Introduction
The dramatic worldwide increase in obesity and its associated metabolic diseases has been widely attributed to an obesogenic environment created by our increasingly sedentary lifestyles and easy access to highly palatable, energy-dense foods. Although there is little doubt that poor diet and lack of exercise contribute to the obesity epidemic, recent studies have identified an estrogen endocrine disrupter chemical bisphenol A (BPA) that may act as an environmental obesogen and either directly or indirectly influence fat accrual.Reference Janesick and Blumberg 1 BPA, a monomer plasticizer, is used in the manufacture of common household goods including polycarbonate plastics (e.g. food and drink containers), paints and adhesivesReference Vandenberg, Hauser, Marcus, Olea and Welshons 37 . In rodents, maternal BPA exposure increases postnatal body weights (BWs) and growth rates with differing sex-specific effects. Some studies show greater susceptibility to BPA-increased adiposity in females, some in males and other studies show effects on both sexes.Reference Richter, Birnbaum and Farabollini 2 – Reference Susiarjo, Xin and Bansal 6 Critically, fetal exposures to BPA at levels equivalent to or below the established daily human safe dose (50 µg BPA/kg BW/day) not only increase BW and postnatal growth rate but also alter body composition in later life.Reference Richter, Birnbaum and Farabollini 2 , Reference Rubin and Soto 3 , Reference Alonso-Magdalena, Morimoto, Ripoll, Fuentes and Nadal 7 – Reference vom Saal, Nagel, Coe, Angle and Taylor 9 Energy (calorie) intake, energy expenditure and energy storage are the three key determinants of energy balance.Reference Hill, Wyatt and Peters 10 Increased adipose mass in offspring exposed to BPA during pregnancy may result from increased numbers of adipocyte cells (hyperplasia) secondary to enhanced differentiation of pre-adipocytes into mature adipocyte cells, as well as increased adipocyte cell size (hypertrophy) secondary to triglyceride storage.Reference Ailhaud, Grimaldi and Negrel 11 , Reference Hausman, DiGirolamo, Bartness, Hausman and Martin 12 Adipogenesis involves pre-adipocyte proliferation, differentiation and lipogenesis.Reference Morrison and Farmer 13 , Reference Rosen, Walkey, Puigserver and Spiegelman 14 This process requires the coordinated interaction of pre-adipocyte proliferative factors (PREF1, SOX9) and several adipogenic differentiation transcription factors, including members of the CCAAT/enhancer-binding family of proteins (C/EBPα, C/EBPβ and C/EBPδ),Reference Darlington, Ross and MacDougald 15 , Reference Lane, Lin, MacDougald and Vasseur-Cognet 16 which activate the peroxisome proliferator-activated receptor (PPARγ2).Reference Rosen and Spiegelman 17 The BPA is associated with increased expression of pro-adipogenic genes in vivo Reference Somm, Schwitzgebel and Toulotte 4 , Reference Miyawaki, Sakayama, Kato, Yamamoto and Masuno 18 and with accelerated terminal differentiation of 3T3-L1 cells in vitro. Reference Chamorro-Garcia, Kirchner and Li 19 , Reference Masuno, Iwanami, Kidani, Sakayama and Honda 20 However, no study to date has determined the effect of prenatal BPA on offspring primary pre-adipocyte proliferation and differentiation and the underlying mechanism.
We sought to confirm the effects of maternal BPA exposure during pregnancy and lactation on offspring BW, while examining effects on measures of adiposity. To more fully explore the mechanisms of BPA-mediated effects, we further utilized established models of newborn rat primary pre-adipocyte stem cells, exploring both proliferative (i.e. trophic) and differentiation effects of BPA.Reference Yee, Lee and Ross 21 We further explored putative signal factors which explain, in part, adipose responses and underlying epigenetic mechanisms mediated by BPA. The data reinforce the marked effects of BPA on adipogenesis and highlight the susceptibility of stem-cell populations during early life with long-term consequence on metabolic homeostasis.
Methods
BPA model
In vivo maternal BPA exposure
Studies were approved by the Animal Care Committee at the Los Angeles Biomedical Research Institute at Harbor-UCLA Medical Center and were in accordance with the American Accreditation Association of Laboratory Care. All animals were treated humanely and with regard for alleviation of suffering. Virgin Sprague–Dawley female rats (Charles River Laboratories, Hollister, CA, USA) were housed in an animal facility with controlled 12/12 h light/dark cycles, constant temperature and humidity conditions and ad libitum access to chow diet (Lab Diet 5001; Brentwood). The chow diet contains soy meal and as this was fed to both Controls and BPA-exposed animals, any comparative differences between the groups are likely due to the BPA exposure rather than the estrogenic activity of the phytoestrogens in the diet. To avoid potential BPA contamination, polypropylene cages and purified water in glass bottles were utilized. Female rats were randomly assigned to Control (n=5) or BPA (n=5) group. To reflect the most likely route of human exposure,Reference Carwile, Luu and Bassett 22 – Reference Le, Carlson, Chua and Belcher 25 dams were exposed to BPA via their drinking water. Control rats had access to purified drinking water, whereas the BPA group received purified drinking water containing BPA (5 mg/l; BPA Sigma-Aldrich, purity ≥99%, CAS no. 80-05-7) for 2 weeks before mating and throughout pregnancy and lactation (Table 1). Studies that administered BPA to pregnant rodents via drinking water, a concentration of 10 mg/l water (consumption of ~1.2 mg/kg BW/day)Reference Mendoza-Rodriguez, Garcia-Guzman and Baranda-Avila 26 yielded BPA tissue concentrations of 10–25 ng/g tissueReference Kabuto, Amakawa and Shishibori 27 , Reference Nakajima, Goldblum and Midoro-Horiuti 28 consistent with that of human samples.Reference Schonfelder, Wittfoht and Hopp 29 A dose five-fold higher (6 mg/kg BW/day) administered via gavage achieved significantly higher maternal plasma BPA levels,Reference Yoshida, Shimomoto and Katashima 30 whereas a water concentration of only 1 mg/l resulted in low maternal plasma-free BPA levels (0.84 ng/ml).Reference Patisaul, Sullivan and Radford 31 Our dose was selected based upon our confirmation (pilot study) of maternal and newborn serum levels within the lower range of demonstrated human levels with normal BPA exposure.
Table 1 In vivo maternal BPA exposure: drinking water
BPA, bisphenol A; DEXA, dual-energy X-ray absorptiometry.
In vivo maternal BPA exposure: Non-pregnant female rats at 9 weeks of age were allowed drinking water that was BPA-free (Control group) or contained BPA (BPA group). At 12 weeks of age, tail blood was obtained for BPA analysis and all females were mated and continued on same drinking water regimen throughout pregnancy and lactation. Parameters measured at various time points are indicated. At each offspring ages, n=5 males were studied from five separate litters.
Maternal BW and water intake were monitored daily, and dual-energy X-ray absorptiometry (DEXA) was undertaken at the end of lactation after which adipose tissue was collected for cell size analysis. Before mating, maternal blood was obtained via tail bleed and at the end of lactation via cardiac puncture in BPA-free tubes for BPA analysis. To avoid inducing maternal stress and fetal resorption,Reference Weinstock 32 blood samples were not collected during pregnancy, especially as maternal stress has been demonstrated to be an independent risk factor for offspring obesity.Reference Mueller and Bale 33 , Reference Hohwu, Li, Olsen, Sorensen and Obel 34 Free (unconjugated) BPA levels were measured using GC/MS (NMS Labs, PA, USA) with an assay sensitivity of 0.25 ng/ml. Insufficient plasma volume from maternal tail bleeds (before BPA administration) and newborns necessitated pooling of samples and hence only mean values are reported. Following BPA administration at the end of lactation, samples were analyzed individually for BPA levels.
Dams gave birth spontaneously and five litters per group (Control and BPA) were utilized for offspring studies. Litter size was standardized to eight per litter (four males and four females) to normalize rearing and all offspring were nursed by their respective mothers till 3 weeks of age. Following weaning, all offspring were given purified water and housed in polypropylene cages.
Offspring studies
DEXA scan
At 3 and 24 weeks of age, one male and one female offspring from one litter (n=5) underwent a non-invasive DEXA scanning using DXA system with a software program for small animal (QDR 4500A; Hologic, Bedford, MA, USA). An in vivo scan of whole body composition was obtained, including fat tissue mass, total mass and percent body fat.
Adipose tissue retrieval
Excess pups (beyond the four males and four females) were sacrificed at day 1 of life and subcutaneous (inguinal) adipose tissue from two pups of each gender from each litter was pooled according to sex (representing n=1) for analysis of tissue protein expression. A total of n=5 litters were studied. At 3 weeks of age, one male and one female from one litter (n=5) were euthanized and visceral (retroperitoneal) adipose tissue was collected for analysis of protein expression.
Blood pressure
At 6 weeks of age, measurements were undertaken in conscious animals using non-invasive tail-cuff sphygmomanometry (ML125 NIPB System; AD Instruments) method. Several cuff sizes are used depending on the weight of the animal. To circumvent the potential problem of restraint-induced stress, the animals were acclimatized for at least 1 week with placement in the restraint. One male and one female offspring from one litter (n=5) were studied.
In vitro BPA exposure
An additional four Control litters (n=4) were studied for in vitro effects of exogenous BPA on adipocyte cultures. From each litter, inguinal adipose tissue was dissected from each of four, 1-day-old Control males (four pooled samples representing n=1) for cell culture studies. Pre-adipocytes were cultured in Dulbecco’s modified Eagle’s medium (DMEM) medium in absence or presence of differentiation cocktail (see below) and treated with dimethyl sulfoxide (DMSO) (control) or BPA (1, 10, 20 µm) for 5 days. The total number studied was n=4 from four litters for adipocyte cultures.
Primary cultures
Primary adipocyte cell cultures were established as previously described.Reference Yee, Lee and Ross 21 Briefly, pooled adipose tissue from 1-day-old newborn males was minced and digested with collagenase type II (5000 U/g) in Krebs-Ringer solution. The pre-adipocyte cultures were resuspended in high glucose (450 mg/dl) DMEM (Invitrogen) with 10% fetal bovine serum (FBS) and 1% antibiotic-antimycotics (Invitrogen) and incubated at 37°C with 5% CO2. Cells were either seeded in 24-well plate (~1 × 104 cells/ml) for proliferation studies or flasks (~1 × 106 cells/ml) for protein expression and differentiation studies. After 24 h, undifferentiated pre-adipocytes were either treated with DMSO or BPA for 5 days. For differentiation studies, 24-h cultured pre-adipocytes were treated with dexamethasone (1 μm), methylisobutylxanthine (0.1 mm), insulin (10 μg/ml) in presence of DMSO or BPA for 5 days. At the end of exposure, all cells were harvested and protein extracted for analysis as described under Western Blot.
Analysis
Pre-adipocyte proliferation assay
Pre-adipocyte (complete medium) proliferation was determined as previously reportedReference Yee, Lee and Ross 21 using [3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide] colorimetric assay.Reference Mosmann 35
Adipocyte cell size
The cross-sectional area of adipocytes was determined as previously described.Reference Desai, Guang, Ferelli, Kallichanda and Lane 36 Briefly, formalin-fixed and paraffin-embedded adipose tissues was sectioned (5 µm), stained with hematoxylin–eosin, photomicrographs captured at 20× magnification and area of adipocytes (three images per section, and three sections per animal) determined using Image PRO software.
Lipid staining
Differentiated adipocytes were fixed with 4% paraformaldehyde, stained with 0.5% oil red O, mounted onto slides with Vectashield mounting medium with 4′,6-diamidino-2 phenylindole (Vector) and images (40× magnification) captured (Zeiss Axioskop 40 microscope with Axiocam HRc camera). For quantification, stained adipocytes were dried for 1 h at 37°C, incubated with a fixed volume of isopropanol for 20 min to elute the oil red O, and absorbance measured at 520 nm.
Western blot
Cell and tissue protein was extracted and Western Blotting performed as described previously.Reference Desai, Guang, Ferelli, Kallichanda and Lane 36 For adipose tissue, protein expression analysis included pre-adipocyte marker (PREF1, 55 kDa; Millipore), suppressor of pre-adipocyte differentiation (SOX9; 67 kDa; Millipore), adipogenic markers (PPARγ, 57 kDa, Cayman; C/EBPα, 42 kDa; Santa Cruz), lipogenic factor (SREBP1, 125 kDa; Santa Cruz) and epigenetic regulators DNA methyltransferase 3a (DNMT3a, 120 kDa; Santa Cruz) and lysine (K)-specific demethylase 1A (LSD1, 110 kDa; Cell Signaling).
Statistical analyses
In vivo responses between BPA and control offspring were compared by unpaired t-test. In vitro responses between BPA-exposed and untreated (DMSO) Control cells were compared by unpaired t-test or analysis of variance with Dunnett’s post-hoc test, as appropriate. P values ⩽0.05 were considered significant.
Results
Maternal BPA effects on maternal phenotype
Of note, the higher BPA levels (shown below) did not impact maternal weight gain and body fat. The maternal BW and water intake during pregnancy and lactation were comparable between BPA and Control dams (Fig. 1a). Further at the end of lactation, body fat and adipocyte cell size were similar in both groups (Fig. 1b).
Fig. 1 (a) Effects of maternal bisphenol A (BPA) exposure on maternal body weights, food intake and water intake: Maternal daily body weights and water intake during pregnancy and lactation of BPA (○) and Control (•) dams. Values are means±se of n=5 litters. (b) Effects of maternal BPA exposure on maternal body fat: Maternal body fat and retroperitoneal adipocyte cell size and images (scale bar=100 µm) at the end of lactation of BPA (□) and Control (■) dams. Values are means±se of n=5 litters.
Maternal BPA effects on plasma BPA levels
The pooled maternal plasma BPA level before BPA administration was 0.46 ng/ml. During the course of pregnancy, the amount of BPA consumed by dams via drinking water was 500–900 µg/kg/day and during lactation, it was higher (approximately 1500 µg/kg/day) due to increased water intake. The maternal BPA plasma levels at the end of lactation were higher in BPA as compared with Control dams (8.2±3.8 v. 0.42±0.04ng/ml). Similarly, newborns of BPA dams had higher plasma BPA levels (0.62 ng/ml) as compared with undetectable levels in newborns of Control dams.
Maternal BPA effects on offspring phenotype
Maternal BPA did not alter birth weights of either male or female newborns. However, by the end of the nursing period, 3-week-old BPA males exhibited significantly increased BW that persisted in 24-week-old adults (Fig. 2). Consistent with this, the percentage of body fat was increased in 3- and 24-week-old BPA males as compared with Controls (Fig. 3). Additionally, the systolic blood pressure was significantly higher in 6-week-old BPA than Control males (Fig. 4). In contrast to males, BWs, percentage body fat and systolic blood pressure of BPA females were comparable to Control females (Figs 2–4).
Fig. 2 Maternal bisphenol A (BPA) effects on offspring body weights: Body weights of 1-day-old newborns, and body weights of 3 and 24-week-old male and female offspring. Values are means±se of n=5 from 5 litters per group; *P<0.05 BPA (□) vs. Control (■).
Fig. 3 Maternal bisphenol A (BPA) effects on offspring adiposity: Percentage body fat of 3- and 24-week-old male and female offspring. Values are means±se of n=5 from five litters per group; *P<0.05 BPA (□) vs. Control (■).
Fig. 4 Maternal bisphenol A (BPA) effects on offspring blood pressure: Systolic blood pressure of 6–week-old male and female offspring. Values are means±se of n=5 from 5 litters per group; *P<0.05 BPA (□) vs. Control (■).
Maternal BPA effects on adipose tissue
As adiposity was not altered in BPA female, we only studied male offspring. At 1 day of age, BPA males had significantly increased protein expression of the adipogenic transcription factor PPARγ though not C/EBPα or the lipogenic factor SREBP1 (Fig. 5a). At 3 weeks of age, BPA males showed increased protein expression of C/EBPα though not PPARγ while the lipogenic factor SREBP1 was increased (Fig. 5b) in conjunction with evidence of hypertrophic adipocytes (Fig. 5c).
Fig. 5 Maternal bisphenol A (BPA) effects on offspring adipogenic and lipogenic factors: (a) inguinal adipose tissue protein expression of peroxisome proliferator-activated receptor (PPARγ) and C/EBPα in 1-day-old male newborns, (b) retroperitoneal adipose protein expression of PPARγ, C/EBPα and SREBP1 and (c) retroperitoneal adipocyte cell size and images (scale bar=100 µm) in 3-week-old male offspring. Values are means±se of n=5 of pooled adipose from each of five litters per group; *P<0.05 BPA (□) vs. Control (■).
Consistent with increased adipose tissue mass and hypertrophic adipocytes in BPA males, protein expression of CD68 (macrophage marker) and TNFα (pro-inflammatory cytokine) was significantly increased at 3 weeks of age (Fig. 6).
Fig. 6 Maternal bisphenol A (BPA) effects on adipose tissue inflammation in 3-week-old male offspring: Retroperitoneal adipose protein expression of CD68 and TNFα in 3-week-old male offspring. Values are means±se of n=5 of pooled adipose from each of five litters per group; *P<0.05 BPA (□) vs. Control (■).
In vitro BPA effects
Pre-adipocytes from Control offspring cultured in standard media with incremental doses of BPA showed a dose-dependent increase in proliferation (Fig. 7a and 7b), consistent with increased expression of the pre-adipocyte marker (PREF1), though expression of the anti-adipocyte differentiation transcription factor (SOX9) expression was decreased at 10 µm BPA (Fig. 7c). In differentiated adipocytes (grown in DMEM medium), BPA increased the number of adipocytes (1.8-fold), consistent with the increased protein expression of adipogenic transcription factors (PPARγ, C/EBPα), and expression of the pro-inflammatory cytokine TNFα (Fig. 8a). In addition, BPA-treated adipocytes showed increased lipid accumulation consistent with increased expression of SREBP1 (Fig. 8b).
Fig. 7 In vitro bisphenol A (BPA) effects on pre-adipocytes: Inguinal adipose tissue from 1-day-old Control newborns were cultured in Dulbecco’s modified Eagle’s medium (DMEM) (undifferentiated) media and treated with dimethyl sulfoxide (DMSO) (Control) or BPA (0, 1, 10, 20 μM) for 5 days. (a) Pre-adipocyte image, (b) proliferative index measured at 520 OD and (c) protein expression of PREF1 and Sox9 (results are shown for two doses due to loss of 20 μM BPA protein extract) with representative blot shown. Values are fold change (mean±se) of n=4 of pooled adipose from each of 4 litters; *P<0.05 BPA (□) vs. Control (■).
Fig. 8 In vitro bisphenol A (BPA) effects on differentiated adipocytes: Inguinal adipose tissue from 1-day-old Control newborns were cultured in Dulbecco’s modified Eagle’s medium (DMEM) media and pre-adipocytes were allowed to differentiate in presence of BPA for 5 days. (a) Protein expression of peroxisome proliferator-activated receptor γ (PPARγ) and C/EBPα with representative blot shown. (b) Adipocytes were fixed with 4% paraformaldehyde and stained for lipid (Oil Red O; red) and nucleus (4′,6-diamidino-2 phenylindole (DAPI); blue). Scale bar=100 µm. Protein expression of SREBP1 with representative blot shown and lipid content. Values are fold change (mean±se) of n=4 of pooled adipose from each of four litters;*P<0.05 BPA (□) vs. Control (■).
Epigenetic factors
BPA-treated adipocytes demonstrated a dose-dependent increase in DNMT3 though not LSD1 (Fig. 9).
Fig. 9 In vitro bisphenol A (BPA) effects on epigenetic factors: Inguinal pre-adipocytes from 1-day-old Control newborns were cultured in Dulbecco’s modified Eagle’s medium (DMEM) media with BPA (1, 10, 20 µm) for 5 days. Protein expression of DNA methyltransferase 3a (DNMT3a) and lysine (K)-specific demethylase 1A (LSD1) with representative blot shown. Values are fold change (mean±se) of pooled cells n=4 from each of 4 litters; *P<0.05 BPA (□) vs. Control (■).
Discussion
The present study determined the effects of BPA exposure in vivo and in vitro on adipogenesis, and the protein expression of regulatory transcription factors in early life. The data suggest that enhanced pre-adipocyte proliferation and differentiation in early life may contribute to the underlying mechanism of BPA-induced obesity.
A wide range of detectable BPA levels are reported in adults and children serum (0.2–20 ng/ml),Reference Vandenberg, Hauser, Marcus, Olea and Welshons 37 including breast milk (1.1 ng/ml), and maternal (0.2 to >10 ng/ml) and fetal/newborn serum (0.2–9.2 ng/ml). More significantly, the higher levels of measurable BPA in amniotic fluid (8.3–8.7 ng/ml) and placental tissues (1.0–104.9 ng/ml)Reference Kosarac, Kubwabo, Lalonde and Foster 38 , Reference Padmanabhan, Siefert and Ransom 39 imply a continued fetal exposure to BPA throughout development. Of note, higher BPA levels are seen in infants and children than in adultsReference Welshons, Nagel and vom Saal 40 and this has been associated with increased adiposity.Reference Harley, Aguilar and Chevrier 41 , Reference Rochester 42 The findings from animal studies corroborate the association of BPA exposure with adiposity. Specifically, studies show that it is lower (≤500 μg/kg/day) rather than higher dose (>5000 μg/kg/day) of maternal BPA that is effective in promoting offspring weight gain.Reference Somm, Schwitzgebel and Toulotte 4 , Reference Angle, Do and Ponzi 5 Sex-specific effects are seen with some studies demonstrating increased postnatal growth in both males and females at maternal doses between 2.4 and 500 μg/kg/day,Reference Richter, Birnbaum and Farabollini 2 , Reference van Esterik, Dolle and Lamoree 43 – Reference Bansal, Rashid and Xin 45 only in males at maternal dose of 500 μg/kg/dayReference Angle, Do and Ponzi 5 and only in females at maternal dose of 100 μg/kg/day.Reference Somm, Schwitzgebel and Toulotte 4 The reported effective maternal BPA dose of 500 μg/kg/day is compatible with our in vivo studies where effects are seen only in the male offspring and consistent with study by Bartolomei et al.Reference Susiarjo, Xin and Bansal 6 Further regardless of differing metabolism of BPA in rodents and humans,Reference Vandenberg, Maffini, Sonnenschein, Rubin and Soto 46 pharmacokinetic studies indicate that BPA exposure of approximately 400–500 μg/kg/day yield blood concentrations of the unconjugated, bioactive form of BPA that is similar to that reported in human blood.Reference Vandenberg, Colborn and Hayes 47
Previous studies on mice and rats indicate that endocrine disruptorsReference Melnick, Lucier and Wolfe 48 , Reference Diel, Schmidt and Vollmer 49 and estrogenReference Diel, Schmidt and Vollmer 49 have species and strain-dependent as well as organ-specific effects.Reference Diel, Schmidt and Vollmer 49 However, the overall sensitivity of various biological endpoints did not have pronounced differences between strains of rats.Reference Diel, Schmidt and Vollmer 49 With regard to adipogenesis, studies show that the two commonly used strains, Sprague–Dawley and Wistar rats, demonstrate increased adiposity in response to BPA exposure. Reference Somm, Schwitzgebel and Toulotte 50 – Reference Wei, Lin and Li 53 Specifically, pregnant Sprague–Dawley exposed to BPA cause increased adipogenesisReference Somm, Schwitzgebel and Toulotte 50 , metabolic pertubationsReference Tremblay-Franco, Cabaton and Canlet 54 and induced epigenetic transgenerational inheritance of obesity.Reference Manikkam, Tracey, Guerrero-Bosagna and Skinner 55
Interestingly unlike in the offspring, increased plasma BPA levels did not adversely impact maternal BW and body fat. This may be a result of continued exposure of developing fetus to BPA as stated above. There is also an implication that α-fetoprotein, which is secreted in high levels during development, may be involved. Normally, α-fetoprotein binds to estrogen and protects the developing fetus from estrogen effects. However, binding properties of some endocrine disruptors to α-fetoprotein may reduce this protection and expose the fetus to estrogen effects.Reference Nagel, vom Saal and Welshons 56 – Reference Milligan, Khan and Nash 59
The present study demonstrates that despite the normal birth weight, male newborns exposed to maternal BPA have a markedly increased postnatal growth rate and fat mass accumulation and hypertension. This elevated adipogenic potential is consistent with increased adipogenic (PPARγ, C/EBPα) and lipogenic (SREBP1) factors, which likely contribute to the hypertrophic adipocytes seen in BPA offspring. The underlying mechanism may be attributed to BPA-induced increased pre-adipocyte proliferation as demonstrated by in vitro BPA exposure studies. Specifically, pre-adipocytes exposed to in vitro BPA exhibit increased cell proliferation and increased PREF1, suggesting an increased pool of committed pre-adipocyte stem cells for potential adipocyte differentiation.Reference Hudak and Sul 60 PREF1 is known to suppress adipocyte differentiation by induction of SOX9.Reference Hudak and Sul 60 – Reference Wang and Sul 62 In our studies of in vitro BPA exposure, the failure of PREF1 to induce SOX9, as evident by the suppression of SOX9, indicates dysregulation of PREF1/SOX9 that may facilitate adipocyte differentiation under appropriate stimuli (e.g. increased insulin/glucocorticoid),Reference Ailhaud, Amri and Bardon 63 high carbohydrate or high fat diet.Reference Faust, Miller and Sclafani 64 – Reference Wabitsch 66 Indeed, when pre-adipocytes are allowed to differentiate in presence of BPA, there is increased expression of adipogenic and lipogenic transcription factors together with more lipid-filled adipocytes. Collectively, this results in an inflammatory response as evident by increased expression of macrophage marker (CD68) and pro-inflammatory cytokine (TNFα). These changes are similar to the effects seen with in vivo maternal BPA exposure and consistent with previously reported ex vivo studies of 3T3-L1 pre-adipocyte cell lines,Reference Chamorro-Garcia, Kirchner and Li 19 , Reference Masuno, Iwanami, Kidani, Sakayama and Honda 20 , Reference Masuno, Kidani and Sekiya 67 – Reference Sargis, Johnson, Choudhury and Brady 69 as well as murine and humanReference Boucher, Boudreau and Atlas 70 , Reference Valentino, D'Esposito and Passaretti 71 primary stem cells. Furthermore in the present study, the early increased PPARγ and later increased C/EBPα and SREBP1 suggest the specific influence of BPA during perinatal and postnatal exposures. It is known that although PPARγ and C/EBPα positively regulate each other’s expression and cooperate to promote adipogenesis,Reference Lee and Ge 72 C/EBPα is required for lipogenesis.Reference Bauer, Sasaki and Cohen 73
The mechanism underlying BPA-induced enhanced adipocyte proliferation and differentiation may involve epigenetic modifications via DNA and/or histone methylation,Reference Bastos, Kamstra and Cenijn 74 , Reference Kundakovic and Champagne 75 particularly of genes such as PPARγReference Fujiki, Kano, Shiota and Murata 77 and C/EBPα. Methylase and demethylase enzymes that have been implicated in determining stem-cell proliferation (self-renewal) and differentiation include DNMT (DNA methyltransferase) and LSD1 (lysine (K)-specific histone demethylase).Reference Adamo, Sese and Boue 76 For example, treatment of 3T3-L1 pre-adipocytes with an inhibitor of DNA methylation (5′-aza-cytideine) or knockdown of LSD1 decreased proliferation and adipocyte differentiation, resulting in downregulation of PPARγ.Reference Fujiki, Kano, Shiota and Murata 77 – Reference Zych, Stimamiglio and Senegaglia 79 This is consistent with our findings on in vitro BPA exposure, that is, increased pre-adipocyte proliferation with increased protein expression of DNMT3a. The lack of dose–response seen in the case of C/EBPα is presently unknown. However, it may involve BPA- or direct transcription factor-mediated chromatin interaction and accessibility.Reference Bastos, Kamstra and Cenijn 74 , Reference Hervouet, Vallette and Cartron 80 Although the current data suggest BPA alters transcriptional regulation of genes involved in proliferation, the role of DNA methylation in pre-adipocytes remains to be established and further studies of site-specific epigenetic modification of specific genes are required to fully elucidate BPA-induced changes in adipogenesis.
Although the sex-specific effects of BPA are well documented including the differential susceptibility of males and females to different doses of BPAReference Vandenberg, Colborn and Hayes 47 , Reference Strakovsky, Wang and Engeseth 81 – Reference Rubin, Paranjpe and DaFonte 85 , the underlying mechanism remains unclear.Reference Lejonklou, Dunder and Bladin 82 The plausible explanation may involve sex hormones, genomic and non-genomic pathway involving nuclear estrogen receptors, differing developmental pattern and/or epigenetic influence.Reference Vandenberg, Colborn and Hayes 47 , Reference Kundakovic, Gudsnuk and Franks 83 , Reference Susiarjo, Sasson, Mesaros and Bartolomei 86
Conclusion
Our data (Table 2) confirm that primary adipose progenitor cells are vulnerable to endocrine disruption by BPA resulting in altered proliferation and differentiation in early life independent of systemic influences. Enhanced adipose proliferation and differentiation indicate the potential for maternal/fetal BPA exposure to program an increased risk of offspring obesity and consequent metabolic abnormalities.Reference Somm, Schwitzgebel and Toulotte 4 , Reference Miyawaki, Sakayama, Kato, Yamamoto and Masuno 18 , Reference Ding, Fan and Zhao 87 , Reference Perreault, McCurdy and Kerege 88
Table 2 Summary data of in vivo exposure to maternal BPA
↑, increased; ↔, no change; –, not studied; PPARγ, peroxisome proliferator-activated receptor γ.
Changes at various ages are summarized for male offspring exposed to maternal bisphenol A.
Acknowledgments
The authors thank Stacy Behare for animal assistance.
Financial Support
This work was supported by the National Institute of Environmental Health Sciences (R21ES023112-01; M.D., M.G.R.), National Center for Advancing Translational Sciences UCLA- CTSI (Grant U11TR000124; M.D.), National Institute on Minority Health and Health Disparities (5U54MD007598-06 (M.G.F.) and Flora Foundation (M.D., M.G.R.).
Conflicts of Interest
None.
Ethical Standards
As stated in the Methods, all studies were approved by the Animal Research Committee of the Los Angles Biomedical Research Institute at Harbor-UCLA Medical Center and were conducted in strict accordance with guidelines provided by the American Accreditation Association of Laboratory Care and the Public Health Service Policy on Humane Care and Use of Laboratory Animals and conform to the principles and regulations as described in the Editorial by Grundy.Reference Grundy 89 All animals were treated humanely and with regard for alleviation of suffering. Virgin Sprague–Dawley female rats (Charles River Laboratories, Hollister, CA) were housed in an animal facility with controlled 12/12-hour light/dark cycles, constant temperature and humidity conditions and ad libitum access to chow diet (Lab Diet 5001; Brentwood, Missouri) and water.
