzed using univariate and multivariate data analyses. Using multivariate discrimination analysis, VAT appeared as a distinct depot in adipose tissue inflammation, while the SAT depots had a similar pattern, with respect to gene expression. A significantly elevated (P < 0.01) expression of the CC chemokine receptor 2 (CCR2) and macrophage migration inhibitory factor (MIF) in VAT contributed strongly to the discrimination. In conclusion, the human adipose tissue depots have unique inflammatory patterns, with CCR2 and MIF distinguishing between VAT and the SAT depots.


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods and Procedures
  5. Results
  6. Discussion
  7. SUPPLEMENTARY MATERIAL
  8. Acknowledgments
  9. Disclosure
  10. REFERENCES
  11. Supporting Information

Adipose tissue is now recognized as an active metabolic–endocrine organ. Proteins secreted from fat have far-reaching effects on body functions including appetite, glucose and lipid metabolism, and immune function (1). Accumulation of body fat in the abdominal region is linked to insulin resistance, with visceral adipose tissue (VAT) enlargement in particular being closely linked to obesity-related complications, including type 2 diabetes and cardiovascular disease (2). Notably, subcutaneous adipose tissue (SAT) can be subdivided into two distinct compartments and deep SAT, but superficial SAT accumulation has not been reported to associate strongly to insulin resistance (3,4).

A putative key link between increasing fat mass and obesity-related complications, including insulin resistance, is a chronic low-grade inflammatory state within adipose tissue, related to infiltration by macrophages (5,6). This includes secretion of specific chemokines thought to attract inflammatory cells to adipose tissue, which are thus key in perpetuating adipose inflammation. Whether this immune activation differs between VAT and the superficial and deep SAT depots has not been studied.

Our hypothesis was that VAT and deep SAT have a higher inflammatory gene expression than superficial SAT. To test this hypothesis, the inflammatory gene expression profile of the different fat depots was measured in apparently healthy women. Because many variables were measured for a series of samples, we used a supervised multivariate classification method, i.e., Partial Least Squares Discriminant Analysis (PLS-DA), combined with crossvalidation to determine which inflammatory markers distinguish the different depots (7,8).

Methods and Procedures

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods and Procedures
  5. Results
  6. Discussion
  7. SUPPLEMENTARY MATERIAL
  8. Acknowledgments
  9. Disclosure
  10. REFERENCES
  11. Supporting Information

Subjects

A total of 17 South African women (6 black and 11 white) undergoing abdominal open surgery for various benign conditions were recruited for the study, as previously described (9). Exclusion criteria included: (i) known human immunodeficiency virus/AIDS; (ii) treatment for any endocrine abnormality (e.g., diabetes, thyroid disease) or intake of other relevant medications; (iii) currently pregnant or breast-feeding; and (iv) known polycystic ovarian syndrome. The study was approved by the Research Ethics Committee of the Faculty of Health Sciences of University of Cape Town, and written informed consent was given by the subjects.

Procedures before surgery

As previously described (9), anthropometric measures, including weight, height, waist (level of umbilicus), and hip (largest gluteal circumference) were taken and body fatness was determined by dual-energy X-ray absorptiometry (Discovery-W; Hologic, Bedford, MA). Computed tomography (X-press Helical Scanner; Toshiba, Tokyo, Japan), at the level of L4/L5, was used to measure regional body fat distribution including quantification of superficial and deep SAT and VAT volumes, as previously described (4). Blood pressure was measured in a seated position after 5–10 min rest using an automated blood pressure monitor (Omron 711; Omron health Care, Hamburg, Germany). The average of three measurements taken at 1-min intervals was used in the analysis. Venous blood samples, for analyses of blood lipids, glucose, and insulin, were collected in the morning after an overnight fast.

Blood analyses

Blood lipids were analyzed using the Roche Modular autoanalyser (Roche, Tarrytown, NY). To analyze total cholesterol, triglycerides, and high-density lipoprotein cholesterol levels, an enzymatic colorimetric assay was used. Low-density lipoprotein (LDL) cholesterol concentrations were determined using the Friedewald's formula (10). Plasma glucose concentrations were measured by the glucose oxidase method (Glucose Analyzer 2; Beckman Instruments, Fullerton, CA). Plasma insulin concentrations were measured by immunochemiluminometric assays using the ADVIA Centaur (Bayer Diagnostics, Penzberg, Germany), described previously (11). Homeostasis model assessment 2 (HOMA2) was calculated as previously described (12).

Procedures during surgery

Adipose tissue biopsies from the VAT, deep and superficial SAT depots in the anterior abdomen were collected during open abdominal surgery. Adipose tissue samples were taken below the umbilicus and VAT was obtained from the inferior portion of the greater omentum. The surgeons discriminated the deep and superficial layers as those lying below and above the fascia superficialis, respectively, as previously described in the plastic surgery literature (13,14,15). The superficial SAT samples were taken just below the skin, whereas the deep SAT samples were taken superficial to the rectus abdominus fascia. Samples were immediately snap frozen in liquid nitrogen and stored at −80 °C until analysis.

RNA extraction, reverse transcription, and real-time PCR

RNA was extracted from whole adipose tissue biopsies using RNeasy Lipid Tissue Midi Kit (Qiagen, Hilden, Germany) according to manufacturer's instructions. The yield and purity of RNA from the three adipose tissue depots in all 17 subjects were determined by spectrophotometer (ND-1000; NanoDrop Technologies, Wilmington, DE) and RNA integrity was evaluated by visual inspection of 28S and 18S RNA on 1% agarose gel electrophoresis in the presence of ethidium bromide. Reverse transcription of 2 µg RNA was performed using High Capacity cDNA Archive Kit (Applied Biosystems, Foster City, CA) with the addition of RNase inhibitor (Applied Biosystems) at a final concentration of 1.0 U/ml. Relative quantification of mRNA expression was analyzed on an ABI Prism 7000 Sequence Detection System (Applied Biosystems) using Universal PCR Master Mix 2X (Roche Molecular Systems, Branchburg, NJ). In total, 12 inflammation-related genes were included in the analysis, including the proinflammatory cytokines tumor necrosis factor-α, interleukin-6, macrophage migration inhibitory factor (MIF), colony-stimulating factor 1; the anti-inflammatory cytokine interleukin-10; the chemokine C–C chemokine ligand 2 and its receptor CCR2, and C-X3-C motif receptor 1 (CX3CR1); the monocyte/macrophage markers (CD68, CD14, CD163, CD206). The following TaqMan gene expression assays (Applied Biosystems) were used: CD14 (Hs 00169122_g1), CD68 (Hs00154355_m1), CD163 (Hs01016657_m1), colony-stimulating factor 1 (Hs00174164_m1), MIF (Hs00036988_g1), C–C chemokine ligand 2 (Hs00234140_m1), CCR2 (Hs00356601_m1), MRC1 (alias: CD206) (Hs00267207_m1), tumor necrosis factor-α (Hs00174128 _m1), interleukin-6 (Hs00985639_m1), interleukin-10 (Hs00174086_m1), CX3CR1 (Hs00365842_m1), PPIA (Hs999999904_m1), LRP10 (Hs00204094_m1), RPLP0 (Hs99999902_m1). Reference genes were evaluated by running PPIA, LRP10, and RPLP0 on a subsample of the study cohort and PPIA was selected due to its low coefficient of variation. Samples were run in triplicate and relative expression levels calculated from a standard curve. Expression levels of target genes were normalized to PPIA.

Multivariate data analysis

To explore the differences between the superficial SAT, deep SAT, and VAT depots based on the gene expression data, the multivariate projection method PLS-DA regression was used (8,16). In PLS-DA, a multidimensional space X is formed by representing each gene expression variable by one dimension and similarly a Y space consisting of one variable for each class (i.e., fat depot). The dimensionality of these spaces are reduced to new low-dimensional spaces by calculating score vectors that summarize the information in the original X variables related to the class assignments of the samples represented by Y.

To evaluate the number of PLS score vectors, crossvalidation is performed. How much of the variance of Y, expressing the class belonging of the samples, that can be predicted by the model is described with Q2, and can be compared with R2 expressing how much of the variance of Y that can be reproduced by the model. A Q2 value >0.1 corresponds to 95% significance of the model and if Q2 is 1, all variance in Y is explained. The similarity/dissimilarity among the fat depots is illustrated by plotting the score vectors summarizing X, against each other in a score plot. The weight plot shows how the genes contribute to the score vectors and the separation between the fat depots. For the multivariate analysis commercial software was used (SIMCA 12.0; Umetrics, Umeå, Sweden). For detailed description, see Supplementary Methods and Procedures online.

Univariate data analysis and statistics

Subject characteristics and gene expression data are expressed as mean ± s.d. The result from the multivariate discrimination analysis was confirmed by conventional univariate analyses. For univariate analyses of intraindividual putative adipose tissue depot differences, the nonparametric Friedman two-way ANOVA was used, followed by Wilcoxon signed-ranks post hoc test for pairwise comparisons (superficial SAT vs. deep SAT; superficial SAT vs. VAT; and, deep SAT vs. VAT). Bonferroni correction was applied for multiple comparisons between adipose tissue depots, so that significance was accepted at P < 0.017. Pearson (r) and, when appropriate, Spearman (rs) correlation coefficients were used to estimate associations between variables. Data were natural log-transformed to achieve normal distribution, if required. In these analyses P < 0.05 was considered significant. All univariate analyses were performed using the SPSS package (version 15.0; SPSS, Chicago, IL).

Results

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods and Procedures
  5. Results
  6. Discussion
  7. SUPPLEMENTARY MATERIAL
  8. Acknowledgments
  9. Disclosure
  10. REFERENCES
  11. Supporting Information

Subject characteristics

Subject characteristics are shown in Table 1. The volumes of the superficial and deep SAT depots were very similar, whereas the volume of the VAT was significantly smaller (P < 0.001 vs. both SAT depots). No differences in anthropometric and metabolic variables, or gene expression data were found between the two ethnic groups (data not shown).

Table 1.  Subject characteristics (N = 17)
inline image

Multivariate analysis of gene expression

Results from the multivariate gene expression analysis of the three abdominal fat depots are shown in Figure 1. The two SAT depots displayed a similar pattern in gene expression, which was confirmed by a separate PLS-DA for these depots where no significant difference was obtained according to crossvalidation. In contrast, the VAT depot differed from the SAT depots and constituted a distinct depot (Figure 1a). The weight plot (Figure 1b) showed how the 12 inflammation-related genes related to the superficial SAT, deep SAT, and VAT depots. The expression profile of MIF and CCR2 contributed strongly to the discrimination of VAT from superficial and deep SAT.

image

Figure 1. Results from multivariate gene expression analyses. (a) Plot of the first two score vectors (t[1]/t[2]) from the Partial Least Squares Discriminant Analysis (PLS-DA) based on 12 gene expression variables in superficial subcutaneous adipose tissue (sSAT; open circles), deep subcutaneous adipose tissue (dSAT; closed circles), and visceral adipose tissue (VAT; open diamonds). The PLS-DA resulted in a significant two component model with R2 = 0.37 and Q2 = 0.25. The rather low R2 and Q2 values were explained by the fact that the superficial SAT and deep SAT did not separate from each other. In contrast, separate PLS-DA models for VAT vs. superficial SAT gave an R2 = 0.60 and Q2 = 0.42 and VAT vs. deep SAT gave R2 = 0.70 and Q2 = 0.47, displaying VAT as a distinct depot. (b) The PLS-DA weight plot shows the contribution of the 12 gene expression variables (w*), denoted with closed triangles, to the scores in a. The weights (c) for the fat depots are denoted with (open squares). VAT separates from the other two depots (superficial and deep SAT), and the gene expression variables that contribute most to this separation are MIF and CCR2.

Download figure to PowerPoint

Univariate analyses of gene expression

To further confirm the results from PLS-DA, we examined the expression of MIF and CCR2, as well as other genes, with univariate analyses (Figure 2, Table 2). CCR2 showed approximately twofold higher expression in VAT compared to superficial and deep SAT (Figure 2a). Similarly, MIF mRNA levels were twofold higher in VAT compared to superficial and deep SAT depots, with higher mRNA levels in VAT being consistent for every subject (Figure 2b). The monocyte/macrophage markers CD14 and CD163 gene expression were significantly higher in VAT vs. superficial SAT (Table 2). Deep SAT had slightly increased expression of CD206 vs. superficial SAT. The expression of the cytokine colony-stimulating factor 1 was slightly higher in deep SAT than superficial SAT (Table 2).

image

Figure 2. Adipose tissue gene expression. (a) Gene expression of CC chemokine receptor 2 (CCR2) in superficial subcutaneous adipose tissue (sSAT), deep subcutaneous adipose tissue (dSAT), and visceral adipose tissue (VAT) in women. (b) Macrophage migration inhibitory factor (MIF) expression in superficial SAT, deep SAT, and VAT. **P < 0.01; ***P < 0.001.

Download figure to PowerPoint

Table 2.  Expression of immune-related genes in superficial subcutaneous adipose tissue (SAT), deep SAT, and visceral adipose tissue (VAT) (N = 17)
inline image

Associations between VAT CCR2 and MIF mRNA levels and anthropometry and metabolic outcomes

Within the VAT depot, CCR2 mRNA correlated with BMI (r = 0.63, P = 0.006), whereas MIF mRNA levels correlated with percentage body fat (r = 0.55, P = 0.02). In contrast, CCR2 and MIF mRNA levels did not correlate with VAT volume, insulin sensitivity, blood lipids, or blood pressure.

Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods and Procedures
  5. Results
  6. Discussion
  7. SUPPLEMENTARY MATERIAL
  8. Acknowledgments
  9. Disclosure
  10. REFERENCES
  11. Supporting Information

This study reveals that the VAT inflammatory gene expression differed from the superficial and deep SAT depots which, in contrast to our original hypothesis, showed an overlapping pattern in gene expression. The multivariate data analysis revealed that the clear separation of VAT from SAT depots was mainly driven by the chemotactic factors CCR2 and MIF. Previous reports comparing inflammation in abdominal adipose tissue depots indicate an elevated inflammation in VAT vs. SAT (17,18,19). However, this is the first study to include both SAT depots in the analysis of inflammatory markers.

CCR2 expression was highest in the VAT depot and contributed strongly to the discrimination of VAT vs. superficial and deep SAT depots. These results are supported by a recent study showing higher CCR2 gene expression in human VAT than SAT with higher levels in severely obese than in lean subjects (20). Similarly, we found that VAT CCR2 expression correlated significantly with BMI. Notably, the chemokine receptor CCR2 seems to be strongly involved in the development of obesity-associated adipose tissue inflammation. Mice deficient in CCR2 fed a high-fat diet had reduced macrophage content and expression of inflammatory markers in adipose tissue compared to wild-type animals matched for adiposity. Furthermore, these mice were more insulin sensitive than adiposity-matched wild-type mice (21). Together, these findings indicate that CCR2 may have a prominent role in human adipose tissue inflammation.

In addition to CCR2, we found significantly higher MIF expression in VAT compared to both superficial and deep SAT. MIF is a cytokine with chemotactic properties that is expressed and released by adipose tissue. In contrast to our findings, Skurk et al. found no difference in the secretion of MIF from isolated human subcutaneous and omental adipocytes (22). This may be due to the fact that cell types other than adipocytes may have contributed to the increased expression of MIF in VAT, as we assessed whole adipose tissue. In addition, Skurk et al. measured protein release from isolated adipocytes in culture (22), where the isolation and culturing procedure of adipocytes may affect their phenotype (23). Interestingly, a recent study investigated the role of MIF in adipose tissue inflammation. LDL-receptor-deficient mice developed insulin resistance and glucose intolerance, whereas double knockout LDL receptor/MIF deficient mice were protected. Furthermore, LDL receptor/MIF deficient animals had reduced chronic inflammation and lower macrophage content and decreased adipose tissue gene expression of several chemokines and cell adhesion factors (24). Notably, both C–C chemokine ligand 2 and CCR2 expression were significantly reduced in adipose tissue of LDL-receptor/MIF deficient mice (24).

The similar outcomes relating to MIF and CCR2 deficiency suggest that elevated levels of MIF and CCR2 in VAT compared to SAT in humans are of major importance for metabolic dysfunction and VAT macrophage infiltration. This is further supported by the elevated expression of the monocyte/macrophage markers CD163 and CD14 in VAT.

In contrast to some previously reported studies of obese subjects (17,25,26), we did not find any depot difference in tumor necrosis factor-α expression. However, in normal-weight subjects, including both men and women, there are no reported depot differences in tumor necrosis factor-α mRNA level (26,27). Because the cohort of this study had a wide BMI range, possible depot differences might be blunted.

Limitations of this study include the relatively low number of subjects. As such, putative effects of ethnicity may need larger cohorts for further evaluation. Differences in fat distribution and metabolic outcomes have thus been reported between black and white women (28,29,30). However, we explored differences in gene expression, with a multivariate statistical approach which is well validated for small cohorts with a relatively large number of variables (31,32). In data sets such as the present one, univariate analysis is not optimal due to the risk of type I and type II errors. This is illustrated by the fact that the univariate analysis for CCR2 and MIF variables shows a larger overlap between VAT and SAT than the PLS-DA analysis (Figures 1a and 2). Thus the univariate analysis results in a type II error in this situation. Multivariate analysis combined with crossvalidation, as in this case PLS, solves these problems. This strategy uses the correlation structure among the variables and calculates one or a few latent variables that are linear combinations of the original variables. These latent variables are thus more stable and less noisy compared to the original variables, in the same way as an average is more stable vs. individual measurements.

We have measured mRNA levels and not protein concentrations in our tissue samples. However, correlations between human adipose tissue mRNA and protein levels from several genes included in this study have been demonstrated by others suggesting that the present data are valid (33,34,35,36). Another consideration is that although VAT seems to be more prone to inflammation than the superficial and deep SAT depots, based on pure mass effects, the impact of the SAT depots on inflammation and metabolic consequences should not be ignored and warrants further investigation. Finally, gene expression data were related to total RNA. It remains to be studied whether tissue weight or cell size may alter our interpretation of the expression of genes of interest.

In summary, we investigated the inflammatory status of three abdominal fat depots in women. Using multivariate data analysis, we found a similar expression pattern of inflammatory genes in superficial and deep SAT, whereas the VAT appeared as a distinct depot. This was mainly due to increased expression of MIF and CCR2 in VAT, suggesting that these proinflammatory factors may play a key role in low-grade inflammation in VAT.

Acknowledgments

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods and Procedures
  5. Results
  6. Discussion
  7. SUPPLEMENTARY MATERIAL
  8. Acknowledgments
  9. Disclosure
  10. REFERENCES
  11. Supporting Information

This study was supported by grants from the Swedish Research Council, the Swedish heart and lung foundation, the Swedish Diabetes Foundation, the NovoNordisk Foundation, the Swedish Society for Medical Research, the Petrus and Augusta Hedlunds Foundation, the Thurings Foundation, the Faculty of Medicine of Umeå University, the Bergwalls Foundation, the Lars Hiertas Minne as well as the Medical Research Council of South Africa, the International Atomic Energy Agency, the National Research Foundation of South Africa, and the University of Cape Town.

REFERENCES

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods and Procedures
  5. Results
  6. Discussion
  7. SUPPLEMENTARY MATERIAL
  8. Acknowledgments
  9. Disclosure
  10. REFERENCES
  11. Supporting Information
  • 1
    Greenberg AS, Obin MS. Obesity and the role of adipose tissue in inflammation and metabolism. Am J Clin Nutr 2006; 83: 461S465S.
  • 2
    Wajchenberg BL. Subcutaneous and visceral adipose tissue: their relation to the metabolic syndrome. Endocr Rev 2000; 21: 697738.
  • 3
    Kelley DE, Thaete FL, Troost F, Huwe T, Goodpaster BH. Subdivisions of subcutaneous abdominal adipose tissue and insulin resistance. Am J Physiol Endocrinol Metab 2000; 278: E941E948.
  • 4
    Smith SR, Lovejoy JC, Greenway F et al. Contributions of total body fat, abdominal subcutaneous adipose tissue compartments, and visceral adipose tissue to the metabolic complications of obesity. Metab Clin Exp 2001; 50: 425435.
  • 5
    Cancello R, Clément K. Is obesity an inflammatory illness? Role of low-grade inflammation and macrophage infiltration in human white adipose tissue. BJOG 2006; 113: 11411147.
  • 6
    Kershaw EE, Flier JS. Adipose tissue as an endocrine organ. J Clin Endocrinol Metab 2004; 89: 25482556.
  • 7
    Sjöström M, Wold S, Söderström B. PLS discriminant plots. In: Geselma ES, Kanal LN (eds). Pattern Recognition in Practice II. Elsevier: Amsterdam, 1986, pp 461470.
  • 8
    Wold S, Eriksson L, Sjöström M. Partial least squares to latent structures (PLS) in chemistry. In: Schleyer PvR, Allinger NL, Clark T et al. (eds). Encyclopedia of Computational Chemistry. Wiley & Sons: Chicester, 1999, pp 20062022.
  • 9
    Goedecke JH, Wake DJ, Levitt NS et al. Glucocorticoid metabolism within superficial subcutaneous rather than visceral adipose tissue is associated with features of the metabolic syndrome in South African women. Clin Endocrinol (Oxf) 2006; 65: 8187.
  • 10
    Friedewald WT, Levy RI, Fredrickson DS. Estimation of the concentration of low-density lipoprotein cholesterol in plasma, without use of the preparative ultracentrifuge. Clin Chem 1972; 18: 499502.
  • 11
    El Kenz H, Bergmann P. Evaluation of immunochemiluminometric assays for the measurement of insulin and C-peptide using the ADVIA Centaur. Clin Lab 2004; 50: 171174.
  • 12
    Wallace TM, Levy JC, Matthews DR. Use and abuse of HOMA modeling. Diabetes Care 2004; 27: 14871495.
  • 13
    Gasperoni C, Salgarello M. Rationale of subdermal superficial liposuction related to the anatomy of subcutaneous fat and the superficial fascial system. Aesthetic Plast Surg 1995; 19: 1320.
  • 14
    Johnson D, Dixon AK, Abrahams PH. The abdominal subcutaneous tissue: computed tomographic, magnetic resonance, and anatomical observations. Clin Anat 1996; 9: 1924.
  • 15
    Markman B, Barton FE Jr. Anatomy of the subcutaneous tissue of the trunk and lower extremity. Plast Reconstr Surg 1987; 80: 248254.
  • 16
    Wold S, Sjöström M, Eriksson L. PLS-Regression: a basic tool of chemometrics. Chemometr Intell Lab Syst 2001; 58: 109130.
  • 17
    Cao YL, Hu CZ, Meng X, Wang DF, Zhang J. Expression of TNF-α protein in omental and subcutaneous adipose tissue in obesity. Diabetes Res Clin Pract 2008; 79: 214219.
  • 18
    Poulain-Godefroy O, Lecoeur C, Pattou F, Frühbeck G, Froguel P. Inflammation is associated with a decrease of lipogenic factors in omental fat in women. Am J Physiol Regul Integr Comp Physiol 2008; 295: R1R7.
  • 19
    Fain JN, Madan AK, Hiler ML, Cheema P, Bahouth SW. Comparison of the release of adipokines by adipose tissue, adipose tissue matrix, and adipocytes from visceral and subcutaneous abdominal adipose tissues of obese humans. Endocrinology 2004; 145: 22732282.
  • 20
    Huber J, Kiefer FW, Zeyda M et al. CC chemokine and CC chemokine receptor profiles in visceral and subcutaneous adipose tissue are altered in human obesity. J Clin Endocrinol Metab 2008; 93: 32153221.
  • 21
    Weisberg SP, Hunter D, Huber R et al. CCR2 modulates inflammatory and metabolic effects of high-fat feeding. J Clin Invest 2006; 116: 115124.
  • 22
    Skurk T, Herder C, Kräft I et al. Production and release of macrophage migration inhibitory factor from human adipocytes. Endocrinology 2005; 146: 10061011.
  • 23
    Ruan H, Zarnowski MJ, Cushman SW, Lodish HF. Standard isolation of primary adipose cells from mouse epididymal fat pads induces inflammatory mediators and down-regulates adipocyte genes. J Biol Chem 2003; 278: 4758547593.
  • 24
    Verschuren L, Kooistra T, Bernhagen J et al. MIF deficiency reduces chronic inflammation in white adipose tissue and impairs the development of insulin resistance, glucose intolerance, and associated atherosclerotic disease. Circ Res 2009; 105: 99107.
  • 25
    Cao YL, Wang YX, Wang DF, Meng X, Zhang J. Correlation between omental TNF-α protein and plasma PAI-1 in obesity subjects. Int J Cardiol 2008; 128: 399405.
  • 26
    Zha JM, Di WJ, Zhu T et al. Comparison of gene transcription between subcutaneous and visceral adipose tissue in Chinese adults. Endocr J 2009; 56: 935944.
  • 27
    Walker GE, Verti B, Marzullo P et al. Deep subcutaneous adipose tissue: a distinct abdominal adipose depot. Obesity (Silver Spring) 2007; 15: 19331943.
  • 28
    Lovejoy JC, de la Bretonne JA, Klemperer M, Tulley R. Abdominal fat distribution and metabolic risk factors: effects of race. Metab Clin Exp 1996; 45: 11191124.
  • 29
    van der Merwe MT, Crowther NJ, Schlaphoff GP et al. Evidence for insulin resistance in black women from South Africa. Int J Obes Relat Metab Disord 2000; 24: 13401346.
  • 30
    Goedecke JH, Levitt NS, Lambert EV et al. Differential effects of abdominal adipose tissue distribution on insulin sensitivity in black and white South African women. Obesity (Silver Spring) 2009; 17: 15061512.
  • 31
    Pérez-Enciso M, Tenenhaus M. Prediction of clinical outcome with microarray data: a partial least squares discriminant analysis (PLS-DA) approach. Hum Genet 2003; 112: 581592.
  • 32
    Nguyen DV, Rocke DM. Tumor classification by partial least squares using microarray gene expression data. Bioinformatics 2002; 18: 3950.
  • 33
    Harman-Boehm I, Blüher M, Redel H et al. Macrophage infiltration into omental versus subcutaneous fat across different populations: effect of regional adiposity and the comorbidities of obesity. J Clin Endocrinol Metab 2007; 92: 22402247.
  • 34
    Weisberg SP, McCann D, Desai M et al. Obesity is associated with macrophage accumulation in adipose tissue. J Clin Invest 2003; 112: 17961808.
  • 35
    Westerbacka J, Cornér A, Kolak M et al. Insulin regulation of MCP-1 in human adipose tissue of obese and lean women. Am J Physiol Endocrinol Metab 2008; 294: E841E845.
  • 36
    Skopková M, Penesová A, Sell H et al. Protein array reveals differentially expressed proteins in subcutaneous adipose tissue in obesity. Obesity (Silver Spring) 2007; 15: 23962406.

Supporting Information

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods and Procedures
  5. Results
  6. Discussion
  7. SUPPLEMENTARY MATERIAL
  8. Acknowledgments
  9. Disclosure
  10. REFERENCES
  11. Supporting Information

supporting Information

FilenameFormatSizeDescription
oby_1713_sm_1.doc41KSupporting info item

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.