As the major cell type of adipose tissue, adipocytes secrete several proinflammatory cytokines that may contribute to inflammation associated with obesity [1, 2]. We have recently demonstrated increased levels of expression of inflammation genes in adipocytes of obese Pima Indian subjects [3].

In addition to adipocytes, adipose tissue contains cell types such as preadipocytes/stromal vascular cells (SVC), mesenchymal stem cells, macrophages, endothelial cells and fibroblasts. These various cell types may play different roles in adipose tissue inflammation associated with obesity. For example, it has been postulated that macrophages resident in adipose tissue contribute to obesity-associated inflammation and insulin resistance [4, 5]. Furthermore, it has been shown that preadipocytes/SVC can function as macrophage-like cells and may play a direct role in inflammation response and obesity [6, 7].

In this study, we compared the gene expression profiles of cultured abdominal subcutaneous preadipocytes/SVC from obese and non-obese non-diabetic Pima Indians to determine the potential role of this cell type in obesity, particularly in obesity-related inflammation.

Subjects, materials and methods

Subjects, metabolic measurements and fat biopsies

The subject selection criteria, metabolic measurements and fat biopsies are described in detail in the accompanying manuscript [3]. The anthropometric and metabolic variables of the subjects who participated in the preadipocyte experiments are described in Table 1, while the measurements for the subjects who took part in the adipocyte experiments are detailed in the accompanying manuscript [3].

Table 1 Clinical characteristics of the subjects who participated in the microarray and quantitative RT-PCR confirmation studies

Preadipocyte isolation and culture

Collagenase digestion of the subcutaneous abdominal adipose tissue biopsy samples was carried out as described previously [8]. The floating adipocytes were isolated as described by Lee et al [3]. The preadipocyte-containing infranatant was collected into a separate tube and washed several times with Hank’s Balanced Salt Solution. The stromal vascular fraction pellet containing preadipocytes/SVC was resuspended in standard medium consisting of Medium 199 (Life Technologies, Grand Island, NY, USA) supplemented with 1 μg/ml amphotericin B, 100 U/ml penicillin G sodium, 100 μg/ml streptomycin sulphate, 2 mmol/l Glutamax-1 and 10% heat-inactivated FBS (Life Technologies). The suspension was strained through a sterile 25-μm stainless steel tissue sieve (Thermo EC, Holbrook, NY, USA). The filtrate was transferred to a T75 culture flask and maintained in an incubator at 37°C in 5% CO2. Cells were allowed to attach and, the next day, floating red blood cells were removed by aspiration and the culture media was replenished. At sub-confluency, the cultured cells were trypsinised and plated at a concentration of approximately 1.5×106 cells/15-cm dish for RNA extraction, which was carried out approximately 14 days after the biopsy date. Media was changed every 2–3 days throughout the culturing period. There were no differences (in terms of cell number or visible morphology) between the cultures from the non-obese and obese subjects.


Microarray analyses were carried out on approximately 33,000 genes represented by approximately 45,000 probe sets on Affymetrix Genechip arrays (HG-U133 A and B; Santa Clara, CA, USA) in 14 non-obese and 14 obese Pima Indian subjects. Target preparation, array hybridisation and scanning and processing were as described in Lee et al [3]. The data for absolute analysis have been deposited in NCBI Gene Expression Omnibus (GEO;, GEO Series accession number GSE2510).

Quantitative real-time PCR

Quantification of transcripts was carried out using TaqMan Assays-on-Demand Gene Expression Products (Applied Biosystems, Foster City, CA, USA) as described previously [8] in an independent group of subjects (Table 1). The mRNA levels of each gene were normalised to those of the gene encoding the TATA-box binding protein (TBP) as described in the preceding article [3]. Ten subjects (six non-obese and four obese) were members of both independent subject groups for quantitative RT-PCR in preadipocytes (Table 1) and adipocytes [3], i.e. preadipocytes and adipocytes were simultaneously isolated from these ten individuals. The clinical characteristics of the two independent subject groups were not statistically different.

Statistical analyses

The anthropometric characteristics of the subjects who participated in the microarray study and the independent group of subjects who took part in the quantitative RT-PCR experiment were statistically analysed using the Student’s t-test and Wilcoxon tests (Statistical Analysis System software, SAS Institute, Cary, NC, USA). The Mann–Whitney U-test (MAS5.0, Affymetrix) was carried out to analyze differentially expressed genes in the microarray data. Q-RT-PCR data of selected inflammation-related genes between cultured preadipocytes/SVC of the obese and the non-obese subjects were analyzed using the two-tailed, Student’s t-test with p≤0.05 considered significant.


Of the approximately 45,000 probe sets analysed by microarray hybridisation, 341 probe sets were differentially expressed between the two groups of subjects (p<0.01), as determined by Affymetrix software. Excluding replicate probe sets and genes with unknown functions, 218 genes were annotated as described in the preceding article [3] (Electronic Supplementary Material [ESM] Table 1). The inflammation category consisted of 23 genes, 17 of which were upregulated in obese subjects (Table 2).

Table 2 Inflammation-related genes

We measured the levels of expression of ten representative inflammation-related genes by quantitative RT-PCR in samples of preadipocytes/SVC from an independent group of non-obese and obese subjects. Of the ten genes, six (IL8, CTSS, ITGB2, HLA-DRA, CD53 and PLA2G7) showed significant (p<0.05) differential expression. CCL18 showed a trend (p=0.07) to be upregulated in samples from obese subjects, whereas the levels of expression of ICAM1, CD59 and DAF were not significantly different between the two groups (p>0.09) (Table 2). In addition, we confirmed the upregulation of MMP9, a gene in the extracellular matrix category that encodes a matrix metalloprotease (microarray, p=0.005; quantitative RT-PCR, p=0.004).

According to the microarray data, only six inflammation-related genes (IL8, CTSS, ITGB2, HLA-DRA, CD53, and CD59) were differentially regulated between obese and non-obese subjects in both adipocytes (see ESM Table 1 in [3]) and preadipocytes/SVC (Table 2). In obese subjects, CD59 was downregulated in preadipocytes/SVC and upregulated in adipocytes. We compared the expression levels of the six inflammation-related genes by quantitative RT-PCR in the independent groups of subjects (ESM Fig. 1). The results indicated that IL8, CTSS, ITGB2, HLA-DRA and CD53 were significantly upregulated in cultured preadipocytes/SVC of obese subjects; HLA-DRA and CD59 were upregulated in the adipocytes of obese subjects; and CD53 only showed a trend to be upregulated in these samples (p=0.06).

Gene expression levels of macrophage markers (EMR1, CD68, CSF1R and CCR2) were also determined by quantitative RT-PCR. Only CD68 expression levels were upregulated (p=0.05) in preadipocytes/SVC of obese subjects. The CSF1R transcript levels were not different between the non-obese and obese subjects (p=0.7), and expression of EMR1 and CCR2 was not detectable in any subject.


In the present study we have demonstrated that cultured preadipocytes/SVC of obese Pima Indian subjects have elevated levels of expression of inflammation-related genes. This finding highlights the role of preadipocytes/SVC in adipose tissue inflammation in obesity, similar to the upregulation of inflammation-related genes in isolated adipocytes of obese subjects [3]. Adipocyte microarray gene profiling demonstrated the upregulation of several inflammation-related genes that were not upregulated according to the preadipocyte/SVC microarray data, and vice versa. Of the six inflammation-related genes that were differentially expressed in both adipocyte and preadipocyte microarray profiling experiments, only HLA-DRA was upregulated in both the preadipocytes/SVC and adipocytes of obese subjects. The genes IL8, CTSS, CD53 and ITGB2 were upregulated only in the preadipocytes/SVC of the obese subjects. Other inflammation-related genes that were upregulated in adipocytes of obese subjects, e.g. TNF, IL6, CCL2 (encoding monocyte chemoattractant protein-1), and CCL3 (encoding macrophage inflammatory protein 1α), were not differentially expressed in preadipocytes/SVC. The results demonstrate cell-specific patterns of activated inflammation-related genes. Preadipocytes/SVC and adipocytes may play different, but complementary, roles in obesity-related inflammation. However, we cannot rule out the possibility that the differences we observed were due, at least in part, to differences between freshly isolated and cultured cells.

The gene encoding the chemokine IL-8 (IL8) was upregulated in cultured preadipocytes/SVC of obese subjects in this study. Our data support previous work [9, 10] indicating that IL-8 levels secreted by non-adipocyte cells in human adipose tissue correlate positively with BMI. The proinflammatory IL-8 upregulates the expression of MMP9, a Type IV matrix metalloprotease involved in extracellular matrix remodelling in adipose tissue [11, 12]. As expected, MMP9 expression was elevated in preadipocytes/SVC of obese subjects compared with those of non-obese subjects. Both MMP9 and IL-8 are involved in leucocytosis and stem cell mobilisation [13]. Thus, we hypothesise that the IL-8 and MMP9 secreted by the preadipocytes/SVC may play a role in extracellular matrix changes or adipose tissue remodelling. The remodelling may be required to accommodate adipose tissue expansion and/or recruitment of immune cells into the tissue.

Inflammation-related genes can also be expressed by macrophages; thus, we examined whether the increased expression of these genes in cultured preadipocytes/SVC from obese subjects was due to the presence of potentially contaminating macrophages. Only one (CD68) of the four macrophage markers tested was upregulated in cultured preadipocytes/SVC of obese subjects. These data need to be further corroborated by other methods, such as immunohistochemistry. Nevertheless, the results indicate that although macrophages may be present in the preadipocyte/SVC cultures, they are not likely to be a major contaminant. Thus, we suggest that other cell types in the cultures, probably the preadipocytes/SVC themselves, are the major contributors to the elevated expression of inflammation-related genes.

Collagenase digestion of adipose tissue during isolation of preadipocytes/SVC could potentially stimulate transcription of inflammation-related genes; however, this is rather unlikely. After culturing the cells for at least 48 h, the inflammatory state of preadipocytes/SVC (as assessed by the expression of pro-inflammatory genes TNF and IL1B) should return to the level observed prior to collagenase digestion (X. Xu and A. Ferrante, Naomi Berrie Diabetes Center, Columbia University, NY, USA, personal communication).

To summarise, increased expression of inflammation-related genes in cultured preadipocytes/SVC from obese individuals indicate that these cells, like adipocytes, may contribute to adipose tissue inflammation associated with obesity.