| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Research Centre for Endocrinology and Metabolism (K.S., J.P., L.E.O., P.-A.S., B.C., L.M.S.C.), Division of Body Composition and Metabolism (A.G., T.C.L., J.B., J.S.T., B.C.), Department of Internal Medicine, Department of Radiology (J.B.), Sahlgrenska University Hospital, and Institute of Anatomy and Cell Biology (E.J.), The Sahlgrenska Academy at Göteborg University, SE-413 45 Göteborg, Sweden
Address all correspondence and requests for reprints to: Kajsa Sjöholm, Research Centre for Endocrinology and Metabolism, Vita Stråket 12, Sahlgrenska University Hospital, SE-413 45 Göteborg, Sweden. E-mail: kajsa.sjoholm{at}medic.gu.se.
 |
Abstract |
|---|
 Top Abstract IntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
We show that adipose tissue is a major expression site of A-SAA during the nonacute-phase reaction condition. This provides a direct link between adipose tissue mass and a marker for low-grade inflammation and cardiovascular risk.
|
Introduction |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Expression profiling by microarray has emerged as a powerful tool when screening for regulated genes (2, 3, 4, 5). Another application for expression profiling is the identification of tissue- or cell-specific genes (6). This approach demands a high number of expression profiles from different tissues and cells to obtain a reliable analysis. Due to the specialized function of adipose tissue, it can be assumed that adipocytes express a specific set of genes that other types of cells do not express or only express at low levels. It is possible that the identification of genes predominantly expressed in adipocytes can give new insights into adipocyte function and may thereby provide important information about genes involved in the development of obesity and obesity-associated metabolic disease. Several studies have shown that central obesity, in particular an increase in visceral fat mass, is tightly linked to components of the metabolic syndrome (7). The aim of this study was to search for genes that are predominantly expressed in omental adipocytes. The search was performed in gene expression profiles from human omental adipocytes and 32 other human tissues and cells (2, 3, 4, 6, 8).
We here report that adipose tissue is a major production site of acute-phase serum amyloid A (A-SAA), a known risk factor for coronary artery disease (9, 10). Therefore, our results provide additional support for an important role of adipose tissue and obesity in the development of cardiovascular disease.
|
Subjects and Methods |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Adipose tissue biopsies from the abdominal sc depot and the major omentum were obtained from 10 obese men (2) and four obese women undergoing laparoscopic bariatric surgery. Adipocytes were prepared from sc and omental adipose tissue from three of the obese men (2) using the procedure outlined by Smith et al. (11). A healthy liver sample was used for immunohistochemistry. Abdominal sc adipose tissue biopsies and fasting blood samples were obtained from obese women and men before (wk 0), during (wk 8 and 16), and after (wk 18) diet-induced weight loss using a very-low-calorie diet (VLCD, 450 kcal/d). Subjects were treated with VLCD for 16 wk and then a normal diet was gradually reintroduced. The adipose tissue RNA content was analyzed using microarray and/or real-time PCR. Adipose tissue from one female subject undergoing abdominal surgery was used for Northern blot analysis of A-SAA expression in adipose tissue and adipocytes. Monocytes from healthy blood donors were prepared and differentiated into macrophages as previously described (12). Pooled macrophage RNA was used for microarray analysis as well as for verification with real-time PCR. RNA pools from nasal mucosa biopsies (3), as well as T cells (4), were analyzed by microarray. The Medical Ethics Committee at Göteborg University approved all studies, and all participants gave written informed consent. RNA from the Human Total RNA Master Panel II (Clontech, Palo Alto, CA) was used for verification of the gene expression in different human tissues by real-time PCR. Liver RNA from the RNA panel was used for Northern blot analysis of A-SAA expression.
Measurements
Body composition was determined with computed tomography (CT) before (wk 0) and during (wk 16) the VLCD treatment, using the same parameters as previously described (2). Total, sc, and visceral adipose tissue areas, from the CT scan at the lumbar 4 vertebrae, were used for correlations. The serum A-SAA concentration was measured in duplicate using an ELISA (Biosource International, Camarillo, CA). Other blood chemistry analyses [high sensitive C-reactive protein (hs-CRP), high-density lipoprotein (HDL), and insulin] were performed at the Department of Clinical Chemistry, Sahlgrenska University Hospital. Patient characteristics used for correlations are stated in Table 1
(adipose tissue samples from these patients were analyzed using U133A microarrays; Affymetrix, Santa Clara, CA).
|
Hu95A microarray (Affymetrix) hybridization to RNA from omental adipocytes, sc and omental adipose tissue, nasal mucosa, T cells, and macrophages was performed as previously described (2, 3, 4). RNA from sc adipose tissue biopsies before, during, and after VLCD treatment was hybridized to U133A microarrays (Affymetrix). Hu95A expression profiles from fetal brain, whole brain, cortex, caudate nucleus, amygdala, thalamus, corpus callosum, pituitary gland, cerebellum, spinal cord, dorsal root ganglia, whole blood, testis, prostate, ovary, uterus, salivary gland, trachea, lung, thymus, spleen, adrenal gland, pancreas, thyroid, kidney, fetal liver, liver, and heart were obtained from Su et al. (http://expression.gnf.org; Ref. 6). The expression profile from normal muscle generated by Bakay et al. (8) was obtained from http://microarray.cnmcresearch.org. The Hu95A microarrays of sc and omental adipose tissue as well as omental adipocytes, nasal mucosa, T cells, and macrophages were scaled to an average intensity of 200, the same intensity as the public expression profiles. All VLCD samples were analyzed with wk 0 as baseline. An average difference (AD) was calculated for the replicates of each tissue or cell type and used for all analysis.
Data analysis for Hu95A microarrays
Genes having an AD of less than 20 were given an AD of 20 (13). To identify genes predominantly expressed in adipocytes, two criteria were used: 1) the genes should have an AD of 10 SD higher than the mean AD of all the other tissues, and 2) the genes should have at least a 3-fold higher expression level than the tissue with the second highest AD. These two criteria allow us to identify genes that have an expression level that is clearly separated from both the average expression level of all the tissues, but also from any other highly expressing tissue or cell type (14).
Real-time PCR analysis
Reagents for real-time PCR analysis of aquaporin 7 (AQP7), transport secretion protein-2.2 (TTS-2.2), peptidyl-prolyl isomerase A (PPIA), and large ribosomal protein P0 (RPLP0) (Assays-on-Demand, TaqMan Reverse Transcriptase reagents, and TaqMan Universal PCR Master mix) were purchased from Applied Biosystems (Foster City, CA) and the conditions used were according to the manufacturer’s protocol. The high sequence similarity between the SAA1 and SAA2 isoforms made it impossible to study these isoforms separately. Consequently, probe and primer sequences for A-SAA (Cybergene, Huddinge, Sweden) were designed to span an exon-exon boundary and detect both the SAA1 and SAA2 isoforms, denoted A-SAA (sequences are available upon request). cDNA synthesis was performed in a total reaction volume of 50 µl using 500 ng total RNA. cDNA corresponding to 10 ng RNA per reaction was used for real-time PCR amplification. Amplification and detection of specific products was performed with the ABI PRISM 7900HT Sequence Detection System (Applied Biosystems) using default cycle parameters. A standard curve was plotted for each primer-probe set with a serial dilution of pooled adipose tissue cDNA. Human PPIA or RPLP0 was used as reference to normalize the expression levels between samples. All standards and samples were analyzed in triplicate.
RT-PCR and Northern blot
RT-PCR amplification of A-SAA (primer sequences are available upon request) and SAA3 (primers as previously described in Ref. 15) was performed. The PCRs were purified using the Qiaquick PCR purification kit (Qiagen, Hilden, Germany) and sequenced using the BigDye Terminator v3.1 cycle sequencing kit and an ABI PRISM 3100 Genetic Analyzer (Applied Biosystems). The A-SAA fragment was subcloned into the pCRII vector (Invitrogen, San Diego, CA). ß-Actin (ACTB) (Ambion, Austin, TX) was used as an internal control. Northern blot analysis was performed using total RNA (5 µg) from liver, adipocytes, and adipose tissue. Membranes were probed with 32P-labeled antisense A-SAA and ACTB RNA.
Immunohistochemistry
Omental and sc adipocytes were fixed in buffered paraformaldehyde (4%), embedded in agar (3%), dehydrated in a graded series of ethanol, and embedded in paraffin using standard procedures. Biopsy sections were prepared as previously described (2). SAA1 was localized by incubation with a monoclonal antibody to human SAA1 (HyCult Biotechnology, Uden, The Netherlands) as previously described (2).
Statistical analysis
All analyses were initially performed with standard linear regression. Correlation of within-person longitudinal measurements was addressed with the use of generalized estimating equations (16). Tests of parameters within these models were performed with generalized Wald tests (17), which yielded 
2 statistics. Due to the skewed nature of the measurements, logarithmic transformations were applied to the various A-SAA variables and predictors. For each gene analyzed in Fig. 1, the full model considered had main effects for both depot and gender, as well as an interaction term for potential gender-specific depot effects. In the tests for depot effect, the null hypothesis was that the measurements differed only by gender. In the tests for gender effect, the null hypothesis was that the measurements differed only by depot type. Under each null hypothesis, the test statistic was distributed as 2 with two degrees of freedom. An exact binomial test was used to assess down- (or up-) regulation. In the absence of down-regulation, maximum levels are equally likely to occur at any time point. In the presence of down-regulation, maximum levels will have a disproportionately higher frequency at baseline.
|
|
Results |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Using microarray datasets from 33 human tissues and cells, 28 genes were identified as predominantly expressed in omental adipocytes (Table 2). Some of these genes were expected (e.g. adiponectin, adipose specific 2, hormone-sensitive lipase, AQP7), whereas others were more unexpected (e.g. A-SAA) or not well studied (e.g. TTS-2.2). Literature and database searches were performed to select genes for further validation. To validate the approach used in this study, we confirmed the predominant expression in omental adipocytes for AQP7, TTS-2.2, and A-SAA. In line with previous results (18), AQP7 was shown to be predominantly expressed in adipocytes as determined by real-time PCR analysis. Heart displayed the second highest expression in both the microarray and real-time PCR analysis (Fig. 1
, A and B). The AQP7 expression levels in women were significantly higher than in men after adjusting for depot differences (P < 0.0001; Fig. 1C). However, no difference in AQP7 expression between sc and omental adipose tissue depots could be detected (Fig. 1C). TTS-2.2, which has recently been shown to catalyze the initial step in triglyceride hydrolysis (19), displayed very high expression levels in omental adipocytes, when analyzed by microarray. According to the microarray analysis, the second highest expression was found in heart (Fig. 1A). The TTS-2.2 expression level was three times higher in omental adipocytes as compared with the second highest expression level, detected in skeletal muscle as analyzed with real-time PCR, whereas heart showed the third highest expression (Fig. 1B). The TTS-2.2 expression levels were significantly higher in women compared with men after adjusting for depot differences (P < 0.0001; Fig. 1C). No difference in TTS-2.2 expression between the adipose tissue depots or during diet-induced weight loss (data not shown) could be detected.
|
Tissue panel and adipose tissue depot expression of A-SAA.
Liver has previously been reported to be the main production site for A-SAA (reviewed in Refs. 20 and 21). However, analysis of the microarray datasets showed that A-SAA expression was even higher in human adipocytes, whereas fetal and adult liver displayed the second and third highest expression level (Fig. 1A). This finding was confirmed in two subjects using real-time PCR (Fig. 1B), showing that the omental adipocyte A-SAA expression was four times higher than the expression level in fetal liver and adult liver. There was a significant difference between expression levels both between the adipose tissue depots, where sc adipose tissue showed the highest expression, and gender, where women displayed the highest expression (P < 0.0001 and P < 0.01, respectively; Fig. 1C).
Northern blot, RT-PCR, and immunohistochemistry.
A-SAA expression was also analyzed using Northern blot and RT-PCR. Northern blot confirmed a high expression of A-SAA in adipocytes and adipose tissue, but unexpectedly low signal was detected in human liver (Fig. 2A). Using RT-PCR, A-SAA expression was detected in adipocytes and adipose tissue, as well as in liver (data not shown). To confirm that the amplified A-SAA fragment corresponds to that reported in GenBank, it was verified by sequencing. Immunohistochemistry showed that SAA1 is expressed also at the protein level in both human omental and sc adipocytes as well as in liver (Fig. 2, B and C). The SAA1 staining in adipocytes is focused around the nucleus, where most of the cytoplasm is gathered. In the liver, only some of the hepatocytes stained positive for SAA1, and the most intense staining was seen in the hepatocytes around the central venules.
|
mRNA and serum levels of A-SAA during diet-induced weight loss.
Together with C-reactive protein (CRP), A-SAA has recently been recognized as a cardiovascular risk factor (9, 10, 25). Short-term studies on intentional weight loss have shown drastic improvement in all aspects of the metabolic syndrome. This, together with our observation that A-SAA is highly expressed in adipocytes, prompted us to examine the possible regulation of A-SAA during diet-induced weight loss. During VLCD, A-SAA mRNA levels were down-regulated, according to the microarray (P < 0.05; Fig. 3A). The A-SAA serum levels were also regulated in response to diet-induced weight loss (P < 0.0001; Fig. 3B), and correlated with A-SAA mRNA expression in sc adipose tissue (P < 0.001) as measured by microarray. The A-SAA mRNA expression was significantly higher in women compared with men as analyzed by microarray (P < 0.0001). The down-regulation of mRNA levels between baseline and wk 8 during VLCD was confirmed in a different population, using real-time PCR (n = 30, P < 0.0001; Fig. 3C).
|
|
Discussion |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Serum levels of CRP and A-SAA are highly correlated (28), and this correlation was confirmed also in our patient material. In contrast to A-SAA, we failed to detect CRP mRNA expression in adipose tissue. In a recent publication, obesity was shown to be the major determinant of elevated CRP in subjects with the metabolic syndrome (30). However, it has been proposed that the correlation between BMI and hs-CRP may be driven by adipose-derived cytokines, such as IL-6 (31). It should be emphasized that our data demonstrate that adipose tissue is a major site of A-SAA production under conditions that do not evoke a strong systemic acute-phase response. We show SAA1 immunoreactivity in the cytoplasm around the adipocyte nucleus. In the liver, SAA1 staining is concentrated to hepatocytes surrounding central venules. It may be speculated that the more than 1000-fold increase in serum A-SAA, seen during an infection, could be due to instant release of the A-SAA protein that has been stored in hepatocytes close to the central venules, whereas the more moderate increased A-SAA serum levels, seen in obese patients (10, 28), may be due to a constant secretion of A-SAA from the adipocytes.
The finding that adipose tissue expression of A-SAA was regulated in response to diet-induced weight loss raises the question of the biological role of A-SAA. Although most studies have focused on A-SAA as a marker of inflammation, there are indications that it may influence HDL function. During the acute-phase reaction, A-SAA rapidly associates with HDL, thereby displacing apolipoprotein A1 (20), which normally plays a central role in reverse cholesterol transport from peripheral cells to the liver. Altered activities of HDL-associated enzymes are probably responsible for the impaired antioxidative properties of HDL during the acute phase, suggesting that SAA-HDL is a proinflammatory molecule (32). It has also been speculated that increased levels of A-SAA are responsible for the lowered HDL levels seen during the acute-phase response (33). Therefore, A-SAA production in adipose tissue may provide at least a partial explanation for the low HDL-cholesterol levels frequently seen in obesity. Furthermore, it has been proposed that, in the mouse, A-SAA is a ligand of the cell surface receptor Tanis (29), which is expressed in rat liver and 3T3-L1 cells (29). Tanis displays homology with the human protein AD-015 (29), which we detected in human adipose tissue using real-time PCR (data not shown). Tanis has been suggested as a link between diabetes, inflammatory response, and cardiovascular disease, because it is regulated by glucose and the expression is altered in type 2 diabetes in an animal model (29). Furthermore, we found a significant correlation between SAA serum levels and fasting insulin levels in our patient material.
This report describes an approach for identification of genes predominantly expressed by human omental adipocytes, using microarray expression profiles from a large number of tissues and cells. Efforts have previously been made to identify genes specifically expressed in human adipose tissue. Maeda et al. (34) constructed an adipose tissue cDNA library and compared the expressed genes with those detected in cDNA libraries of 10 other tissues. The approach used in that study is limited by the number of clones that are analyzed and how well they represent the genes expressed in the specific tissue. However, microarray data, in combination with bioinformatic tools, can be used to more extensively and rapidly explore gene expression. In the future, the strategy used here can be applied to datasets from the new generation of microarrays containing all known genes present in the human genome. Hence, it is likely that the list of adipocyte-specific genes will grow. The simplicity of this approach makes it useful for detection of specific gene expression in any particular tissue or cell type of interest.
In the current study, several genes previously known to be expressed predominantly in human adipose tissue were identified, including PPAR-
, adiponectin, adipose specific 2, adipsin, and AQP7 (1, 18, 34). To validate the approach used in this study, the expression of AQP7, was verified. The predominant expression of TTS-2.2, which has recently been shown to catalyze the initial step in triglyceride hydrolysis (19), was also verified. TTS-2.2 was first identified in a cDNA screening project (35), and the protein sequence displays similarities with adiponutrin, an adipose-specific gene that was recently discovered in mice (36, 37) and gene sequence 2 (37). Both the adiponutrin and TTS-2.2 genes are induced early in the differentiation of 3T3-L1 cells (19, 36, 37).
In conclusion, we show in this study that adipose tissue is a major site of production for A-SAA during conditions of the nonacute phase. Previous studies indicate that adipose tissue is an important source of proinflammatory cytokines. Taken together, this suggests that adipose tissue may contribute to the low-grade systemic inflammation seen in obesity. Because SAA-HDL has been proposed to be proatherogenic, the production of A-SAA in adipose tissue may be a link between obesity and atherosclerosis.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
First Published Online December 28, 2004
Abbreviations: ACTB, ß-Actin; AD, average difference; AQP7, aquaporin 7; A-SAA, acute-phase serum amyloid A; BMI, body mass index; CRP, C-reactive protein; CT, computed tomography; hs-CRP, high sensitive CRP; PPIA, peptidyl-prolyl isomerase A; RPLP0, large ribosomal protein P0; TTS-2.2, transport secretion protein-2.2; VLCD, very-low-calorie diet.
Received September 15, 2004.
Accepted December 20, 2004.
|
References |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
B to DNA and the subsequent expression of tumor necrosis factor-
and interleukin-1ß in macrophages. J Clin Invest 98:78–89[Medline]
This article has been cited by other articles:
 |
|
 |
|
  Y. Wu and C. M. Smas Wdnm1-like, a new adipokine with a role in MMP-2 activation Am J Physiol Endocrinol Metab, July 1, 2008; 295(1): E205 - E215. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 B. A. Henry, F. R. Dunshea, M. Gould, and I. J. Clarke Profiling Postprandial Thermogenesis in Muscle and Fat of Sheep and the Central Effect of Leptin Administration Endocrinology, April 1, 2008; 149(4): 2019 - 2026. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 T Lappalainen, M Kolehmainen, U Schwab, L Pulkkinen, D E Laaksonen, R Rauramaa, M Uusitupa, and H Gylling Serum concentrations and expressions of serum amyloid A and leptin in adipose tissue are interrelated: the Genobin Study Eur. J. Endocrinol., March 1, 2008; 158(3): 333 - 341. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. Gummesson, M. Jernas, P.-A. Svensson, I. Larsson, C. A. M. Glad, E. Schele, L. Gripeteg, K. Sjoholm, T. C. Lystig, L. Sjostrom, et al. Relations of Adipose Tissue CIDEA Gene Expression to Basal Metabolic Rate, Energy Restriction, and Obesity: Population-Based and Dietary Intervention Studies J. Clin. Endocrinol. Metab., December 1, 2007; 92(12): 4759 - 4765. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
V. Ceperuelo-Mallafre, M. Miranda, M. R. Chacon, N. Vilarrasa, A. Megia, C. Gutierrez, J. M. Fernandez-Real, J. M. Gomez, E. Caubet, G. Fruhbeck, et al. Adipose Tissue Expression of the Glycerol Channel Aquaporin-7 Gene Is Altered in Severe Obesity But Not in Type 2 Diabetes J. Clin. Endocrinol. Metab., September 1, 2007; 92(9): 3640 - 3645. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. Palming, K. Sjoholm, M. Jernas, T. C. Lystig, A. Gummesson, S. Romeo, L. Lonn, M. Lonn, B. Carlsson, and L. M. S. Carlsson The Expression of NAD(P)H:Quinone Oxidoreductase 1 Is High in Human Adipose Tissue, Reduced by Weight Loss, and Correlates with Adiposity, Insulin Sensitivity, and Markers of Liver Dysfunction J. Clin. Endocrinol. Metab., June 1, 2007; 92(6): 2346 - 2352. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 I J Bujalska, M Quinkler, J W Tomlinson, C T Montague, D M Smith, and P M Stewart Expression profiling of 11{beta}-hydroxysteroid dehydrogenase type-1 and glucocorticoid-target genes in subcutaneous and omental human preadipocytes. J. Mol. Endocrinol., October 1, 2006; 37(2): 327 - 340. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. Jernas, J. Palming, K. Sjoholm, E. Jennische, P.-A. Svensson, B. G. Gabrielsson, M. Levin, A. Sjogren, M. Rudemo, T. C. Lystig, et al. Separation of human adipocytes by size: hypertrophic fat cells display distinct gene expression FASEB J, July 1, 2006; 20(9): 1540 - 1542. [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 |
