Adipokine Concentration in Adipose Tissue of Obese Mice: Location Dependency

Hypertrophy and hyperplasia of adipocytes leading to enlargement of adipose tissue are the characteristics of obesity. Adipocytes secrete adipokines, resulting in elevated levels of adipokines in the circulation. However, whether differential secretion of adipokines in obesity is a result of secretion from enlarged adipocytes or due to location-specific production in adipose tissue is not known. Our prior study showed that Perivascular Adipose Tissue (PVAT) plays a crucial role in increased intima-media thickness and vascular smooth muscle cell proliferation. Whether PVAT adipokines play a crucial role in vascular cells changes is unclear. Moreover, PVAT adipokine production compared to other adipose tissue is not known. In the present study, we examined differential adipokine concentrations in PVAT compared to subcutaneous and gonadal adipose tissue and investigated the correlation between adipocyte cell surface area and adipokine concentrations among these tissues. We measured the concentrations of leptin, macrophage chemoattractant protein-1, IL-6, plasminogen activator inhibitor-1, and resistin in plasma and adipose tissues in three locations (subcutaneous, gonadal, perivascular) as well as plasma insulin concentrations in diet-induced obese mice using Millplex Immunoassays. We also examined the surface area of adipocyte cells in these adipose tissues in obese and lean mice. Finally, we investigated the correlation of the surface area of adipocyte cells in these adipose tissues with adipokine concentrations. Our study revealed that cell surface area of PVAT adipocytes is the smallest among three adipocytes and elevated level of IL-6, but not other four adipokines was observed in PVAT of obese mice. More importantly, differential adipokine concentrations in different adipose tissues were dependent primarily on the location of adipose tissue. Our prior and current studies suggest that increased IL-6 level in PVAT contributes to vascular remodeling.


Introduction
Obesity is a risk factor for cardiovascular, cerebrovascular and metabolic disorders. Obesity occurs as a result of the enlargement of adipose tissue, due to hypertrophy and hyperplasia of adipocytes. Adipocytes are critical repositories of free fatty acids and also act as an endocrine organ. Adipose tissue secretes several pro-inflammatory cytokines and chemokines (collectively known as "adipokines"). Circulatory levels of inflammatory adipokines are elevated in obese subjects [1].
The notion that elevatedlevels of circulatory adipokines are a result of secretion of adipokines from hypertrophic adipocytes, or due to different stimuli at a specific location of adipose tissue, is controversial. Some studies have suggested that elevated levels of adipokines are dependent upon the size of adipocytes, whereas other studies have shown that they are dependent upon specific location of adipose tissue, such as Subcutaneous Adipose Tissue (SAT) and Visceral Adipose Tissue (VAT) [2,3]. Levels of leptin, IL-6, IL-8,Tumor Necrosis Factor (TNF)-α, Macrophage Chemoattractant Protein (MCP)-1, Interferon Gamma-Induced Protein (IP)-10, Macrophage Inflammatory Protein (MIP)-1β, and Granulocyte-Colony Stimulating Factor (G-CSF) from freshly isolated adipocytes were positively associated with adipocyte volume and the greatest secretion was observed in the largest adipocyte fraction obtained from the SAT [2]. Other reports have noted location-specific secretion of adipokines from SAT and VAT. Linder et al., have shown that calcyclin, adipsin and PAC clone 12p13.3 are expressed only in SAT, and that Ras homolog gene family member G and phospholipid transfer protein are expressed in VAT [4]. Other reports have shown upregulation of expression of MIP-1 in SAT and IL-6 in VAT [3]. Differential secretion between SAT and VAT, of leptin, TNF-α, angiotensin, and Plasminogen Activator Inhibitor (PAI)-1 has been reported [5]. Leptin mRNA secretion is higher in SAT than VAT in humans, in contrast to rodents, where leptin mRNA levels in Gonadal Adipose Tissue (GAT) and retroperitoneal adipose tissue are higher than those in SAT [6,7]. Another adipose tissue-specific protein, resistin, is secreted abundantly in GAT. Resistin secretion is reduced as rodents become obese, but it is noteworthy that resistin levels in the circulation may remain elevated [8].
PVAT exerts an anti-contractile effect on vessels through both endothelium-dependent and -independent mechanisms in healthy subjects [9]. However, in obesity, high Free Fatty Acids (FFA) directly cause attenuation of the anti-contractile response of PVAT through an endothelium-dependent pathway [10]. We also reported that PVAT around the abdominal aorta plays a crucial part in vascular remodeling in chronic inflammation and suggested that PVAT produces inflammatory mediators that initiated vascular remodeling [11]. However, adipokine secretion in PVAT, in comparison with adipose tissue in other locations, is not known. Understanding PVAT adipokine secretion and its contribution to vascular remodeling may yield novel, important strategies for management of cardiovascular diseases. We examined the correlation of adipokine concentrations with the cell-surface area of adipocytes from adipose tissue from three locations. We investigated the differential expression of adipokines from the SAT, GAT and PVAT of diet-induced obese mice compared with control lean mice.

Diet-induced obese mice
All protocols and experimental procedures were undertaken in accordance with the Guidelines for the Care and Use of Research Animals (National Institutes of Health, Bethesda, MD, USA) as approved by the Animal Care and Use Committee of SingHealth (Singapore).
Eight-week-old male C57BL/6 mice were purchased from the National University of Singapore. One group of mice were fed a high-fat diet (60% kcal as fat; Research Diets, New Brunswick, USA) and control mice were fed a low-fat diet (10% kcal as fat; research diet, New Brunswick, USA) for 6 weeks (n=26 in each group) following procedures described previously [12]. After 6 weeks, mice were sacrificed. Samples of SAT, GAT, PVAT and plasma were collected. Plasma was stored at -80⁰C immediately after isolation.

Measurement of adipocyte sizes
Sample tissue from SAT, GAT and PVAT were fixed in 4% paraformaldehyde, embedded in paraffin, sectioned and stained with Hematoxylin and Eosin (H&E). Measurement of the surface area of adipocyte cells was undertaken on 10 randomly chosen fields on ≥3 sections for each mouse. Adipocyte area was traced and obtained manually with Image Pro Plus ver6.0 (Media Cybernetics, Silver Spring, MD, USA) following methods described previously [11].

Statistical analysis
Data are summarized as the mean ± and Standard Deviation (SD). Differences in means were tested for statistical significance using Student's t-test, and P<0.05 was considered statistically significant.
No adjustment was made for multiple comparisons; however, all P-values are reported permitting adjustment at the discretion of the reader. Data were analyzed using GraphPad Prism ver4.02 (GraphPad, San Diego, CA, USA). Least-squares regression was used to fit trend lines to log-transformed data.

Results and Discussion
Several roles for adipose tissue in the regulation of metabolism have been discovered. Among these are hormonal regulation, energy intake, and fat storage as well as endocrine, paracrine and autocrine functions [16]. In obesity, adipose tissue undergoes hypertrophy and hyperplasia, and secretes several adipokines, resulting in elevated levels of circulatory adipokines. However, adipocyte hypertrophy and location of adipose tissue in relation to adipokine secretion are poorly understood. Some studies have shown a positive correlation between adipocyte size and secretion of adipokines from adipocytes, whereas others suggest that the location of adipose tissue is a predominant factor in regulating adipokine secretion [2,3]. In the present study, we investigated the levels of different adipokines in SAT, GAT and PVAT in diet-induced obese mice compared with control lean mice. We further examined correlations of adipocyte cell surface area with adipokine concentrations among these adipose tissues.
Comparison of cell surface area of adipocytes from different adipose tissue in obese mice has not been previously reported. Our study demonstrated that in lean control mice the surface area of adipocyte cells obtained from GAT (2007 ± 198 µm 2 ) was significantly larger than that from SAT (1078 ± 210 µm 2 , P=0.001) and PVAT (333 ± 46 µm 2 , P=0.001). Similarly, the surface area of adipocyte cells in obese mice obtained from GAT (6793 ± 2273 µm 2 ) was significantly larger than those from SAT (3846 ± 1233 µm 2 , P=0.001) and PVAT (622 ± 57 µm 2 , P=0.001). Moreover, the surface areas of adipocyte cells from GAT, SAT, and PVAT of obese mice were significantly larger than those of lean mice (P=0.03, P=0.03, and P=0.01 respectively, figure 1). The surface   Elevated levels of insulin and adipokines in plasma have been reported in obese subjects as well as several diet-induced and genetically modified animal models. Our study found significantly elevated levels of insulin (P=0.009), leptin (P=0.03), IL-6 (P=0.01) and resistin (P=0.002) in plasma of obese mice compared with control lean mice (Figure 2A).
Adipose-derived hormone leptin mRNA levels in obese humans are known to be substantially higher in SAT than in omental adipose tissue. However, in obese rats these levels are significantly higher in GAT and retroperitoneal tissue than in inguinal adipose tissue [6,7]. Our study revealed significantly higher leptin concentrations in GAT than in SAT and PVAT in both lean mice (P=0.004 and P=0.005, respectively, figure 2B) and obese mice (P=0.02 and P=0.01). Moreover, leptin concentrations in GAT, SAT and PVAT of obese mice were significantly higher than those of control lean mice (P=0.03, P=0.02, and P=0.001). Our data suggest that leptin concentration in adipose tissue is dependent upon the location of adipose tissue. MCP-1 is produced predominantly by macrophages and endothelialcells, and is a potent chemotactic factor for monocytes. Adipose tissue secretes MCP-1 in animal models of obesity, suggesting that MCP-1 in adipose tissue contributes to macrophage infiltration into tissue as well as insulin resistance [17]. Longer induction of high fat diet induced obesity (12 weeks) in these studies compared to shorter induction of high fat diet induced obesity (6 weeks) in our study could explain different severity of obesity and inflammatory changes in these mice. Our study did not detect a significant increase in MCP-1 levels in three adipose tissues.
IL-6 is a proinflammatory cytokine and is elevated in chronic lowgrade inflammation such as obesity. Adipose tissue in obese humans and in animal models produces several pro-inflammatory cytokines, including IL-6. IL-6 is secreted by T cells and macrophages to stimulate immune response. It is also produced by skeletal muscle in response to exercise and reported to have beneficial effects in regulating metabolism in tissue and organs [18]. Accumulating evidence indicates that IL-6 is produced abundantly by visceral areas of gonadal adipocyte cells from lean and obese mice in our study were similar to those reported in other studies [13]. adipose tissue. In addition, plasma concentrations of IL-6 correlate with insulin resistance. In our study, IL-6 concentrations in lean mice did not differ significantly among SAT, GAT and PVAT. However, IL-6 concentrations in PVAT of obese mice were significantly elevated compared with those of lean mice (P=0.03). Moreover, IL-6 concentrations in PVAT were significantly higher than those of SAT and GAT in obese mice (P=0.03 and P=0.005, respectively, figure 2D). Our present study confirms our previous finding of elevated PVAT IL-6 level in chronic inflammation and is consistent with PVAT inflammatory mediation in the development of increased intima-media thickness of abdominal aorta [11]. Our data suggest that higher IL-6 levels in the PVAT of obese mice contribute to elevated IL-6 concentrations in the circulation, but not to IL-6 concentrations in SAT and GAT. Furthermore, our data indicate that increased IL-6 secretion in obesity is dependent upon adipose tissue location.
PAI-1 plays a crucial role in fibrinolytic activity and contributes to the pathogenesis of atherothrombosis. PAI-1 is produced primarily by SAT in human and rodent models of obesity following insulin resistance. Our data are consistent with published work and show that PAI-1 is produced by SAT as mice become obese [19]. In obese mice, PAI-1 concentrations were significantly elevated in SAT but not in GAT or PVAT, (P=0.01, figure 2E), suggesting that PAI-1 production from adipose tissue is dependent upon adipose tissue location.
Resistin (adipose tissue-specific secretory factor) is produced by adipose tissue in humans and rodents. In obesity, plasma levels of resistin are elevated and several studies have shown a positive correlation with insulin resistance. Our study confirmed significantly elevated resistin plasma levels in obese mice and demonstrated the complex nature of resistin levels in adipose tissue in different locations. In lean mice, resistin concentrations in GAT were significantly higher than those of SAT and PVAT (P=0.001 in both, figure 2F). However, in obese mice, significantly increased concentrations of resistin were observed in SAT (P=0.02) but not in GAT and PVAT. Our data suggest that adipose tissues at different locations have different functions as obesity develops. The secretion of adipokines from adipose tissue may be dependent upon the location of adipose tissue. The correlation between adipokine concentration and surface area of adipocyte cells was examined. Among the five adipokine concentrations in three types of adipose tissues in our study, adipokine concentrations of MCP-1, IL-6, PAI-1 and resistin did not correlate with adipocyte cell surface area (data not shown).
However, leptin concentration was significantly correlated with adipocyte cell surface area (Figure 3) in both lean and obese mice. The high r 2 values obtained from the regression analysis demonstrated a strong association between leptin concentration and adipocyte cell surface area for all three types of adipose tissues. Moreover, the highest concentrations of leptin and the strongest correlations with adipocyte cell surface area were found in GAT in both lean and obese mice. Our study demonstrated that leptin concentration was primarily dependent upon location of the adipose tissues rather than the surface area of adipocyte cells.
In conclusion, the present study clearly demonstrated that adipocyte tissue cell surface area was primarily dependent upon location-subcutaneous, gonadal or perivascular. Moreover, we suggest that differential adipokine concentrations in different types of adipose tissues may lead to elevated levels of adipokine concentrations in plasma in a mouse model of obesity. Furthermore, our prior and current studies suggest that increased IL-6 level in PVAT may contribute to vascular remodeling.