Women with polycystic ovary syndrome (PCOS) have high circulating androgens, thought to originate from ovaries and adrenals, and frequently suffer from the metabolic syndrome including obesity. However, serum androgens are positively associated with body mass index (BMI) not only in PCOS, but also in simple obesity, suggesting androgen synthesis within adipose tissue. Thus we investigated androgen generation in human adipose tissue, including expression of 17β-hydroxysteroid dehydrogenase (17β-HSD) isozymes, important regulators of sex steroid metabolism. Paired omental and subcutaneous fat biopsies were obtained from 27 healthy women undergoing elective abdominal surgery (age range 30–50 years; BMI 19.7–39.2 kg/m2). Enzymatic activity assays in preadipocyte proliferation cultures revealed effcient conversion of androstenedione to testosterone in both subcutaneous and omental fat. RT-PCR of whole fat and preadipocytes of subcutaneous and omental origin showed expression of 17β-HSD types 4 and 5, but no relevant expression of 17β-HSD types 1, 2, or 3. Microarray analysis confirmed this expression pattern (17β-HSD5>17β-HSD4) and suggested a higher expression of 17β-HSD5 in subcutaneous fat. Accordingly, quantitative real-time RT-PCR showed significantly higher expression of 17β-HSD5 in subcutaneous compared with omental fat (P<0.05). 17β-HSD5 expression in subcutaneous, but not omental, whole fat correlated significantly with BMI (r=0.51, P<0.05). In keeping with these findings, 17β-HSD5 expression in subcutaneous fat biopsies from six women taking part in a weight loss study decreased significantly with weight loss (P<0.05). A role for 17β-HSD5 in adipocyte differentiation was further supported by the observed increase in 17β-HSD5 expression upon differentiation of stromal preadipocytes to mature adipocytes (n=5; P<0.005), which again was higher in cells of subcutaneous origin. Functional activity of 17β-HSD5 also significantly increased with differentiation, revealing a net gain in androgen activation (androstenedione to testosterone) in subcutaneous cultures, contrasting with a net gain in androgen inactivation (testosterone to androstenedione) in omental cultures. Thus, human adipose tissue is capable of active androgen synthesis catalysed by 17β-HSD5, and increased expression in obesity may contribute to circulating androgen excess.
Hyperandrogenaemia is the principal biochemical finding in the polycystic ovary syndrome (PCOS) (Dunaif & Thomas 2001). In addition, women with PCOS frequently present with obesity and have an increased risk of developing the metabolic syndrome including type 2 diabetes mellitus and hypertension (Dunaif & Thomas 2001, Elting et al. 2001). High circulating androgens in PCOS have been thought to originate from increased androgen generation in ovaries and adrenals (Franks 1995). However, there is evidence for a significant contribution of peripheral steroidogenesis to androgen synthesis and activation in PCOS (Stewart et al. 1990, Fassnacht et al. 2003). Interestingly, serum androgens are positively associated with body mass index (BMI) not only in PCOS, but also in women with simple obesity (Taponen et al. 2003). This, together with a fall in luteinising hormone levels with increasing BMI (Holte et al. 1994), suggests that androgen synthesis may take place not only in adrenals and ovaries but also in adipose tissue.
Recent research focused on sex steroid conversion in human adipose tissue, specifically investigating the expression and activity of 17β-hydroxysteroid dehydrogenases (17β-HSDs), which represent a major switch regulating sex steroid activation and inactivation at the pre-receptor level (Adamski & Jakob 2001, Luu-The 2001). By employing RNA protection assays and Southern blot, Corbould et al.(1998) detected low mRNA expression levels of 17β-HSD isozymes 2 and 3, but not of 17β-HSD1, in human adipose tissue. In another study, Corbould et al.(2002) found that preadipocytes readily converted androstenedione to testosterone and that the ratio of 17β-HSD3 to aromatase expression in omental adipose tissue was positively correlated with BMI. However, in recent years 17β-HSD5 (AKR1C3) has been described as a major player in androgen interconversion (Dufort et al. 1999, Penning et al. 2000, Luu-The et al. 2001). Recently, Blouin et al.(2003) described mRNA expression of 17β-HSD5 (AKR1C3) and 3α-HSD type 3 (AKR1C2) in human female adipose tissue. They reported an increased inactivation of 5α-dihydrotestosterone by AKR1C2 in omental fat tissue from women with visceral obesity and suggested that local androgen inactivation is a predominant reaction in female abdominal tissue.
Here, we studied androgen generation in human female adipose tissue, particularly focusing on the site-specific expression of 17β-HSD5 (AKR1C3) in omental and subcutaneous fat, and investigated the impact of changes in BMI. In addition, we explored 17β-HSD5 expression and androgen conversion across adipocyte differentiation.
Subjects and Methods
Paired subcutaneous and omental biopsies Paired subcutaneous (sc) and omental (om) adipose tissue biopsies were obtained from 27 female patients undergoing elective abdominal hysterectomy for myoma or fibroma (median age 43 years, range 30–50 years; median BMI 27.8 kg/m2, range 19.7–39.2 kg/m2). All were premenopausal, non-diabetic, and none of them had been treated with glucocorticoids within the preceding twelve months or had been on any hormonal treatment, e.g. contraceptives. In 24 of the 27 subjects suffcient material for both activity and expression studies was obtained, while biopsy material in the remaining three subjects was only suffcient for expression analysis.
Subcutaneous biopsies (weight loss study) Six healthy women (median age 45 years, range 23–57 years; median BMI 35.2 kg/m2, range 30.4–38.3 kg/m2) with no significant past medical history were recruited for a weight loss study as previously reported (Tomlinson et al. 2004). A sc buttock biopsy was performed under local anaesthetic to obtain 1–2 g adipose tissue. After a median duration of 10 weeks of a weight loss program with a very low calorie diet (425 kcal/day) and a weight loss of more than 10% of the initial body weight, the patients returned to a normal diet for at least 1 week before they received a second sc buttock biopsy (Tomlinson et al. 2004).
All studies had the approval of the local research ethics committee and written informed consent was obtained in every case.
Tissue and primary cell culture
Samples of whole sc and om fat tissue were used for RNA extraction and for the isolation of preadipocytes 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, Berks, UK) in 1× Hank’s balanced salt solution (Life Technologies Inc., Paisley, Strathclyde, 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 then centrifuged again at 90 × g for 5 min. The pellet containing preadipocytes was removed, and cells were washed with DMEM/Nutrient Mixture F-12 (Life Technologies Inc.) containing 15% fetal calf serum (Life Technologies Inc.), 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 (Fig. 1A). Subsequently, preadipocyte cultures were either cultured towards confluence to study proliferated preadipocytes or subjected to differentiation stimuli to study the effects of maturation of stromal preadipocytes to adipocytes.
For preadipocyte proliferation cultures, human preadipocytes were maintained in DMEM/F-12 medium containing 15% fetal calf serum until confluence, which was achieved after 5 and 7 days of cell culture (Fig. 1B). Preadipocytes were then used for enzyme assays and RNA extraction.
For preadipocyte differentiation assays, human subcutaneous and omental preadipocytes were cultured for 14 days in media known to promote adipocyte differentiation (Hauner et al. 1987) following an established protocol (Bujalska et al. 1999) (Fig. 1C). On days 1 and 14, total RNA was extracted and differentiation of adipose stromal cells to adipocytes was confirmed by analysing lipoprotein lipase and glycerol-3-phosphate dehydrogenase expression, as described previously (Bujalska et al. 2002b). In addition, sc and om preadipocyte cultures were used for enzymatic activity assays on day 1 and day 14 of differentiation.
Paired sc and om preadipocyte proliferation cultures from 19 women were incubated in serum-free DMEM without phenol for 20 h with 40 nM androstenedione containing 100 000 c.p.m. [1,2,6,7-3H] androstenedione (specific activity 25 Ci/mmol; Perkin-Elmer, Cambridge, Cambs, UK).
Paired sc and om preadipocyte differentiation cultures from 5 women were incubated on day 1 and day 14 in serum-free DMEM without phenol for 20 h with 40 nM androstenedione containing 100 000 c.p.m. [1,2,6,7-3H] androstenedione or with 40 nM testosterone containing 100 000 c.p.m. [1,2,6,7-3H] testosterone (specific activity 78.5 Ci/mmol; Perkin-Elmer).
Steroids were extracted in 5 vols dichloromethane, concentrated by evaporation, and separated by thin layer chromatography (TLC) on silica gel/aluminium foil-coated TLC plates (Fluka Chemika, Buchs, Switzerland) with chloroform/ethylacetate (80:20) as the solvent system. Fractional conversion of steroids was quantified using a Bioscan 2000 Imager (LabLogic, Sheffeld, UK). Steroids were identified by referring to co-migration of unlabelled steroid standards (Sigma, UK) visualised by Liebermann-Burchard reagent as previously described (Quinkler et al. 1999). Cell protein content was determined with a modified Bradford method (Bio-Rad Laboratories, Munich, Germany), and enzyme activities were expressed as pmol or fmol product formed per mg protein per h. All incubations were carried out at least in triplicate.
RNA extraction and RT
Total RNA was extracted from whole fat tissue and preadipocyte cultures using a single step extraction method (Tri reagent, Sigma, UK (preadipocytes), or Genelute total mammalian RNA extraction kit, Sigma, UK (whole adipose tissue)). 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).
PCR for 17β-HSD isozymes
Qualitative PCR analysis for expression of 17β-HSD isozymes was carried out using cDNA generated from paired sc and om fat samples of 5 women, employing the following primers: 17β-HSD1 (HSD17B1), forward: 5′-TTTCA ATGACGTTTATTGCGCCAGCA-3′, reverse: 5′-TTG CTGTGGGCGAGGTATTGGTAGAA-3′; 17β-HSD2 (HSD17B2), forward: 5′-CTGGTGACAGGTGGTGA TTG-3′, reverse: 5′-TTATCTGCACTGGCTTCGTG-3′; 17β-HSD3 (HSD17B3), forward: 5′-TCCTGAACGCACCGGATGAAAT-3′, reverse: 5′-TACCTGACCTTG GTGTTGAGCTTCAG-3′; 17β-HSD4 (HSD17B4), forward: 5′-AGTTCTCTCTCTTTCTTGTTGGCTC TGGA-3′, reverse: 5′-GCGTCCTATTTCCTCAAA TACAAAGGTACTCT-3′; 17β-HSD5 (AKR1C3), forward: 5′-ACTTCATGCCTGTATTGGGATTTG-3′, reverse: 5′-CTGCCTGCGGTTGAAGTTTGACA-3′. 17β-HSD5 primers differ in two bases (forward primer) and three bases (reverse primer) respectively, from the AKR1C2 sequence (see underlined bases above) and are specific for 17β-HSD5 (AKR1C3) as previously shown by Bumke-Vogt et al.(2002). In addition to the primer sequences for 17β-HSD isozymes 1, 2, and 3 given above, we also performed specific PCR reactions with the 17β-HSD primers used by Corbould et al.(1998). Specific PCR reactions (20 μl) were carried out in 1 × PCR buffer (Promega) containing 50 mM KCl, 10 mM Tris–HCl (pH 9.0), and 0.1% Triton X-100, 2 mM MgCl2, 0.2 μM of each dNTP, 0.6 μM of forward and reverse primer and 1 U Taq DNA polymerase (Promega). Samples were amplified using initial denaturation at 94 °C (5 min) followed by 34 cycles at 94 °C (30 s), 60 °C (30 s), and 72 °C (30 s) with final elongation at 72 °C (7 min). After gel electrophoresis, PCR products of reactions with 17β-HSD5 primers were excised under UV light, extracted with QIAquick Gel Extraction Kit (Qiagen) and subjected to direct sequencing.
Total RNA was extracted from paired sc and om whole fat tissue as well as from preadipocyte proliferation cultures from five women. Biotin-labelled cRNA was reverse transcribed from 10 μg total RNA (pool of 2 μg total RNA from each cell preparation) according to the Affymetrix technical protocol (GeneChip Expression Analysis, Technical Manual, http://www.a3ymetrix.com/support/index.a3x). Integrity of the samples was tested on Affymetrix Test3 array and samples were accepted if their 3′/5′ratio of human housekeeping genes was lower than 3, referring to the house-keeping genes beta-actin (ACTB, Affymetrix probe HSAC07/X00351) and glyceraldehyde-3-phosphate dehydrogenase (GAPD, Affymetrix probe HUMGAPDH/M33197) were accepted. Fragmented cRNA (15 μg) was hybridised to each of the array of the Human Genome U133 set (Affymetrix, High Wycombe, UK). Affymetrix software (Data Mining Tool, Version 3.0) and Microarray Data Suite 5.0 (Affymetrix) were used for analysing microarray data. 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 probe set for 17β-HSD5 (209160_at) on the HG U133A chip includes 11 probes, all matching the published sequence. Due to the sequence difference between 17β-HSD5 and other 17β-HSD enzymes, none of the probes coding for 17β-HSD5 match the probe sets of 17β-HSD isozymes 1, 2, 3 or 4. The enzymes AKR1C1 (probe set 204151_x_at) and AKR1C2 (probe set 209699_x_at) have a high homology to 17β-HSD5 (AKR1C3); however only 2 and 3 probes, respectively, out of 11 probes matched the 17β-HSD5 sequence.
17β-HSD5 mRNA expression was analysed using an ABI Prism 7700 sequence detection system (Perkin–Elmer Applied Biosystems, Warrington, UK) that employs Taq–Man chemistry for highly accurate mRNA quantification as previously described (Bujalska et al. 2002a). Reactions were performed in 25-μl vol on 96-well plates, in buffer containing TaqMan Universal PCR Master Mix (Applied Biosystems, Foster City, CA, USA) and 25 ng cDNA template. 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 at 95 °C for 15 s and 60 °C for 1 min. Oligonucleotide primers and a Taqman probe for 17β-HSD5 were as follows: forward, GGGATCTCAACGAGACAAACG; reverse, AAAGGACTGGGTCCTCCAAGA; probe, TGG ACCCGAACTCCCCGGTG. Data were expressed as CT values (=cycle number at which logarithmic PCR plots cross a calculated threshold line) and were 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 ΔCT].
Data are expressed as means±s.d or s.e.m., as specified. Statistical analysis on real-time PCR data was performed on mean ΔCT values to exclude potential bias owing to averaging data that had been transformed through the equation 2−ΔCT. 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 coeffcient.
PCR for 17β-HSD isozymes
The conversion of androstenedione to testosterone and vice versa requires the presence of reductive and oxidative 17β-HSD activity. Thus, to identify the enzymes responsible for testosterone synthesis and inactivation in adipose tissue, we performed RT-PCR analysis employing specific primers for the amplification of human 17β-HSD isozymes. When analysing preadipocyte and whole adipose mRNA obtained from sc and om depots of five individual patients, abundant expression of 17β-HSD4 and 17β-HSD5 mRNA was detected (Fig. 2). The data suggest that 17β-HSD4 expression is higher in om than sc whole adipose tissue. 17β-HSD5 expression appears to be weakest in om preadipocytes. In whole fat and preadipocytes of both sc and om origin no relevant expression of 17β-HSD1, 2 or 3 was detected, except for a weak band for 17β-HSD1 and 2 in omental tissue of one subject (Fig. 2). Using the primers and methods reported by Corbould et al.(1998) we were not able to detect mRNA expression of 17β-HSD isozymes 1, 2 or 3, either in whole fat or in preadipocytes, despite a strong band in the positive control lane (data not shown).
Microarray analysis confirmed the absence of 17β-HSD1, 2 and 3 and the presence of 17β-HSD4 and 5 in preadipocytes and whole fat tissue of sc and om origin (Table 1). 17β-HSD5 microarray expression signals were higher in sc than in om fat (511 vs 73 in preadipocytes, P<0.001; 1525 vs 428 in whole fat, P<0.001), with an apparently higher abundance in whole fat than in preadipocytes of sc origin. By contrast, 17β-HSD4 expression signals did not differ significantly between sc and om fat, either in preadipocytes (263 vs 222) or in whole fat tissue (330 vs 303). The expression levels of AKR1C1 and AKR1C2 were very high (2800 and 2200 respectively) and exceeded those of 17β-HSD5 (AKR1C3) in whole fat tissue with no significant difference between sc or om tissue.
Enzyme assays in preadipocyte proliferation cultures
Enzyme activity studies in primary preadipocyte proliferation cultures from 19 patients revealed significant androgen generation, as assessed by the conversion of androstenedione to testosterone, with no apparent difference between sc and om preadipocyte cultures (0.83 ± 0.24 vs 0.86 ± 0.23 pmol/mg/h).
Quantitative PCR for site-specific expression of 17β-HSD5
For validation of the results on 17β-HSD5 expression in fat achieved by microarray analysis, we performed quantitative real-time PCR on cDNA generated from paired sc and om fat biopsies of 19 women. In keeping with the micorarray data, 17β-HSD5 mRNA expression levels were significantly higher in sc than in om fat, both in preadipocytes (P<0.01) and in whole fat (P=0.05) (Fig. 3). However, in both sc and om depots, 17β-HSD5 mRNA expression was significantly higher in whole fat tissue than in preadipocytes (P<0.001) (Fig. 3). This suggests a much higher abundance of 17β-HSD5 expression in mature adipocytes than in preadipocytes.
A potential site-specific role for 17β-HSD5, in the light of higher abundance of expression in sc than in om fat, was also supported by correlation of real-time 17β-HSD5 mRNA expression with BMI in the 19 patients studied (Fig. 4A and B). 17β-HSD5 mRNA expression in whole adipose tissue was significantly correlated with BMI in the sc compartment (r=0.51, P=0.027) but not in the om depot (r=0.11) (Fig. 4A). By contrast, expression of 17β-HSD5 in preadipocytes, which was considerably lower than in whole fat tissue, did not correlate with BMI, either in cells of om (r=0.078) or sc (r=0.014) origin (Fig 4B). There was no significant correlation between 17β-HSD5 mRNA expression and age in whole fat or preadipocytes.
Quantitative PCR for 17β-HSD5 expression in weight loss study
We performed quantitative real-time PCR on cDNA generated from subcutaneous buttock biopsies of six women who underwent a weight loss study. The mRNA expression of 17β-HSD5 was significantly reduced (P<0.05) in the biopsies taken after the weight loss compared with those taken before the weight loss (Fig. 5).
Quantitative PCR for 17β-HSD5 expression across differentiation
To further explore the expression of 17β-HSD5 across differentiation, primary preadipocyte cultures from sc and om fat of eight individuals were differentiated to mature adipocytes in adipogenic medium. 17β-HSD5 mRNA levels were assessed by real-time RT-PCR and again were found to be significantly higher (P<0.05) in sc than in om preadipocytes on day 1. Upon differentiation to adipocytes (day 1 compared with day 14), 17β-HSD5 mRNA levels in both om and sc cultures increased significantly (P<0.005) (Fig. 6A).
Enzyme assays in preadipocyte differentiation cultures
We studied the conversion of androstenedione to testosterone and vice versa across differentiation using primary preadipocyte differentiation cultures from sc and om fat of five women. On day 1 we could not detect any conversion of androstenedione to testosterone in sc and om preadipocytes (Fig. 6B). Omental, but not subcutaneous, preadipocytes readily converted testosterone to androstenedione on day 1. After differentiation to mature adipocytes on day 14, om and sc cultures exhibited similar conversion rates of androstenedione to testosterone, whereas the reverse conversion of testosterone to androstenedione was much stronger in the om than in cultures of sc origin (Fig 6B). When calculating the net conversion in preadipocyte differentiation cultures, om preadipocytes predominantly inactivated testosterone to androstenedione both before (day 1) and after 14 days of differentiation (Fig. 6C). By contrast, in differentiated cells (day 14) of sc origin net conversion revealed a significant gain in androgen activation, i.e. conversion of androstenedione to testosterone (Fig. 6C).
Conversion of androstenedione to testosterone in human adipose tissue may represent an intracrine source of androgen synthesis, possibly also contributing to the previously described positive correlation between circulating androgens and BMI (Taponen et al. 2003). It is important to recognize that this may follow a sex-specific pattern as this correlation has been found in women only (Taponen et al. 2003). Epidemiological studies have found that circulating levels of dehydroepiandrosterone (DHEA) sulphate, the sulphate ester of the crucial androgen precursor DHEA, are positively correlated with truncal fat in women (Williams et al. 1993) whilst there is an inverse correlation in men (Haffner et al. 1993). Thus androgen metabolism within male adipose tissue may differ from the pattern we report here in women.
Previous studies by Corbould et al.(1998, 2002) suggested that 17β-HSD3, which primarily shows gonadal expression, is also expressed in female adipose tissue. They concluded that 17β-HSD3 may be responsible for the conversion of androstenedione to testosterone, because no functional 17β-HSD1 enzyme was found in adipose tissue (Corbould et al. 1998). However, we were not able to detect any 17β-HSD3 mRNA expression in our fat tissue samples, even using the primers described by Corbould et al. and highly sensitive microarray techniques. There are some differences between the study presented by Corbould et al.(1998) and our study, in particular with regard to tissue handling and preadipocyte cultures. We have grown the preadipocytes in proliferation culture until confluence, i.e. for 5–7 days, before isolating RNA, whereas Corbould et al. used preadipocytes immediately after isolation and without further culturing. However, Corbould et al. did not investigate 17β-HSD5 expression in their fat samples, which is a likely candidate enzyme for the catalysis of the conversion of androstenedione to testosterone (Dufort et al. 1999, El-Alfy et al. 1999, Qiu et al. 2004). Very recently, 17β-HSD5 has been shown to be expressed in human adipose tissue (Blouin et al. 2003). We did not observe expression of 17β-HSD3, but did define abundant expression of 17β-HSD5. Thus we consider it highly likely that 17β-HSD5 represents the main reductive 17β-HSD isozyme responsible for conversion of androstenedione to testosterone in human fat. The only other 17β-HSD enzyme we found to be expressed in adipose tissue was 17β-HSD4, which only catalyses the oxidative reaction and therefore cannot convert androstenedione to testosterone (Adamski & Jakob 2001, Luu-The 2001, Mindnich et al. 2004).
According to our microarray data and quantitative real-time PCR results comparing whole fat with preadipocytes, we found 17β-HSD5 to be predominantly expressed in mature adipocytes, suggesting a role for this enzyme in adipocyte differentiation (Singh et al. 2003). This is further supported by the outcome of our preadipocyte differentiation assays, which showed a significant increase in 17β-HSD5 expression upon differentiation from stromal preadipocytes to mature adipocytes. However, it is not clear if the increased expression of 17β-HSD5 represents cause or consequence of ongoing differentiation. The increase in intracrine generation of androgens via the conversion of androstenedione to testosterone may influence differentiation (Singh et al. 2003). 17β-HSD5 also catalyses the conversion of DHEA to androstenediol, which has been shown to occur in female adipose tissue following incubation with radio-labelled DHEA (Schindler & Aymar 1975). Androstenediol has very similar structural characteristics to 5α-androstanediol, which recently has been shown to exhibit differentiating effects in the prostate via selective activation of the oestrogen receptor β(ERβ) (Weihua et al. 2002).
The results of the enzymatic activity assay in our preadipocyte differentiation cultures revealed predominant androgen inactivation, i.e. conversion of testosterone to androstenedione, which is likely to be be catalysed by 17β-HSD4, as we did not detect expression of the other oxidative 17β-HSD isozyme, 17β-HSD2. Blouin et al.(2003) previously suggested that omental adipose tissue represents an androgen inactivating rather than an androgen activating tissue. This is further supported by their finding of increased inactivation of 5α-dihydrotestosterone by 3α-HSD activity, possibly catalysed by AKR1C2 (Blouin et al. 2003).
In contrast to the predominant androgen inactivation in om cultures, calculation of net conversion in our preadipocyte differentiation cultures of sc origin revealed significant androgen activation, i.e. conversion of androstenedione to testosterone, suggesting that sc fat depots participate in systemic androgen production.
Our findings strongly suggest a site-specific role for 17β-HSD5. First, 17β-HSD5 expression was significantly higher in sc than in om adipose tissue. Secondly, we found a significant positive correlation of BMI with 17β-HSD5 expression in sc, but not in om fat in our female cohort. Conversely, we could show that 10% weight loss resulted in a significant decrease in 17β-HSD5 expression in sc fat. This lends some support to the view that 17β-HSD5 expression plays a site-specific role in female obesity. However, the weight loss data should be interpreted with some caution; in this study we have used gluteal sc adipose tissue, which may differ from abdominal sc adipose tissue with respect to both enzymatic expression patterns and differentiation behaviour.
Interestingly, our findings with regard to 17β-HSD5 expression exactly mirror previous observations on the expression of the androgen receptor (AR) in human adipose tissue, describing an increased expression in om compared with sc fat and a decrease in AR expression with differentiation of preadipocytes to mature adipocytes (Dieudonne et al. 1998). Higher conversion to more active androgen in subcutaneous adipose tissue may reflect a compensatory mechanism to counteract the lower AR expression. Our findings therefore highlight the importance of understanding site-specific regulation of adipogenesis by sex steroids.
In conclusion, we have shown that human female preadipocytes are capable of active androgen generation with the conversion of androstenedione to testosterone most likely catalysed by 17β-HSD5. We observed a site-and cell type-specific expression pattern for 17β-HSD5 with predominant expression in sc mature adipocytes, therefore suggesting a possible role for this enzyme in the regulation of adipocyte differentiation. We have shown that 17β-HSD5 expression increases across differentiation from stromal preadipocytes to mature adipocytes and that the interconversion of androstenedione and testosterone is shifted towards testosterone generation with ongoing differentiation of sc preadipocytes. In addition, we observed a positive correlation between BMI and sc 17β-HSD5 expression and a decrease in 17β-HSD5 expression with weight loss. This supports the hypothesis that human adipose tissue may represent an intracrine source of androgen synthesis, possibly also contributing to the previously described positive correlation between circulating androgens and BMI. Further studies are warranted to explore adipose tissue-specific metabolism of sex steroids in both males and females and to elucidate the clinical significance of these findings, bearing in mind the importance of regional fat distribution for cardiovascular complications of obesity.
Expression of 17β-HSD isozymes according to cDNA arrays with RNA extracted from preadipocytes and whole fat tissue of both subcutaneous (sc) and omental (om) origin. For each cell compartment the corresponding specific RNAs of 5 women were pooled
|Detection||Signal||Detection||Signal||Fold change sc vs om|
|Detection: qualitative measure indicating if the transcript is reliably detected (present, P), not detected (absent, A). Signal: quantitative measure of the relative abundance of a transcript. Fold change: change in expression level for a transcript between sc and om.|
|Whole fat tissue|
We would like to thank David M Smith, Carl Montague, and Kay Garnett (AstraZeneca, UK) for help with micro-array analysis. We also gratefully acknowledge the support of Dr Malcolm Mattrass and the Diabetes Research Fund, University Hospital Birmingham NHS Trust. We thank the consultants and theatre staffat the Birmingham Women’s Hospital, Birmingham, UK, for obtaining the fat biopsies. We thank all women who volunteered to participate in the study for their contribution.
This work was supported by the Deutsche Forschungs-gemeinschaft (Research Fellowship to M Q, and Senior Clinical Fellowship to W A), the Medical Research Council (Research Training Fellowship to J W T, and Senior Clinical Fellowship to P M S) and by a University of Birmingham Rowbotham Bequest Research Grant (to W A).
CorbouldAM Bawden MJ Lavranos TC Rodgers RJ & Judd SJ 2002 The effect of obesity on the ratio of type 3 17β-hydroxysteroid dehydrogenase mRNA to cytochrome P450 aromatase mRNA in subcutaneous abdominal and intra-abdominal adipose tissue of women. International Journal of Obesity and Related Metabolic Disorders26165–175.
El-AlfyM Luu-The V Huang X-F Berger L Labrie F & Pelletier G 1999 Localization of type 5 17β-hydroxysteroid dehydrogenase 3β-hydroxysteroid dehydrogenase and androgen receptor in the human prostate by in situ hybridization and immunocytochemistry. Endocrinology1401481–1491.
PenningTM Burczynski ME Jez JM Hung CF Lin HK Ma H Moore M Palackal N & Ratnam K 2000 Human 3α-hydroxysteroid dehydrogenase isoforms (AKR1C1-AKR1C4) of the aldo-keto reductase superfamily: functional plasticity and tissue distribution reveals roles in the inactivation and formation of male and female sex hormones. Biochemical Journal35167–77.
QuinklerM Johanssen S Grossmann C Bahr V Muller M Oelkers W & Diederich S 1999 Progesterone metabolism in the human kidney and inhibition of 11β-hydroxysteroid dehydrogenase type 2 by progesterone and its metabolites. Journal of Clinical Endocrinology and Metabolism844165–4171.
TaponenS Martikainen H Jarvelin MR Laitinen J Pouta A Hartikainen AL Sovio U McCarthy MI Franks S & Ruokonen A 2003 Hormonal profile of women with self-reported symptoms of oligomenorrhea and/or hirsutism: Northern Finland Birth Cohort 1966 Study. Journal of Clinical Endocrinology and Metabolism88141–147.