Human primary adipocytes exhibit immune cell function: adipocytes prime inflammation independent of macrophages

Abstract

Background: Obesity promotes inflammation in adipose tissue (AT) and this is implicated in pathophysiological complications such as insulin resistance, type 2 diabetes and cardiovascular disease. Although based on the classical hypothesis, necrotic AT adipocytes (ATA) in obese state activate AT macrophages (ATM) that then lead to a sustained chronic inflammation in AT, the link between human adipocytes and the source of inflammation in AT has not been in-depth and systematically studied. So we decided as a new hypothesis to investigate human primary adipocytes alone to see whether they are able to prime inflammation in AT.

Methods and results: Using mRNA expression, human preadipocytes and adipocytes express the cytokines/chemokines and their receptors, MHC II molecule genes and 14 acute phase reactants including C-reactive protein. Using multiplex ELISA revealed the expression of 50 cytokine/chemokine proteins by human adipocytes. Upon lipopolysaccharide stimulation, most of these adipocyte-associated cytokines/chemokines and immune cell modulating receptors were up-regulated and a few down-regulated such as (ICAM-1, VCAM-1, MCP-1, IP-10, IL-6, IL-8, TNF-α and TNF-β highly up-regulated and IL-2, IL-7, IL-10, IL-13 and VEGF down-regulated. In migration assay, human adipocyte-derived chemokines attracted significantly more CD4+ T cells than controls and the number of migrated CD4+ cells was doubled after treating the adipocytes with LPS. Neutralizing MCP-1 effect produced by adipocytes reduced CD4+ migration by approximately 30%.

Conclusion: Human adipocytes express many cytokines/chemokines that are biologically functional. They are able to induce inflammation and activate CD4+ cells independent of macrophages. This suggests that the primary event in the sequence leading to chronic inflammation in AT is metabolic dysfunction in adipocytes, followed by production of immunological mediators by these adipocytes, which is then exacerbated by activated ATM, activation and recruitment of immune cells. This study provides novel knowledge about the prime of inflammation in human obese adipose tissue, opening a new avenue of investigations towards obesity-associated type 2 diabetes.

Figure 1. Differentiation of human pre-adipocytes into adipocytes.

Panel A shows human pre-adipocytes (long, thin and flatted cells without detectable fat organelles). Panel B shows that differentiated fat cells (adipocytes) are spherical with differences in fat organelle size. Most space in adipocytes is occupied by fat organelles.

Figure 2. The differentiation markers for adipocytes.

Illumina beadarray was used to screen adipocytes differentiation markers. These markers are Adiponectin (AdipoQ), perilipin (PLIN), adipose triglyceride lipase (PNPLA2) and fatty acid binding protein (FABP4). Gene expression was expressed as ²log (e.g. the difference between pre-adipocytes and adipocytes for adiponectin expression is approximately 10, thus the true difference is 2¹⁰ ( = 1024-fold) and this is shown on the y-axis).

Figure 3. Confocal laser scan microscopy (CLSM) analysis of Perilipin in adipocytes.

CLSM was used to validate and localize perilipin protein in human adipocytes as shown in figure 3. Fat organelles ( =  lipid droplets) were stained with FITC (green color), indicated by green arrow, and Perilipin was visualized with Texas red (red color) shown by red arrow. Perilipin was localized at the surface of fat organelles.

Figure 4. The expression of immune-associated components (cytokine and chemokine genes) by human primary adipocytes.

Figure 4 A–C shows cytokines and chemokines and their receptor gene expression by human adipocytes, using Illumine BeadArray. Illumina gene profiling analysis shows the expression of cytokines and chemokines in human pre- adipocytes (red bars) and adipocytes (blue bars). Gene expression was expressed as ²log and is shown on the y-axis. Illumina results were derived from two independent experiments. Each experiment was done in duplicate.

Figure 5. A comparison analysis of 50 cytokines/chemokines between adipocytes and adipocytes treated with LPS.

Figures 5 A and B show up-regulation above ratio 3 (panel A) and from ratio 1 to 3 (panel B) and down-regulation (panel C, ratio under 1), respectively, of cytokine and chemokine proteins before and after lipopolysaccharide (LPS) stimulation (200 ng/ml with a final concentration of 1 µg for 24 h), using a multiplex ELISA cytokine/chemokine protein BeadArray. Protein expression was expressed as the ratio of LPS-treated/untreated and is shown on the y-axis. Preadipocytes are shown in red bars and adipocytes in blue bars.

Figure 6. The correlation between cytokines/chemokines and their receptors.

Figure 6 shows a significant negative correlation between cytokines/chemokines and their receptor genes. Illumina results were derived from two independent experiments. Each experiment was done in duplicate.

Figure 7. The localization and visualization of C-RP in the human adipocytes using confocal microscopy strategy.

Confocal microscopy analysis was used to localize C-reactive protein (CRP) in the human adipocytes. An adipocyte wet slide was stained only with IgG isotype and alexa 647 coupled goat anti-mouse (panel A); considered as a negative control). This slide was not incubated with primary monoclonal antibody (Mab) against human CRP. Another adipocyte wet slide was also stained with Mab against human CRP and bound antibodies were displayed with alexa 647 coupled goat anti-mouse (panel A and E; red color). Fat organelle (lipid droplets) was visualized with FITC (C; green color), and adipocytes were stained with DAPI to detect nuclei (panel D; blue color). Panel E shows a merging of all three labels. Fluorescent labeling was used for all detections. CRP was found to be positive in cytoplasm and plasma membrane of human adipocytes and this confirmed the gene expression of CRP in both pre-adipocytes and adipocytes obtained by Illumina BeadArray ( Table 1 ). White and red arrows indicate the presence of CRP on cytoplasm and plasma membrane, respectively (panel B).

Figure 8. Localization of C-RP in human adipocytes using Immunoelectron microscopy.

Panel A served as a negative control and this slide was not incubated with primary monoclonal antibody (Mab) against human CRP, but with IgG isotype. Panel B slide was labeled with Mab against human CRP and bound antibodies were displayed with coupled goat anti-mouse 15 nm gold particles. White arrow shows one CRP gold-stained particle (Black dot).

Figure 9. The expression of the MHC class II genes, receptors involved in (co)stimulation of T cells and immune-associated cluster differentiation (CD) by human primary pre-adipocytes and adipocytes.

Panel A shows the expression of MHC class II genes and some genes involved in co-stimulation of T cells in human pre-adipocytes (red bars) and adipocytes (blue bars). Gene expression was given as ²log and shown on the y-axis. Illumina results were obtained from two independent experiments. Each experiment was done in duplicate, thus, each sample was measured four times. Panel B shows CD gene expression by human adipocytes, using Illumina BeadArray. Illumina gene profiling analysis shows the expression of CD genes in human pre-adipocytes (red bars) and adipocytes (blue bars). Gene expression was expressed as ²log and shown on the y-axis.

Figure 10. The activation of CD4+ cells by human adipocytes.

Panel A shows the analysis of CD4+ cells migration caused by adipocytes-produced chemokines after 10 h incubation. The lower compartment (insert) contained attached adipocytes treated with LPS (200 ng/ml) or untreated, as well as antibodies against MCP-1 and IP-10 (2 µg/ml) to neutralize the effect of these two chemo-attractants. The upper chamber (insert) contained purified and rested CD4+ T cells. The cells were incubated at 37°C for 10 h. The inserts were then removed and the cells migrated toward the adipocytes were collected in culture medium without serum. The number of migrated cells was obtained by cell counting using a Bürker-Türk chamber. This figure represents two independent experiments with similar results (only differ in the number of migrated cells). Each experiment was done in triplicate. Error bars show standard deviations of means for triplicate cultures per condition. The p-value obtained from an independent two-tailed test samples (T-test). P<0.05 was accepted as statistically significant. The controls in the migration assay were 1- culture medium without serum, 2- culture medium serum free with LPS, and culture medium serum free with IgG isotype (this is not shown in Figure 10 A and 10 B, because this was executed separately and the obtained results were the same as culture medium serum free with LPS). Panel B shows the analysis of CD4+ cells migration caused by adipocyte-produced chemokines after 30 h incubation. All cell migration conditions were similar to the conditions described in panel A, except that the cells were incubated at 37°C for 30 h.