- Email: [email protected]
Depot-specific prostaglandin synthesis in human adipose tissue: A novel possible mechanism of adipogenesis
Gene 380 (2006) 137 – 143 www.elsevier.com/locate/gene
Depot-specific prostaglandin synthesis in human adipose tissue: A novel possible mechanism of adipogenesis Marcus Quinkler a,c , Iwona J. Bujalska a , Jeremy W. Tomlinson a , Dave M. Smith b , Paul M. Stewart a,⁎ a
Division of Medical Sciences, Institute of Biomedical Research, University of Birmingham, Queen Elizabeth Hospital, Birmingham, B15 2TH, UK b Diabetes and Obesity Drug Discovery, AstraZeneca, Cheshire, UK c Department of Clinical Endocrinology, Centre for Internal Medicine, Charité Campus Mitte, Berlin, Germany Received 29 March 2006; received in revised form 22 May 2006; accepted 23 May 2006 Available online 10 June 2006 Received by J.A. Engler
Abstract Despite the magnitude of the obesity epidemic, the mechanisms that contribute to increases in fat mass and to differences in fat depots are still poorly understood. Prostanoids have been proposed as potent adipogenic hormones, e.g. metabolites of prostaglandin J2 (PGJ2) bind and activate PPARγ. We hypothesize that an altered expression of enzymes in PGJ2 synthesis may represent a novel pathogenic mechanism in human obesity. We characterized adipose depot-specific expression of enzymes in PGJ2 synthesis, prostaglandin transporter and PPARγ isoforms. Paired omental and subcutaneous adipose tissue samples were obtained from 26 women undergoing elective abdominal surgery and gene expression examined in whole tissue and cultured preadipocytes using an Affymetrix cDNA microarray technique and validated with quantitative real-time PCR. All enzymes involved in prostaglandin synthesis were expressed in both adipose tissues. Expression of prostaglandin synthase-1 (PGHS1), prostaglandin D synthase (PTGDS), human prostaglandin transporter (hPGT) and PPARγ2 was higher in OM adipose tissue compared to SC, whereas 17β-hydroxysteroid dehydrogenase 5 (AKR1C3) showed predominance in SC adipose tissue. In SC adipose tissue, PGHS1 mRNA expression increased with BMI. The differential, depot-specific expression of key enzymes involved in transport, synthesis and metabolism of prostaglandins may have an important impact upon fat cell biology and may help to explain some of the observed depot-specific differences. In addition, the positive correlation between PGHS1 and BMI offers the novel hypothesis that the regulation of PG synthesis may have a role in determining fat distribution in human obesity. © 2006 Elsevier B.V. All rights reserved. Keywords: Obesity; Prostaglandins; PPAR gamma; AKR1C3; Omental and subcutaneous fat
1. Introduction The peroxisome proliferator-activated receptor gamma (PPARγ) has a fundamental role in glucose homeostasis and Abbreviations: AKR, aldo-keto reductase; BMI, body mass index; FCS, fetal calf serum; HSD, hydroxysteroid dehydrogenase; hPGT, human prostaglandin transporter; IL-1β, interleukin 1β; iNOS, inducible NO synthase; OM, omental; PG, prostaglandin; PGHS1, prostaglandin synthase-1; PGJ2, prostaglandin J2; PPARγ, peroxisome proliferator-activated receptor gamma; PTGDS, prostaglandin D synthase; SC, subcutaneous; SD, standard deviation; TNFα, tumor necrosis factor alpha. ⁎ Corresponding author. Tel.: +44 121 627 2380; fax: +44 121 627 2384. E-mail address: [email protected] (P.M. Stewart). 0378-1119/$ - see front matter © 2006 Elsevier B.V. All rights reserved. doi:10.1016/j.gene.2006.05.026
adipocyte differentiation (Auwerx, 1999; Gurnell et al., 2003; Tontonoz et al., 1994). Synthetic PPARγ ligands, such as thiazolidinediones (Lehmann et al., 1995), are increasingly used in the treatment of type 2 diabetes mellitus (Gurnell et al., 2003), improving insulin sensitivity, but also increasing adipogenesis. Several laboratories have begun to search for subtype-specific natural PPAR activators. Besides linoleate, linolenate and arachidonate, the most notable PPAR ligand is 15-deoxy-delta12-14-prostaglandin J2 (15d-PGJ2), a natural derivative of prostaglandin D2 (PGD2) and prostaglandin J2 (PGJ2) formed by spontaneous dehydration in aqueous solutions or by albumin (Fig. 4) (Kikawa et al., 1984; Straus and Glass, 2001). It exerts an EC50 of 2 μM for PPARγ
138
M. Quinkler et al. / Gene 380 (2006) 137–143
mediated activation and a Ki for competitive binding with rosiglitazone of 2.5 μM. 15d-PGJ2 acting via PPARγ seems to play a critical role in fat cell differentiation, inducing the expression of adipocyte-specific genes and promoting the formation of mature lipid-laden adipocytes (Fajas et al., 1997; Hollenberg et al., 1997; Spiegelman, 1998). It is therefore plausible that the production of 15d-PGJ2 within adipose tissue may act as an endogenous mediator of adipocyte differentiation. The metabolism of prostanoids within adipose tissue may also be crucial. The 15d-PGJ2 precursor, PGD2, can be converted by the enzyme aldo-keto reductase 1 C3 (AKR1C3) into PGF2α, which has PPARγ antagonist properties (Fig. 4). We and others have shown that this enzyme is expressed in human adipocytes (Quinkler et al., 2004) and shows high affinity for PGD2 with less preference for the androgen precursor androstenedione and for progesterone (Nishizawa et al., 2000). Therefore, within adipose tissue the ability to generate prostanoid metabolites that can either activate or antagonize PPARγ is a novel ‘pre-receptor’ mechanism that may contribute to the pathogenesis of obesity. On this background we have performed a detailed depotspecific characterization of mRNA expression of enzymes involved in PGD2 and PGJ2 synthesis, of PPARγ and of the human prostaglandin transporter (hPGT) in human omental and subcutaneous adipose tissue. 2. Subjects and methods 2.1. Subjects Paired subcutaneous (SC) and omental (OM) adipose tissue biopsies were obtained from 26 female patients undergoing elective abdominal hysterectomy for uterine myoma or fibroma (median age 43 years, range 30–50, median BMI 27.8 kg/m2, range 19.7–39.2). All were premenopausal, non-diabetic, and none had been treated with glucocorticoids within the preceding twelve months or had been on any hormonal treatment. This study had the approval of the local research ethics committee and written informed consent was obtained in every case. 2.2. Adipocyte and preadipocyte isolation Samples of whole SC and OM fat tissue were used for RNA extraction and for preadipocyte isolation as previously reported (Bujalska et al., 1997). Briefly, adipose tissue biopsies were washed in PBS, then chopped and digested with collagenase class 1 (2 mg/ml; Worthington Biochemical Corp., Reading, UK) in 1× Hanks' Balanced Salt Solution (Life Technologies Inc. Paisley, UK) for 60 min at 37 °C in a shaking water bath. Samples were centrifuged at 90 ×g for 1 min, the top adipocyte layer was removed and stromal-vascular cell suspension was centrifuged again at 90 ×g for 5 min. 2.3. Human preadipocyte cell culture The pellet containing preadipocytes was washed with 1× Hank's balanced salt solution, resuspended in DMEM/Nutrient Mixture F-12 (Life Technologies Inc., Paisley, UK) containing
15% fetal calf serum (Life Technologies Inc., Paisley, UK), penicillin (50,000 U/l) and streptomycin (50,000 μg/l), and seeded in 24 well plates. Cells were left overnight and washed the following day with 1× Hank's balanced salt solution. Preadipocytes were cultured to confluence for 4 to 6 days. DMEM/F12 media with 10% FCS was changed every other day. 2.4. RNA extraction and RT Total RNA was extracted from whole fat tissue biopsies and from preadipocyte cultures using either a single step extraction method (Tri reagent, Sigma, UK [for preadipocytes]), or a Genelute total mammalian RNA extraction kit (Sigma, UK [for whole adipose tissue biopsies]). RNA integrity was assessed by gel electrophoresis and concentrations were measured by spectrophotometry at A260 and checked for purity using the ratio A260/280. Reverse transcription of RNA was performed as previously described (Bujalska et al., 1999) employing reagents from Promega (Southampton, UK). 2.5. Microarray Total RNA was extracted from paired SC and OM whole fat tissue as well as from confluent preadipocytes cultures from 5 women. All experiments were performed using Affymetrix HgU133A oligonucleotide array, as described at http://www. affymetrix.com/products/arrays/specific/hgu133.affyx and complied with MIAME standard. Total RNA was used to prepare biotinylated target RNA with minor modifications from the manufacturer's recommendations (http://www.affymetrix. com/support/technical/manual/expression_manual.affx). Briefly, 10 μg of mRNA (pool of 2 μg of total RNA from each cell preparation) was used to generate first-strand cDNA by using a T7-linked oligo(dT) primer. After second-strand synthesis, in vitro transcription was performed with biotinylated UTP and CTP (Enzo Diagnostics, USA), resulting in approximately 100fold amplification of RNA. The target cDNA generated from each sample was processed as per manufacturer's recommendation using an Affymetrix GeneChip Instrument System (http://www.affymetrix.com/support/technical/manual/ expression_manual.affx). Briefly, spike controls were added to 15 μg fragmented cDNA before overnight hybridisation. Arrays were then washed and stained with streptavidin– phycoerythrin, before being scanned on an Affymetrix GeneChip scanner. A complete description of these procedures is available at http://bioinf.picr.man.ac.uk/mbcf/downloads/ GeneChip_Hyb_Wash_Scan_Protocol_v_2_web.pdf. Additionally, quality and amount of starting RNA were confirmed using an agarose gel. After scanning, array images were assessed by eye to confirm scanner alignment and the absence of significant bubbles or scratches on the chip surface. 3′/5′ ratios for GAPDH and beta-actin were confirmed to be within acceptable limits (0.96–1.14), and BioB spike controls were found to be present on all chips, with BioC, BioD and CreX also present in increasing intensity. When scaled to a target intensity of 100 (using Affymetrix MAS 5.0 array analysis software), scaling factors for all arrays were within acceptable limits (0.8–1.1), as
M. Quinkler et al. / Gene 380 (2006) 137–143
were background, Q values and mean intensities. Affymetrix software Microarray Data Suite 5.0 was used for analyzing microarray data and returned as an Excel spreadsheet. Accordingly, results are presented as qualitative (= detection) and quantitative (= signal) measures of expression level, which represent arbitrary data. The inter-assay variability between microarrays was below 5%. The software Gene Expression Pattern Analysis Suite v2.0 (http://gepas.bioinfo.cipf.es/cgi-bin/tools) was used to normalise the microarray chips and assess the RNA quality. The methods were normalisation (quantiles), perfect match correction (pmonly) and summary (medianpolish). The RNA digestion (degradation) plot showed expression as a function of 5′–3′ position of the probes for all 4 microarrays. Although the slope of the plots indicated some RNA degradation, parallel trends in all samples showed similar degradation pattern. The Boxplot graph of the raw Perfect Match (PM)-intensities showed minor differences in the distribution of intensities across arrays, however, after the raw data was processed (background corrected, normalised, PM-MM adjusted and summarized in a single number within each probe set), the box plots for each array looked the same: being centred in the same points and showing similar variability. 2.6. Quantitative PCR PGHS1, PTGDS, hPGT, PPARγ and PPARγ2 mRNA expression was analyzed using an ABI Prism 7700 sequence detection system (Perkin-Elmer Applied Biosystems, Warrington, UK) that employs TaqMan chemistry for highly accurate mRNA quantification as previously described (Bujalska et al., 2002). All reactions were multiplexed with the housekeeping gene 18S (Perkin-Elmer). Reactions were as follows: 50 °C for 2 min, 95 °C for 10 min, and then 44 cycles of 95 °C for 15 s and 60 °C for 1 min. 5 sequences of primers and probe used for real-time PCR were designed using PrimerExpress software (Applied Biosystems, UK) and shown in Table 1. Data were expressed as CT values (= cycle number at which logarithmic PCR plots cross a calculated threshold line) and used to determine ΔCT values (= CT of the target gene minus CT of the housekeeping gene; high ΔCT values represent low levels of expression). Fold changes were calculated using transformation [fold increase = 2− difference in dCT]. 2.7. Adipocyte size measurements OM and SC adipose tissue samples from 10 patients were analyzed for adipocyte size in SC and OM depots. Tissue sections
139
were stained with hematoxylin and eosin. The cell diameter was measured by two independent investigators blinded to both the depot and the patients' details. Cells were counted from 3 regions of known area in each section, and the mean cell diameter calculated (expressed in μm). 2.8. Statistical analysis Data are expressed as mean ± SD. Statistical analysis on realtime PCR data was performed on mean ΔCT values to exclude potential bias owing to averaging data that had been transformed through the equation 2− CT. Statistical analysis of comparisons between groups was undertaken using paired and unpaired t tests where appropriate, otherwise the Mann– Whitney Rank Sum Test was used. Linear regression analysis was performed using Pearson's correlation coefficient. 3. Results 3.1. Affymetrix microarray analysis PGHS1, PTGDS, AKR1C3, and PPARγ were expressed in human adipose tissue from both SC and OM depots (Table 2A). PGHS1 and PTGDS were more highly expressed (2.1 and 3.2 respectively) in OM adipose tissue compared to SC tissue, whereas PPARγ expression was similar in both depots (Table 2A). Human prostaglandin transporter (hPGT) was expressed in OM whole adipose tissue, but was below detection level in the SC depot (Table 2A). PGHS1, PTGDS, and AKR1C3 were expressed in both SC and OM undifferentiated human preadipocytes cultured to confluence, whereas PPARγ and hPGT were not expressed in preadipocytes (Table 2B). 3.2. Quantitative PCR for site specific expression Quantitative real-time PCR on cDNA generated from paired SC and OM adipose biopsies from 26 women was used to validate the microarray findings. Endorsing these results, PGHS1 and PTGDS mRNA expression levels were higher in OM compared with SC adipose tissue (p < 0.005 and p < 0.001, respectively) (Fig. 1). In addition, we found no significant difference in total PPARγ mRNA expression between SC and OM whole adipose tissue, once again confirming our microarray data (Fig. 1). We further characterized PPARγ isoform expression, because Affymetrix microarray probes for PPARγ were not isoform specific. When we used our designed isoform-
Table 1 Primer and probe sequences (5′ to 3′ orientation) used for real-time PCR in the present study Forwarda
Reverseb
Probec
Gene
GenBank no.
AGCAGCTGAGTGGCTATTTCCT AGAAGAAGGCGGCGTTGTC GCCAAGCTGCTCCAGAAAAT AGAAGAAGGCGGCGTTGTC GGGATCTCCTATGTGGATGACTT
CGGTTGCGGTATTGGAACTG TCAGGTTGAGGCCACCATC TGATCACCTGCAGTAGCTGCA TCAGGTTGAGGCCACCATC GCCGGTCCAAATACAGAGAT
ACACCGAACAGCAGCTCTGGGTCAAATT AGGGGCCACCACAGACTTGCACAT ACAGACCTCAGACAGATTGTCACGGAACA AGGGGCCACCACAGACTTGCACAT CCCAGCAACTCGCCCCTGTACATCT
PGHS1 PTGDS PPARγ PPARγ2 hPGT
AH001520.2 NM_000954 NM_138712.1 NM_015869.2-33 NM_005630
a,b,c
Forward and reverse primers and probes used for quantification by real-time PCR. Sequences are indicated from 5′ end to 3′ end of oligonucleotides.
140
M. Quinkler et al. / Gene 380 (2006) 137–143
Table 2 Expression of genes involved in PGD and PGJ2 synthesis according to microarray from whole adipose tissue (panel A) and preadipocytes (panel B) Enzyme (gene)
SCa Detectionc
A Whole adipose tissue PGHS1 P PTGDS P AKR1C3 P PPARγ P hPGT A B Preadipocytes PGHS1 P PTGDS P AKR1C3 P PPARγ A hPGT A
OMb
Fold changee
Signald
Detectionc
Signald
SC vs OM
54 234 1535 513 67
P P P P P
156 1100 460 316 105
2.1 (I)⁎ 3.2 (I)⁎ 3.2 (D)⁎ NS NS
152 726 511 56 45
P P P A A
325 686 73 26 46
2.0 (I)⁎ NS 3.8 (D)⁎ NS NS
a,b
Subcutaneous (SC) and omental (OM) fat samples from five women were pooled. Detection = qualitative measure indicating if the transcript is reliably detected (P = present) or not detected (A = absent). d Signal = quantitative measure of the relative abundance of a transcript. e Fold change = change in expression level (more than 2-fold) for a transcript between SC and OM. I = increase, D = decrease, NS = not significant. Change p values (⁎) <0.001. c
Fig. 2. In subcutaneous human adipose tissue from 26 women undergoing elective abdominal surgery, expression of PGHS1, the rate limiting step in prostanoid biosynthesis, correlates positively with body mass index (BMI). High 1 / ΔCT values represent high levels of expression and vice versa.
3.4. Adipocyte size measurements Thiazolidinediones influence the number of small and large adipocytes (Okuno et al., 1998). Assuming PGJ2 has similar properties to thiazolidinediones, we investigated the adipocyte size in OM and SC adipocytes. Where H and E sections were obtained (n = 10), adipocyte diameter was significantly higher in
specific primers/probe for PPARγ2, there was significantly higher expression of PPARγ2 in OM compared to SC whole adipose tissue (p < 0.001) (Fig. 1). The human prostaglandin transporter (hPGT) showed a significant higher expression in OM than in SC whole adipose tissue (p < 0.05) (Fig. 1). 3.3. Correlation of gene expression with BMI The expression of many of the enzymatic steps in prostanoid synthesis as well as expression of PPARγ isoforms in both OM and SC adipocytes was independent of BMI (data not shown). PGHS1 mRNA expression levels correlated positively with BMI in SC but not OM adipocytes (R2 = 0.22, p < 0.05) (Fig. 2).
Fig. 1. Depot-specific quantitative mRNA expression (in ΔCT values) of key enzymes in PGD and PGJ2 biosynthesis as measured by real-time PCR in omental (OM) and subcutaneous (SC) whole adipose tissue from 26 women undergoing elective abdominal surgery. High ΔCT values represent low levels of expression and vice versa. Mean ΔCT levels with appropriate standard errors of the mean (SEM) are shown. PGHS1 = prostaglandin synthase 1; PTGDS = prostaglandin D synthase; AKR1C3 = aldo-keto reductase 1C3 = 17βhydroxysteroid dehydrogenase type 5; PPAR = peroxisome proliferatoractivated receptor; hPGT = human prostaglandin transporter (hPGT; SLC21A2). n = 5 for hPGT.
Fig. 3. Human adipocytes (hematoxylin and eosin staining) from omental (A) and subcutaneous (B) biopsies from women undergoing elective abdominal hysterectomy. Magnification ×200 in both figures. (C) Cell diameters were significant higher in subcutaneous adipocytes than in omental adipocytes (n = 10 paired samples).
M. Quinkler et al. / Gene 380 (2006) 137–143
the subcutaneous cells (mean ± SD; OM = 90.4 ± 4.7 μm; SC = 119.5 ± 4.3 μm, p < 0.001) (Fig. 3). 4. Discussion We have performed a detailed characterization of depotspecific mRNA expression of enzymes involved in prostaglandin D and J2 synthesis in adipose tissue in human obesity. Depotspecific differences in gene expression in human adipose tissue are well described (Montague et al., 1998). Furthermore, the relative contribution of depot-specific fat mass to morbidity and mortality is striking. Increases in central, omental adiposity are associated with increased morbidity and mortality (Fontbonne et al., 1992), whilst increased subcutaneous adiposity is relatively benign. Characterizing gene expression in adipose tissue depots may help us understand these fundamental observations. We have shown that all of the key enzymes in PGD and PGJ2 synthesis are expressed in SC and OM human adipose tissue, but PGHS1 (or COX1) and PTGDS are more highly expressed in OM adipose tissue. One could hypothesize that this might lead to increased adipocyte differentiation within the OM depot. However, in reality the situation is likely to be more complex, certainly given the observation that PPARγ activation by thiazolidinediones has a far more potent impact upon SC adipose tissue (Adams et al., 1997). Whilst the biosynthesis and metabolism of prostaglandins are complex (Straus and Glass, 2001), PGHS1 is one of the rate limiting steps. The positive correlation with BMI in SC adipose tissue (Fig. 2) is interesting and may point to a possible pathogenic mechanism through increased generation of prostaglandin metabolites able to activate PPARγ. The lack of this relationship in the omental depot could be a reflection of the reported relative insensitivity of PPARγ activation by thiazolidinediones. However, the correlation of PGHS1 mRNA expression
141
and BMI in sc depot is weak and may not necessarily represent a causative relationship. The product of PGHS1, PGH2, is a substrate for PTGDS, the latter showing a very high expression in both adipose tissue depots, but especially high in omental fat (low ΔCT value in Fig. 1). The central role of PTGDS is emphasized by the observations that PGD2 is the major prostaglandin found in most tissues of adult rats (Ujihara et al., 1988). Expression of AKR1C3 (17β-HSD type 5) adds an additional layer of complexity by potentially generating PPARγ antagonizing ligands (Fig. 4) and depleting the adipose tissue pool of PGD2. Expression of AKR1C3 is higher in SC adipose tissue, and in SC tissue only, expression correlates with BMI (Quinkler et al., 2004). The relative balance of PGHS1 and AKR1C3 expression may be important as this is likely to control the proportion of PPARγ activating and antagonizing ligands. Due to our large and well defined cohort of women, we believe that our data explains a possible mechanism for depot-specific adipocyte differentiation in humans. PPARγ expression is fundamental to our hypothesis. Mutations in PPARγ causing a constitutively active PPARγ lead to increased adipocyte differentiation and obesity (Ristow et al., 1998), whereas loss of function mutations of PPARγ results in a lean body stature with characteristic body fat distribution (lipodystrophy) (Gurnell, 2003). Addition of 15d-PGJ2, a natural PPARγ ligand (Kliewer et al., 1995; Nosjean and Boutin, 2002), enhances markedly the differentiation of human preadipocytes from SC sites. In contrast, preadipocytes from OM sites in the same individuals were refractory to thiazolidinediones, although PPARγ was expressed at similar levels in both depots (Adams et al., 1997). The mechanism of this depotspecific thiazolidinedione response is unknown. It is also of interest that patients with loss of function mutations of PPARγ lose subcutaneous yet gain central fat (Gurnell, 2003). The
Fig. 4. Suggested schematic pathway of prostaglandin D and J2 synthesis in subcutaneous (left side) and omental (right side) human fat. PGHS1 = prostaglandin synthase 1; PTGDS = prostaglandin D synthase; AKR1C3 = aldo-keto reductase 1C3 = 17β-hydroxysteroid dehydrogenase type 5; PPAR = peroxisome proliferatoractivated receptor; hPGT = human prostaglandin transporter (hPGT; SLC21A2).
142
M. Quinkler et al. / Gene 380 (2006) 137–143
presence of tissue-specific distribution of PPARγ isoforms and the variable ratio of PPARγ1 to PPARγ2 raises the possibility that isoform-specific expression may be modulated in disease states such as non-insulin-dependent diabetes mellitus (Mukherjee et al., 1997). In response to low ligand concentrations, PPARγ2 shows a quantitatively greater ability to induce adipogenesis than PPARγ1 (Mueller et al., 2002; Ren et al., 2002; Saladin et al., 1999; Werman et al., 1997). In addition during the adipogenic pathway, cells undergo a large increase in PPARγ2 expression, whereas the levels of PPARγ1 do not change (Fajas et al., 1997). Although we did not find differences in total PPARγ expression between SC and OM depots, expression of PPARγ2 was higher in OM adipose tissue. Interestingly, the PPARγ2 knockout mouse underpins the hypothesis that changes in the relative expression of PPARγ1 and PPARγ2 may cause depot-specific adipose tissue hypertrophy or hyperplasia in vivo (Medina-Gomez et al., 2005). Adipose tissue is an important source of cytokines (Ahima and Flier, 2000) and obesity contributes to a proinflammatory milieu. The PPARγ ligand 15d-PGJ2 is described as a key anti-inflammatory mediator inhibiting production of inducible NO synthase (iNOS), TNFα and interleukin 1β (IL-1β) (Rossi et al., 2000; Castrillo et al., 2000), and there is evidence that these effects of 15d-PGJ2 are partly independent of PPARγ activation (Straus et al., 2000; Vaidya et al., 1999). However, the influence of 15dPGJ2 and AKR1C3 on the inflammatory role of adipose tissue needs to be established. Recently, the role of 15d-PGJ2 as a PPARγ agonist was questioned by showing low picomolar concentrations in the medium of mouse-derived 3T3-L1 preadipocytes and no increase in its concentrations during adipocyte differentiation (Bell-Parikh et al., 2003; Tzameli et al., 2004). Although these observation were made in murine preadipocytes, 15d-PGJ2 concentrations were also reported at very low levels in human biological fluids (Bell-Parikh et al., 2003). However, an unknown endogenous PPARγ ligand is present during differentiation in mouse-derived 3T3-L1 (Tzameli et al., 2004), that may possibly be derived from the prostaglandin or eicosanoid pathway. Thiazolidinediones increased the number of small adipocytes in adipose tissue of obese rats, whereas the number of large adipocytes was decreased (Okuno et al., 1998). Assuming PGJ2 has similar properties to thiazolidinediones, increased concentrations of PGJ2 within omental fat would result in an increase in small adipocytes and a decrease in the number of large adipocytes. In keeping with other studies we confirmed a significant increase in cell size in subcutaneous adipocytes compared to omental adipocytes (Fig. 3); it remains to be seen whether site specific differences in PGJ2 explain the adipocyte size difference observed between subcutaneous and omental fat depots of women. Ragolia et al. (2005) described mice with a knockout of the gene encoding for PTGDS, which had adipocyte hypertrophy independent of the diet. This is in keeping with our data that PTGDS is nearly 20-times higher expressed in omental than subcutaneous adipocytes and that omental adipocytes are smaller than subcutaneous adipocytes. Prostaglandins diffuse better through the plasma membrane than thiazolidinediones, although at physiological pH prostaglandins dominate as organic anions, which diffuse poorly through the plasma membrane. Lu et al. (1996) have identified
the human prostaglandin transporter hPGT, which is encoded by the gene SLC21A2. hPGT is expressed in several human tissues, such as kidney, heart, liver and brain (Lu et al., 1996) as well as in endothelial cells (Topper et al., 1998), but until now expression of hPGT in human adipose tissue has not been discussed. To our knowledge, this is the first description of hPGT expression in human adipose tissue. hPGT was not expressed in preadipocytes, and showed a higher expression in OM whole adipose tissue. The higher hPGT expression is associated with a higher expression of enzymes involved in prostaglandin synthesis (Table 2, Fig. 1) in the OM depot. In conclusion, the differential, depot-specific expression of key enzymes involved in synthesis and metabolism of PGD and PGJ2 may have an important impact upon fat cell biology and may help to explain some of the observed depot-specific differences in adipocyte differentiation. In addition, it offers a novel pathogenic mechanism for increasing omental adiposity through increased endogenous production of PPARγ agonists that can act in a paracrine and/or autocrine manner. Acknowledgements We thank the consultants and theatre staff at the Women's Hospital Birmingham, UK, for the fat biopsies. We thank all women who participated in the study for their excellent collaboration. This work was supported by the Deutsche Forschungsgemeinschaft (Research Fellowship QU142/1-1 to M.Q.), and the Medical Research Council (Research Training Fellowship to J.W.T., and Senior Clinical Fellowship to P.M.S.). References Adams, M., et al., 1997. Activators of peroxisome proliferator-activated receptor gamma have depot-specific effects on human preadipocyte differentiation. J. Clin. Invest. 100, 3149–3153. Ahima, R.S., Flier, J.S., 2000. Adipose tissue as an endocrine organ. Trends Endocrinol. Metab. 11, 327–332. Auwerx, J., 1999. PPARgamma, the ultimate thrifty gene. Diabetologia 42, 1033–1049. Bell-Parikh, L.C., Ide, T., Lawson, J.A., McNamara, P., Reilly, M., FitzGerald, G.A., 2003. Biosynthesis of 15-deoxy-delta12, 14-PGJ2 and the ligation of PPARgamma. J. Clin. Invest. 112, 945–955. Bujalska, I.J., Kumar, S., Stewart, P.M., 1997. Does central obesity reflect “Cushing's disease of the omentum”? Lancet 349, 1210–1213. Bujalska, I.J., Kumar, S., Hewison, M., Stewart, P.M., 1999. Differentiation of adipose stromal cells: the roles of glucocorticoids and 11beta-hydroxysteroid dehydrogenase. Endocrinology 140, 3188–3196. Bujalska, I.J., Walker, E.A., Hewison, M., Stewart, P.M., 2002. A switch in dehydrogenase to reductase activity of 11beta-hydroxysteroid dehydrogenase type 1 upon differentiation of human omental adipose stromal cells. J. Clin. Endocrinol. Metab. 87, 1205–1210. Castrillo, A., az-Guerra, M.J., Hortelano, S., Martin-Sanz, P., Bosca, L., 2000. Inhibition of IkappaB kinase and IkappaB phosphorylation by 15-deoxyDelta(12, 14)-prostaglandin J(2) in activated murine macrophages. Mol. Cell. Biol. 20, 1692–1698. Fajas, L., et al., 1997. The organization, promoter analysis, and expression of the human PPARgamma gene. J. Biol. Chem. 272, 18779–18789. Fontbonne, A., Thibult, N., Eschwege, E., Ducimetiere, P., 1992. Body fat distribution and coronary heart disease mortality in subjects with impaired glucose tolerance or diabetes mellitus: the Paris Prospective Study, 15-year follow-up. Diabetologia 35, 464–468.
M. Quinkler et al. / Gene 380 (2006) 137–143 Gurnell, M., 2003. PPARgamma and metabolism: insights from the study of human genetic variants. Clin. Endocrinol. (Oxf) 59, 267–277. Gurnell, M., Savage, D.B., Chatterjee, V.K., O'Rahilly, S., 2003. The metabolic syndrome: peroxisome proliferator-activated receptor gamma and its therapeutic modulation. J. Clin. Endocrinol. Metab. 88, 2412–2421. Hollenberg, A.N., et al., 1997. Functional antagonism between CCAAT/ enhancer binding protein-alpha and peroxisome proliferator-activated receptor-gamma on the leptin promoter. J. Biol. Chem. 272, 5283–5290. Kikawa, Y., Narumiya, S., Fukushima, M., Wakatsuka, H., Hayaishi, O., 1984. 9-Deoxy-delta 9, delta 12–13, 14-dihydroprostaglandin D2, a metabolite of prostaglandin D2 formed in human plasma. Proc. Natl. Acad. Sci. U.S.A. 81, 1317–1321. Kliewer, S.A., Lenhard, J.M., Willson, T.M., Patel, I., Morris, D.C., Lehmann, J.M., 1995. A prostaglandin J2 metabolite binds peroxisome proliferator-activated receptor gamma and promotes adipocyte differentiation. Cell 83, 813–819. Lehmann, J.M., Moore, L.B., Smith-Oliver, T.A., Wilkison, W.O., Willson, T.M., Kliewer, S.A., 1995. An antidiabetic thiazolidinedione is a high affinity ligand for peroxisome proliferator-activated receptor gamma (PPAR gamma). J. Biol. Chem. 270, 12953–12956. Lu, R., Kanai, N., Bao, Y., Schuster, V.L., 1996. Cloning, in vitro expression, and tissue distribution of a human prostaglandin transporter cDNA (hPGT). J. Clin. Invest. 98, 1142–1149. Medina-Gomez, G., et al., 2005. The link between nutritional status and insulin sensitivity is dependent on the adipocyte-specific peroxisome proliferatoractivated receptor-gamma2 isoform. Diabetes 54, 1706–1716. Montague, C.T., et al., 1998. Depot-related gene expression in human subcutaneous and omental adipocytes. Diabetes 47, 1384–1391. Mueller, E., et al., 2002. Genetic analysis of adipogenesis through peroxisome proliferator-activated receptor gamma isoforms. J. Biol. Chem. 277, 41925–41930. Mukherjee, R., Jow, L., Croston, G.E., Paterniti Jr., J.R., 1997. Identification, characterization, and tissue distribution of human peroxisome proliferatoractivated receptor (PPAR) isoforms PPARgamma2 versus PPARgamma1 and activation with retinoid X receptor agonists and antagonists. J. Biol. Chem. 272, 8071–8076. Nishizawa, M., et al., 2000. Close kinship of human 20alpha-hydroxysteroid dehydrogenase gene with three aldo-keto reductase genes. Genes Cells 5, 111–125. Nosjean, O., Boutin, J.A., 2002. Natural ligands of PPARgamma: are prostaglandin J(2) derivatives really playing the part? Cell Signal 14, 573–583. Okuno, A., et al., 1998. Troglitazone increases the number of small adipocytes without the change of white adipose tissue mass in obese Zucker rats. J. Clin. Invest. 101, 1354–1361. Quinkler, M., et al., 2004. Androgen generation in adipose tissue in women with simple obesity—a site-specific role for 17beta-hydroxysteroid dehydrogenase type 5. J. Endocrinol. 183, 331–342.
143
Ragolia, L., Palaia, T., Hall, C.E., Maesaka, J.K., Eguchi, N., Urade, Y., 2005. Accelerated glucose intolerance, nephropathy, and atherosclerosis in prostaglandin D2 synthase knock-out mice. J. Biol. Chem. 280, 29946–29955. Ren, D., Collingwood, T.N., Rebar, E.J., Wolffe, A.P., Camp, H.S., 2002. PPARgamma knockdown by engineered transcription factors: exogenous PPARgamma2 but not PPARgamma1 reactivates adipogenesis. Genes Dev. 16, 27–32. Ristow, M., Muller-Wieland, D., Pfeiffer, A., Krone, W., Kahn, C.R., 1998. Obesity associated with a mutation in a genetic regulator of adipocyte differentiation. N. Engl. J. Med. 339, 953–959. Rossi, A., et al., 2000. Anti-inflammatory cyclopentenone prostaglandins are direct inhibitors of IkappaB kinase. Nature 403, 103–108. Saladin, R., Fajas, L., Dana, S., Halvorsen, Y.D., Auwerx, J., Briggs, M., 1999. Differential regulation of peroxisome proliferator activated receptor gamma1 (PPARgamma1) and PPARgamma2 messenger RNA expression in the early stages of adipogenesis. Cell Growth Differ. 10, 43–48. Spiegelman, B.M., 1998. PPAR-gamma: adipogenic regulator and thiazolidinedione receptor. Diabetes 47, 507–514. Straus, D.S., Glass, C.K., 2001. Cyclopentenone prostaglandins: new insights on biological activities and cellular targets. Med. Res. Rev. 21, 185–210. Straus, D.S., et al., 2000. 15-deoxy-delta 12, 14-prostaglandin J2 inhibits multiple steps in the NF-kappa B signaling pathway. Proc. Natl. Acad. Sci. U. S. A. 97, 4844–4849. Tontonoz, P., Hu, E., Spiegelman, B.M., 1994. Stimulation of adipogenesis in fibroblasts by PPAR gamma 2, a lipid-activated transcription factor. Cell 79, 1147–1156. Topper, J.N., et al., 1998. Human prostaglandin transporter gene (hPGT) is regulated by fluid mechanical stimuli in cultured endothelial cells and expressed in vascular endothelium in vivo. Circulation 98, 2396–2403. Tzameli, I., et al., 2004. Regulated production of a peroxisome proliferatoractivated receptor-gamma ligand during an early phase of adipocyte differentiation in 3T3-L1 adipocytes. J. Biol. Chem. 279, 36093–36102. Ujihara, M., Urade, Y., Eguchi, N., Hayashi, H., Ikai, K., Hayaishi, O., 1988. Prostaglandin D2 formation and characterization of its synthetases in various tissues of adult rats. Arch. Biochem. Biophys. 260, 521–531. Vaidya, S., Somers, E.P., Wright, S.D., Detmers, P.A., Bansal, V.S., 1999. 15Deoxy-Delta12, 1412, 14-prostaglandin J2 inhibits the beta2 integrindependent oxidative burst: involvement of a mechanism distinct from peroxisome proliferator-activated receptor gamma ligation. J. Immunol. 163, 6187–6192. Werman, A., Hollenberg, A., Solanes, G., Bjorbaek, C., Vidal-Puig, A.J., Flier, J.S., 1997. Ligand-independent activation domain in the N terminus of peroxisome proliferator-activated receptor gamma (PPARgamma). Differential activity of PPARgamma1 and -2 isoforms and influence of insulin. J. Biol. Chem. 272, 20230–20235.
Depot-specific prostaglandin synthesis in human adipose tissue: A novel possible mechanism of adipogenesis Marcus Quinkler a,c , Iwona J. Bujalska a , Jeremy W. Tomlinson a , Dave M. Smith b , Paul M. Stewart a,⁎ a
Division of Medical Sciences, Institute of Biomedical Research, University of Birmingham, Queen Elizabeth Hospital, Birmingham, B15 2TH, UK b Diabetes and Obesity Drug Discovery, AstraZeneca, Cheshire, UK c Department of Clinical Endocrinology, Centre for Internal Medicine, Charité Campus Mitte, Berlin, Germany Received 29 March 2006; received in revised form 22 May 2006; accepted 23 May 2006 Available online 10 June 2006 Received by J.A. Engler
Abstract Despite the magnitude of the obesity epidemic, the mechanisms that contribute to increases in fat mass and to differences in fat depots are still poorly understood. Prostanoids have been proposed as potent adipogenic hormones, e.g. metabolites of prostaglandin J2 (PGJ2) bind and activate PPARγ. We hypothesize that an altered expression of enzymes in PGJ2 synthesis may represent a novel pathogenic mechanism in human obesity. We characterized adipose depot-specific expression of enzymes in PGJ2 synthesis, prostaglandin transporter and PPARγ isoforms. Paired omental and subcutaneous adipose tissue samples were obtained from 26 women undergoing elective abdominal surgery and gene expression examined in whole tissue and cultured preadipocytes using an Affymetrix cDNA microarray technique and validated with quantitative real-time PCR. All enzymes involved in prostaglandin synthesis were expressed in both adipose tissues. Expression of prostaglandin synthase-1 (PGHS1), prostaglandin D synthase (PTGDS), human prostaglandin transporter (hPGT) and PPARγ2 was higher in OM adipose tissue compared to SC, whereas 17β-hydroxysteroid dehydrogenase 5 (AKR1C3) showed predominance in SC adipose tissue. In SC adipose tissue, PGHS1 mRNA expression increased with BMI. The differential, depot-specific expression of key enzymes involved in transport, synthesis and metabolism of prostaglandins may have an important impact upon fat cell biology and may help to explain some of the observed depot-specific differences. In addition, the positive correlation between PGHS1 and BMI offers the novel hypothesis that the regulation of PG synthesis may have a role in determining fat distribution in human obesity. © 2006 Elsevier B.V. All rights reserved. Keywords: Obesity; Prostaglandins; PPAR gamma; AKR1C3; Omental and subcutaneous fat
1. Introduction The peroxisome proliferator-activated receptor gamma (PPARγ) has a fundamental role in glucose homeostasis and Abbreviations: AKR, aldo-keto reductase; BMI, body mass index; FCS, fetal calf serum; HSD, hydroxysteroid dehydrogenase; hPGT, human prostaglandin transporter; IL-1β, interleukin 1β; iNOS, inducible NO synthase; OM, omental; PG, prostaglandin; PGHS1, prostaglandin synthase-1; PGJ2, prostaglandin J2; PPARγ, peroxisome proliferator-activated receptor gamma; PTGDS, prostaglandin D synthase; SC, subcutaneous; SD, standard deviation; TNFα, tumor necrosis factor alpha. ⁎ Corresponding author. Tel.: +44 121 627 2380; fax: +44 121 627 2384. E-mail address: [email protected] (P.M. Stewart). 0378-1119/$ - see front matter © 2006 Elsevier B.V. All rights reserved. doi:10.1016/j.gene.2006.05.026
adipocyte differentiation (Auwerx, 1999; Gurnell et al., 2003; Tontonoz et al., 1994). Synthetic PPARγ ligands, such as thiazolidinediones (Lehmann et al., 1995), are increasingly used in the treatment of type 2 diabetes mellitus (Gurnell et al., 2003), improving insulin sensitivity, but also increasing adipogenesis. Several laboratories have begun to search for subtype-specific natural PPAR activators. Besides linoleate, linolenate and arachidonate, the most notable PPAR ligand is 15-deoxy-delta12-14-prostaglandin J2 (15d-PGJ2), a natural derivative of prostaglandin D2 (PGD2) and prostaglandin J2 (PGJ2) formed by spontaneous dehydration in aqueous solutions or by albumin (Fig. 4) (Kikawa et al., 1984; Straus and Glass, 2001). It exerts an EC50 of 2 μM for PPARγ
138
M. Quinkler et al. / Gene 380 (2006) 137–143
mediated activation and a Ki for competitive binding with rosiglitazone of 2.5 μM. 15d-PGJ2 acting via PPARγ seems to play a critical role in fat cell differentiation, inducing the expression of adipocyte-specific genes and promoting the formation of mature lipid-laden adipocytes (Fajas et al., 1997; Hollenberg et al., 1997; Spiegelman, 1998). It is therefore plausible that the production of 15d-PGJ2 within adipose tissue may act as an endogenous mediator of adipocyte differentiation. The metabolism of prostanoids within adipose tissue may also be crucial. The 15d-PGJ2 precursor, PGD2, can be converted by the enzyme aldo-keto reductase 1 C3 (AKR1C3) into PGF2α, which has PPARγ antagonist properties (Fig. 4). We and others have shown that this enzyme is expressed in human adipocytes (Quinkler et al., 2004) and shows high affinity for PGD2 with less preference for the androgen precursor androstenedione and for progesterone (Nishizawa et al., 2000). Therefore, within adipose tissue the ability to generate prostanoid metabolites that can either activate or antagonize PPARγ is a novel ‘pre-receptor’ mechanism that may contribute to the pathogenesis of obesity. On this background we have performed a detailed depotspecific characterization of mRNA expression of enzymes involved in PGD2 and PGJ2 synthesis, of PPARγ and of the human prostaglandin transporter (hPGT) in human omental and subcutaneous adipose tissue. 2. Subjects and methods 2.1. Subjects Paired subcutaneous (SC) and omental (OM) adipose tissue biopsies were obtained from 26 female patients undergoing elective abdominal hysterectomy for uterine myoma or fibroma (median age 43 years, range 30–50, median BMI 27.8 kg/m2, range 19.7–39.2). All were premenopausal, non-diabetic, and none had been treated with glucocorticoids within the preceding twelve months or had been on any hormonal treatment. This study had the approval of the local research ethics committee and written informed consent was obtained in every case. 2.2. Adipocyte and preadipocyte isolation Samples of whole SC and OM fat tissue were used for RNA extraction and for preadipocyte isolation as previously reported (Bujalska et al., 1997). Briefly, adipose tissue biopsies were washed in PBS, then chopped and digested with collagenase class 1 (2 mg/ml; Worthington Biochemical Corp., Reading, UK) in 1× Hanks' Balanced Salt Solution (Life Technologies Inc. Paisley, UK) for 60 min at 37 °C in a shaking water bath. Samples were centrifuged at 90 ×g for 1 min, the top adipocyte layer was removed and stromal-vascular cell suspension was centrifuged again at 90 ×g for 5 min. 2.3. Human preadipocyte cell culture The pellet containing preadipocytes was washed with 1× Hank's balanced salt solution, resuspended in DMEM/Nutrient Mixture F-12 (Life Technologies Inc., Paisley, UK) containing
15% fetal calf serum (Life Technologies Inc., Paisley, UK), penicillin (50,000 U/l) and streptomycin (50,000 μg/l), and seeded in 24 well plates. Cells were left overnight and washed the following day with 1× Hank's balanced salt solution. Preadipocytes were cultured to confluence for 4 to 6 days. DMEM/F12 media with 10% FCS was changed every other day. 2.4. RNA extraction and RT Total RNA was extracted from whole fat tissue biopsies and from preadipocyte cultures using either a single step extraction method (Tri reagent, Sigma, UK [for preadipocytes]), or a Genelute total mammalian RNA extraction kit (Sigma, UK [for whole adipose tissue biopsies]). RNA integrity was assessed by gel electrophoresis and concentrations were measured by spectrophotometry at A260 and checked for purity using the ratio A260/280. Reverse transcription of RNA was performed as previously described (Bujalska et al., 1999) employing reagents from Promega (Southampton, UK). 2.5. Microarray Total RNA was extracted from paired SC and OM whole fat tissue as well as from confluent preadipocytes cultures from 5 women. All experiments were performed using Affymetrix HgU133A oligonucleotide array, as described at http://www. affymetrix.com/products/arrays/specific/hgu133.affyx and complied with MIAME standard. Total RNA was used to prepare biotinylated target RNA with minor modifications from the manufacturer's recommendations (http://www.affymetrix. com/support/technical/manual/expression_manual.affx). Briefly, 10 μg of mRNA (pool of 2 μg of total RNA from each cell preparation) was used to generate first-strand cDNA by using a T7-linked oligo(dT) primer. After second-strand synthesis, in vitro transcription was performed with biotinylated UTP and CTP (Enzo Diagnostics, USA), resulting in approximately 100fold amplification of RNA. The target cDNA generated from each sample was processed as per manufacturer's recommendation using an Affymetrix GeneChip Instrument System (http://www.affymetrix.com/support/technical/manual/ expression_manual.affx). Briefly, spike controls were added to 15 μg fragmented cDNA before overnight hybridisation. Arrays were then washed and stained with streptavidin– phycoerythrin, before being scanned on an Affymetrix GeneChip scanner. A complete description of these procedures is available at http://bioinf.picr.man.ac.uk/mbcf/downloads/ GeneChip_Hyb_Wash_Scan_Protocol_v_2_web.pdf. Additionally, quality and amount of starting RNA were confirmed using an agarose gel. After scanning, array images were assessed by eye to confirm scanner alignment and the absence of significant bubbles or scratches on the chip surface. 3′/5′ ratios for GAPDH and beta-actin were confirmed to be within acceptable limits (0.96–1.14), and BioB spike controls were found to be present on all chips, with BioC, BioD and CreX also present in increasing intensity. When scaled to a target intensity of 100 (using Affymetrix MAS 5.0 array analysis software), scaling factors for all arrays were within acceptable limits (0.8–1.1), as
M. Quinkler et al. / Gene 380 (2006) 137–143
were background, Q values and mean intensities. Affymetrix software Microarray Data Suite 5.0 was used for analyzing microarray data and returned as an Excel spreadsheet. Accordingly, results are presented as qualitative (= detection) and quantitative (= signal) measures of expression level, which represent arbitrary data. The inter-assay variability between microarrays was below 5%. The software Gene Expression Pattern Analysis Suite v2.0 (http://gepas.bioinfo.cipf.es/cgi-bin/tools) was used to normalise the microarray chips and assess the RNA quality. The methods were normalisation (quantiles), perfect match correction (pmonly) and summary (medianpolish). The RNA digestion (degradation) plot showed expression as a function of 5′–3′ position of the probes for all 4 microarrays. Although the slope of the plots indicated some RNA degradation, parallel trends in all samples showed similar degradation pattern. The Boxplot graph of the raw Perfect Match (PM)-intensities showed minor differences in the distribution of intensities across arrays, however, after the raw data was processed (background corrected, normalised, PM-MM adjusted and summarized in a single number within each probe set), the box plots for each array looked the same: being centred in the same points and showing similar variability. 2.6. Quantitative PCR PGHS1, PTGDS, hPGT, PPARγ and PPARγ2 mRNA expression was analyzed using an ABI Prism 7700 sequence detection system (Perkin-Elmer Applied Biosystems, Warrington, UK) that employs TaqMan chemistry for highly accurate mRNA quantification as previously described (Bujalska et al., 2002). All reactions were multiplexed with the housekeeping gene 18S (Perkin-Elmer). Reactions were as follows: 50 °C for 2 min, 95 °C for 10 min, and then 44 cycles of 95 °C for 15 s and 60 °C for 1 min. 5 sequences of primers and probe used for real-time PCR were designed using PrimerExpress software (Applied Biosystems, UK) and shown in Table 1. Data were expressed as CT values (= cycle number at which logarithmic PCR plots cross a calculated threshold line) and used to determine ΔCT values (= CT of the target gene minus CT of the housekeeping gene; high ΔCT values represent low levels of expression). Fold changes were calculated using transformation [fold increase = 2− difference in dCT]. 2.7. Adipocyte size measurements OM and SC adipose tissue samples from 10 patients were analyzed for adipocyte size in SC and OM depots. Tissue sections
139
were stained with hematoxylin and eosin. The cell diameter was measured by two independent investigators blinded to both the depot and the patients' details. Cells were counted from 3 regions of known area in each section, and the mean cell diameter calculated (expressed in μm). 2.8. Statistical analysis Data are expressed as mean ± SD. Statistical analysis on realtime PCR data was performed on mean ΔCT values to exclude potential bias owing to averaging data that had been transformed through the equation 2− CT. Statistical analysis of comparisons between groups was undertaken using paired and unpaired t tests where appropriate, otherwise the Mann– Whitney Rank Sum Test was used. Linear regression analysis was performed using Pearson's correlation coefficient. 3. Results 3.1. Affymetrix microarray analysis PGHS1, PTGDS, AKR1C3, and PPARγ were expressed in human adipose tissue from both SC and OM depots (Table 2A). PGHS1 and PTGDS were more highly expressed (2.1 and 3.2 respectively) in OM adipose tissue compared to SC tissue, whereas PPARγ expression was similar in both depots (Table 2A). Human prostaglandin transporter (hPGT) was expressed in OM whole adipose tissue, but was below detection level in the SC depot (Table 2A). PGHS1, PTGDS, and AKR1C3 were expressed in both SC and OM undifferentiated human preadipocytes cultured to confluence, whereas PPARγ and hPGT were not expressed in preadipocytes (Table 2B). 3.2. Quantitative PCR for site specific expression Quantitative real-time PCR on cDNA generated from paired SC and OM adipose biopsies from 26 women was used to validate the microarray findings. Endorsing these results, PGHS1 and PTGDS mRNA expression levels were higher in OM compared with SC adipose tissue (p < 0.005 and p < 0.001, respectively) (Fig. 1). In addition, we found no significant difference in total PPARγ mRNA expression between SC and OM whole adipose tissue, once again confirming our microarray data (Fig. 1). We further characterized PPARγ isoform expression, because Affymetrix microarray probes for PPARγ were not isoform specific. When we used our designed isoform-
Table 1 Primer and probe sequences (5′ to 3′ orientation) used for real-time PCR in the present study Forwarda
Reverseb
Probec
Gene
GenBank no.
AGCAGCTGAGTGGCTATTTCCT AGAAGAAGGCGGCGTTGTC GCCAAGCTGCTCCAGAAAAT AGAAGAAGGCGGCGTTGTC GGGATCTCCTATGTGGATGACTT
CGGTTGCGGTATTGGAACTG TCAGGTTGAGGCCACCATC TGATCACCTGCAGTAGCTGCA TCAGGTTGAGGCCACCATC GCCGGTCCAAATACAGAGAT
ACACCGAACAGCAGCTCTGGGTCAAATT AGGGGCCACCACAGACTTGCACAT ACAGACCTCAGACAGATTGTCACGGAACA AGGGGCCACCACAGACTTGCACAT CCCAGCAACTCGCCCCTGTACATCT
PGHS1 PTGDS PPARγ PPARγ2 hPGT
AH001520.2 NM_000954 NM_138712.1 NM_015869.2-33 NM_005630
a,b,c
Forward and reverse primers and probes used for quantification by real-time PCR. Sequences are indicated from 5′ end to 3′ end of oligonucleotides.
140
M. Quinkler et al. / Gene 380 (2006) 137–143
Table 2 Expression of genes involved in PGD and PGJ2 synthesis according to microarray from whole adipose tissue (panel A) and preadipocytes (panel B) Enzyme (gene)
SCa Detectionc
A Whole adipose tissue PGHS1 P PTGDS P AKR1C3 P PPARγ P hPGT A B Preadipocytes PGHS1 P PTGDS P AKR1C3 P PPARγ A hPGT A
OMb
Fold changee
Signald
Detectionc
Signald
SC vs OM
54 234 1535 513 67
P P P P P
156 1100 460 316 105
2.1 (I)⁎ 3.2 (I)⁎ 3.2 (D)⁎ NS NS
152 726 511 56 45
P P P A A
325 686 73 26 46
2.0 (I)⁎ NS 3.8 (D)⁎ NS NS
a,b
Subcutaneous (SC) and omental (OM) fat samples from five women were pooled. Detection = qualitative measure indicating if the transcript is reliably detected (P = present) or not detected (A = absent). d Signal = quantitative measure of the relative abundance of a transcript. e Fold change = change in expression level (more than 2-fold) for a transcript between SC and OM. I = increase, D = decrease, NS = not significant. Change p values (⁎) <0.001. c
Fig. 2. In subcutaneous human adipose tissue from 26 women undergoing elective abdominal surgery, expression of PGHS1, the rate limiting step in prostanoid biosynthesis, correlates positively with body mass index (BMI). High 1 / ΔCT values represent high levels of expression and vice versa.
3.4. Adipocyte size measurements Thiazolidinediones influence the number of small and large adipocytes (Okuno et al., 1998). Assuming PGJ2 has similar properties to thiazolidinediones, we investigated the adipocyte size in OM and SC adipocytes. Where H and E sections were obtained (n = 10), adipocyte diameter was significantly higher in
specific primers/probe for PPARγ2, there was significantly higher expression of PPARγ2 in OM compared to SC whole adipose tissue (p < 0.001) (Fig. 1). The human prostaglandin transporter (hPGT) showed a significant higher expression in OM than in SC whole adipose tissue (p < 0.05) (Fig. 1). 3.3. Correlation of gene expression with BMI The expression of many of the enzymatic steps in prostanoid synthesis as well as expression of PPARγ isoforms in both OM and SC adipocytes was independent of BMI (data not shown). PGHS1 mRNA expression levels correlated positively with BMI in SC but not OM adipocytes (R2 = 0.22, p < 0.05) (Fig. 2).
Fig. 1. Depot-specific quantitative mRNA expression (in ΔCT values) of key enzymes in PGD and PGJ2 biosynthesis as measured by real-time PCR in omental (OM) and subcutaneous (SC) whole adipose tissue from 26 women undergoing elective abdominal surgery. High ΔCT values represent low levels of expression and vice versa. Mean ΔCT levels with appropriate standard errors of the mean (SEM) are shown. PGHS1 = prostaglandin synthase 1; PTGDS = prostaglandin D synthase; AKR1C3 = aldo-keto reductase 1C3 = 17βhydroxysteroid dehydrogenase type 5; PPAR = peroxisome proliferatoractivated receptor; hPGT = human prostaglandin transporter (hPGT; SLC21A2). n = 5 for hPGT.
Fig. 3. Human adipocytes (hematoxylin and eosin staining) from omental (A) and subcutaneous (B) biopsies from women undergoing elective abdominal hysterectomy. Magnification ×200 in both figures. (C) Cell diameters were significant higher in subcutaneous adipocytes than in omental adipocytes (n = 10 paired samples).
M. Quinkler et al. / Gene 380 (2006) 137–143
the subcutaneous cells (mean ± SD; OM = 90.4 ± 4.7 μm; SC = 119.5 ± 4.3 μm, p < 0.001) (Fig. 3). 4. Discussion We have performed a detailed characterization of depotspecific mRNA expression of enzymes involved in prostaglandin D and J2 synthesis in adipose tissue in human obesity. Depotspecific differences in gene expression in human adipose tissue are well described (Montague et al., 1998). Furthermore, the relative contribution of depot-specific fat mass to morbidity and mortality is striking. Increases in central, omental adiposity are associated with increased morbidity and mortality (Fontbonne et al., 1992), whilst increased subcutaneous adiposity is relatively benign. Characterizing gene expression in adipose tissue depots may help us understand these fundamental observations. We have shown that all of the key enzymes in PGD and PGJ2 synthesis are expressed in SC and OM human adipose tissue, but PGHS1 (or COX1) and PTGDS are more highly expressed in OM adipose tissue. One could hypothesize that this might lead to increased adipocyte differentiation within the OM depot. However, in reality the situation is likely to be more complex, certainly given the observation that PPARγ activation by thiazolidinediones has a far more potent impact upon SC adipose tissue (Adams et al., 1997). Whilst the biosynthesis and metabolism of prostaglandins are complex (Straus and Glass, 2001), PGHS1 is one of the rate limiting steps. The positive correlation with BMI in SC adipose tissue (Fig. 2) is interesting and may point to a possible pathogenic mechanism through increased generation of prostaglandin metabolites able to activate PPARγ. The lack of this relationship in the omental depot could be a reflection of the reported relative insensitivity of PPARγ activation by thiazolidinediones. However, the correlation of PGHS1 mRNA expression
141
and BMI in sc depot is weak and may not necessarily represent a causative relationship. The product of PGHS1, PGH2, is a substrate for PTGDS, the latter showing a very high expression in both adipose tissue depots, but especially high in omental fat (low ΔCT value in Fig. 1). The central role of PTGDS is emphasized by the observations that PGD2 is the major prostaglandin found in most tissues of adult rats (Ujihara et al., 1988). Expression of AKR1C3 (17β-HSD type 5) adds an additional layer of complexity by potentially generating PPARγ antagonizing ligands (Fig. 4) and depleting the adipose tissue pool of PGD2. Expression of AKR1C3 is higher in SC adipose tissue, and in SC tissue only, expression correlates with BMI (Quinkler et al., 2004). The relative balance of PGHS1 and AKR1C3 expression may be important as this is likely to control the proportion of PPARγ activating and antagonizing ligands. Due to our large and well defined cohort of women, we believe that our data explains a possible mechanism for depot-specific adipocyte differentiation in humans. PPARγ expression is fundamental to our hypothesis. Mutations in PPARγ causing a constitutively active PPARγ lead to increased adipocyte differentiation and obesity (Ristow et al., 1998), whereas loss of function mutations of PPARγ results in a lean body stature with characteristic body fat distribution (lipodystrophy) (Gurnell, 2003). Addition of 15d-PGJ2, a natural PPARγ ligand (Kliewer et al., 1995; Nosjean and Boutin, 2002), enhances markedly the differentiation of human preadipocytes from SC sites. In contrast, preadipocytes from OM sites in the same individuals were refractory to thiazolidinediones, although PPARγ was expressed at similar levels in both depots (Adams et al., 1997). The mechanism of this depotspecific thiazolidinedione response is unknown. It is also of interest that patients with loss of function mutations of PPARγ lose subcutaneous yet gain central fat (Gurnell, 2003). The
Fig. 4. Suggested schematic pathway of prostaglandin D and J2 synthesis in subcutaneous (left side) and omental (right side) human fat. PGHS1 = prostaglandin synthase 1; PTGDS = prostaglandin D synthase; AKR1C3 = aldo-keto reductase 1C3 = 17β-hydroxysteroid dehydrogenase type 5; PPAR = peroxisome proliferatoractivated receptor; hPGT = human prostaglandin transporter (hPGT; SLC21A2).
142
M. Quinkler et al. / Gene 380 (2006) 137–143
presence of tissue-specific distribution of PPARγ isoforms and the variable ratio of PPARγ1 to PPARγ2 raises the possibility that isoform-specific expression may be modulated in disease states such as non-insulin-dependent diabetes mellitus (Mukherjee et al., 1997). In response to low ligand concentrations, PPARγ2 shows a quantitatively greater ability to induce adipogenesis than PPARγ1 (Mueller et al., 2002; Ren et al., 2002; Saladin et al., 1999; Werman et al., 1997). In addition during the adipogenic pathway, cells undergo a large increase in PPARγ2 expression, whereas the levels of PPARγ1 do not change (Fajas et al., 1997). Although we did not find differences in total PPARγ expression between SC and OM depots, expression of PPARγ2 was higher in OM adipose tissue. Interestingly, the PPARγ2 knockout mouse underpins the hypothesis that changes in the relative expression of PPARγ1 and PPARγ2 may cause depot-specific adipose tissue hypertrophy or hyperplasia in vivo (Medina-Gomez et al., 2005). Adipose tissue is an important source of cytokines (Ahima and Flier, 2000) and obesity contributes to a proinflammatory milieu. The PPARγ ligand 15d-PGJ2 is described as a key anti-inflammatory mediator inhibiting production of inducible NO synthase (iNOS), TNFα and interleukin 1β (IL-1β) (Rossi et al., 2000; Castrillo et al., 2000), and there is evidence that these effects of 15d-PGJ2 are partly independent of PPARγ activation (Straus et al., 2000; Vaidya et al., 1999). However, the influence of 15dPGJ2 and AKR1C3 on the inflammatory role of adipose tissue needs to be established. Recently, the role of 15d-PGJ2 as a PPARγ agonist was questioned by showing low picomolar concentrations in the medium of mouse-derived 3T3-L1 preadipocytes and no increase in its concentrations during adipocyte differentiation (Bell-Parikh et al., 2003; Tzameli et al., 2004). Although these observation were made in murine preadipocytes, 15d-PGJ2 concentrations were also reported at very low levels in human biological fluids (Bell-Parikh et al., 2003). However, an unknown endogenous PPARγ ligand is present during differentiation in mouse-derived 3T3-L1 (Tzameli et al., 2004), that may possibly be derived from the prostaglandin or eicosanoid pathway. Thiazolidinediones increased the number of small adipocytes in adipose tissue of obese rats, whereas the number of large adipocytes was decreased (Okuno et al., 1998). Assuming PGJ2 has similar properties to thiazolidinediones, increased concentrations of PGJ2 within omental fat would result in an increase in small adipocytes and a decrease in the number of large adipocytes. In keeping with other studies we confirmed a significant increase in cell size in subcutaneous adipocytes compared to omental adipocytes (Fig. 3); it remains to be seen whether site specific differences in PGJ2 explain the adipocyte size difference observed between subcutaneous and omental fat depots of women. Ragolia et al. (2005) described mice with a knockout of the gene encoding for PTGDS, which had adipocyte hypertrophy independent of the diet. This is in keeping with our data that PTGDS is nearly 20-times higher expressed in omental than subcutaneous adipocytes and that omental adipocytes are smaller than subcutaneous adipocytes. Prostaglandins diffuse better through the plasma membrane than thiazolidinediones, although at physiological pH prostaglandins dominate as organic anions, which diffuse poorly through the plasma membrane. Lu et al. (1996) have identified
the human prostaglandin transporter hPGT, which is encoded by the gene SLC21A2. hPGT is expressed in several human tissues, such as kidney, heart, liver and brain (Lu et al., 1996) as well as in endothelial cells (Topper et al., 1998), but until now expression of hPGT in human adipose tissue has not been discussed. To our knowledge, this is the first description of hPGT expression in human adipose tissue. hPGT was not expressed in preadipocytes, and showed a higher expression in OM whole adipose tissue. The higher hPGT expression is associated with a higher expression of enzymes involved in prostaglandin synthesis (Table 2, Fig. 1) in the OM depot. In conclusion, the differential, depot-specific expression of key enzymes involved in synthesis and metabolism of PGD and PGJ2 may have an important impact upon fat cell biology and may help to explain some of the observed depot-specific differences in adipocyte differentiation. In addition, it offers a novel pathogenic mechanism for increasing omental adiposity through increased endogenous production of PPARγ agonists that can act in a paracrine and/or autocrine manner. Acknowledgements We thank the consultants and theatre staff at the Women's Hospital Birmingham, UK, for the fat biopsies. We thank all women who participated in the study for their excellent collaboration. This work was supported by the Deutsche Forschungsgemeinschaft (Research Fellowship QU142/1-1 to M.Q.), and the Medical Research Council (Research Training Fellowship to J.W.T., and Senior Clinical Fellowship to P.M.S.). References Adams, M., et al., 1997. Activators of peroxisome proliferator-activated receptor gamma have depot-specific effects on human preadipocyte differentiation. J. Clin. Invest. 100, 3149–3153. Ahima, R.S., Flier, J.S., 2000. Adipose tissue as an endocrine organ. Trends Endocrinol. Metab. 11, 327–332. Auwerx, J., 1999. PPARgamma, the ultimate thrifty gene. Diabetologia 42, 1033–1049. Bell-Parikh, L.C., Ide, T., Lawson, J.A., McNamara, P., Reilly, M., FitzGerald, G.A., 2003. Biosynthesis of 15-deoxy-delta12, 14-PGJ2 and the ligation of PPARgamma. J. Clin. Invest. 112, 945–955. Bujalska, I.J., Kumar, S., Stewart, P.M., 1997. Does central obesity reflect “Cushing's disease of the omentum”? Lancet 349, 1210–1213. Bujalska, I.J., Kumar, S., Hewison, M., Stewart, P.M., 1999. Differentiation of adipose stromal cells: the roles of glucocorticoids and 11beta-hydroxysteroid dehydrogenase. Endocrinology 140, 3188–3196. Bujalska, I.J., Walker, E.A., Hewison, M., Stewart, P.M., 2002. A switch in dehydrogenase to reductase activity of 11beta-hydroxysteroid dehydrogenase type 1 upon differentiation of human omental adipose stromal cells. J. Clin. Endocrinol. Metab. 87, 1205–1210. Castrillo, A., az-Guerra, M.J., Hortelano, S., Martin-Sanz, P., Bosca, L., 2000. Inhibition of IkappaB kinase and IkappaB phosphorylation by 15-deoxyDelta(12, 14)-prostaglandin J(2) in activated murine macrophages. Mol. Cell. Biol. 20, 1692–1698. Fajas, L., et al., 1997. The organization, promoter analysis, and expression of the human PPARgamma gene. J. Biol. Chem. 272, 18779–18789. Fontbonne, A., Thibult, N., Eschwege, E., Ducimetiere, P., 1992. Body fat distribution and coronary heart disease mortality in subjects with impaired glucose tolerance or diabetes mellitus: the Paris Prospective Study, 15-year follow-up. Diabetologia 35, 464–468.
M. Quinkler et al. / Gene 380 (2006) 137–143 Gurnell, M., 2003. PPARgamma and metabolism: insights from the study of human genetic variants. Clin. Endocrinol. (Oxf) 59, 267–277. Gurnell, M., Savage, D.B., Chatterjee, V.K., O'Rahilly, S., 2003. The metabolic syndrome: peroxisome proliferator-activated receptor gamma and its therapeutic modulation. J. Clin. Endocrinol. Metab. 88, 2412–2421. Hollenberg, A.N., et al., 1997. Functional antagonism between CCAAT/ enhancer binding protein-alpha and peroxisome proliferator-activated receptor-gamma on the leptin promoter. J. Biol. Chem. 272, 5283–5290. Kikawa, Y., Narumiya, S., Fukushima, M., Wakatsuka, H., Hayaishi, O., 1984. 9-Deoxy-delta 9, delta 12–13, 14-dihydroprostaglandin D2, a metabolite of prostaglandin D2 formed in human plasma. Proc. Natl. Acad. Sci. U.S.A. 81, 1317–1321. Kliewer, S.A., Lenhard, J.M., Willson, T.M., Patel, I., Morris, D.C., Lehmann, J.M., 1995. A prostaglandin J2 metabolite binds peroxisome proliferator-activated receptor gamma and promotes adipocyte differentiation. Cell 83, 813–819. Lehmann, J.M., Moore, L.B., Smith-Oliver, T.A., Wilkison, W.O., Willson, T.M., Kliewer, S.A., 1995. An antidiabetic thiazolidinedione is a high affinity ligand for peroxisome proliferator-activated receptor gamma (PPAR gamma). J. Biol. Chem. 270, 12953–12956. Lu, R., Kanai, N., Bao, Y., Schuster, V.L., 1996. Cloning, in vitro expression, and tissue distribution of a human prostaglandin transporter cDNA (hPGT). J. Clin. Invest. 98, 1142–1149. Medina-Gomez, G., et al., 2005. The link between nutritional status and insulin sensitivity is dependent on the adipocyte-specific peroxisome proliferatoractivated receptor-gamma2 isoform. Diabetes 54, 1706–1716. Montague, C.T., et al., 1998. Depot-related gene expression in human subcutaneous and omental adipocytes. Diabetes 47, 1384–1391. Mueller, E., et al., 2002. Genetic analysis of adipogenesis through peroxisome proliferator-activated receptor gamma isoforms. J. Biol. Chem. 277, 41925–41930. Mukherjee, R., Jow, L., Croston, G.E., Paterniti Jr., J.R., 1997. Identification, characterization, and tissue distribution of human peroxisome proliferatoractivated receptor (PPAR) isoforms PPARgamma2 versus PPARgamma1 and activation with retinoid X receptor agonists and antagonists. J. Biol. Chem. 272, 8071–8076. Nishizawa, M., et al., 2000. Close kinship of human 20alpha-hydroxysteroid dehydrogenase gene with three aldo-keto reductase genes. Genes Cells 5, 111–125. Nosjean, O., Boutin, J.A., 2002. Natural ligands of PPARgamma: are prostaglandin J(2) derivatives really playing the part? Cell Signal 14, 573–583. Okuno, A., et al., 1998. Troglitazone increases the number of small adipocytes without the change of white adipose tissue mass in obese Zucker rats. J. Clin. Invest. 101, 1354–1361. Quinkler, M., et al., 2004. Androgen generation in adipose tissue in women with simple obesity—a site-specific role for 17beta-hydroxysteroid dehydrogenase type 5. J. Endocrinol. 183, 331–342.
143
Ragolia, L., Palaia, T., Hall, C.E., Maesaka, J.K., Eguchi, N., Urade, Y., 2005. Accelerated glucose intolerance, nephropathy, and atherosclerosis in prostaglandin D2 synthase knock-out mice. J. Biol. Chem. 280, 29946–29955. Ren, D., Collingwood, T.N., Rebar, E.J., Wolffe, A.P., Camp, H.S., 2002. PPARgamma knockdown by engineered transcription factors: exogenous PPARgamma2 but not PPARgamma1 reactivates adipogenesis. Genes Dev. 16, 27–32. Ristow, M., Muller-Wieland, D., Pfeiffer, A., Krone, W., Kahn, C.R., 1998. Obesity associated with a mutation in a genetic regulator of adipocyte differentiation. N. Engl. J. Med. 339, 953–959. Rossi, A., et al., 2000. Anti-inflammatory cyclopentenone prostaglandins are direct inhibitors of IkappaB kinase. Nature 403, 103–108. Saladin, R., Fajas, L., Dana, S., Halvorsen, Y.D., Auwerx, J., Briggs, M., 1999. Differential regulation of peroxisome proliferator activated receptor gamma1 (PPARgamma1) and PPARgamma2 messenger RNA expression in the early stages of adipogenesis. Cell Growth Differ. 10, 43–48. Spiegelman, B.M., 1998. PPAR-gamma: adipogenic regulator and thiazolidinedione receptor. Diabetes 47, 507–514. Straus, D.S., Glass, C.K., 2001. Cyclopentenone prostaglandins: new insights on biological activities and cellular targets. Med. Res. Rev. 21, 185–210. Straus, D.S., et al., 2000. 15-deoxy-delta 12, 14-prostaglandin J2 inhibits multiple steps in the NF-kappa B signaling pathway. Proc. Natl. Acad. Sci. U. S. A. 97, 4844–4849. Tontonoz, P., Hu, E., Spiegelman, B.M., 1994. Stimulation of adipogenesis in fibroblasts by PPAR gamma 2, a lipid-activated transcription factor. Cell 79, 1147–1156. Topper, J.N., et al., 1998. Human prostaglandin transporter gene (hPGT) is regulated by fluid mechanical stimuli in cultured endothelial cells and expressed in vascular endothelium in vivo. Circulation 98, 2396–2403. Tzameli, I., et al., 2004. Regulated production of a peroxisome proliferatoractivated receptor-gamma ligand during an early phase of adipocyte differentiation in 3T3-L1 adipocytes. J. Biol. Chem. 279, 36093–36102. Ujihara, M., Urade, Y., Eguchi, N., Hayashi, H., Ikai, K., Hayaishi, O., 1988. Prostaglandin D2 formation and characterization of its synthetases in various tissues of adult rats. Arch. Biochem. Biophys. 260, 521–531. Vaidya, S., Somers, E.P., Wright, S.D., Detmers, P.A., Bansal, V.S., 1999. 15Deoxy-Delta12, 1412, 14-prostaglandin J2 inhibits the beta2 integrindependent oxidative burst: involvement of a mechanism distinct from peroxisome proliferator-activated receptor gamma ligation. J. Immunol. 163, 6187–6192. Werman, A., Hollenberg, A., Solanes, G., Bjorbaek, C., Vidal-Puig, A.J., Flier, J.S., 1997. Ligand-independent activation domain in the N terminus of peroxisome proliferator-activated receptor gamma (PPARgamma). Differential activity of PPARgamma1 and -2 isoforms and influence of insulin. J. Biol. Chem. 272, 20230–20235.