PPARγ Regulates Trophoblast Proliferation and Promotes Labyrinthine Trilineage Differentiation

Background Abnormal trophoblast differentiation and function is the basis of many placenta-based pregnancy disorders, including pre-eclampsia and fetal growth restriction. PPARγ, a ligand-activated nuclear receptor, plays essential roles in placental development; null murine embryos die at midgestation due to abnormalities in all placental layers, in particular, small labyrinth and expanded giant cell layer. Previous studies have focused mostly on the role of PPARγ in trophoblast invasion. Based on the previously reported role of PPARγ in preadipocyte differentiation, we hypothesized that PPARγ also plays a pivotal role in trophoblast differentiation. To test this hypothesis, we report derivation of wild-type and PPARγ-null trophoblast stem (TS) cells. Methodology/Principal Findings PPARγ-null TS cells showed defects in both proliferation and differentiation, specifically into labyrinthine trophoblast. Detailed marker analysis and functional studies revealed reduced differentiation of all three labyrinthine lineages, and enhanced giant cell differentiation, particularly the invasive subtypes. In addition, rosiglitazone, a specific PPARγ agonist, reduced giant cell differentiation, while inducing Gcm1, a key regulator in labyrinth. Finally, reintroducing PPARγ into null TS cells, using an adenovirus, normalized invasion and partially reversed defective labyrinthine differentiation, as assessed both by morphology and marker analysis. Conclusions/Significance In addition to regulating trophoblast invasion, PPARγ plays a predominant role in differentiation of labyrinthine trophoblast lineages, which, along with fetal endothelium, form the vascular exchange interface with maternal blood. Elucidating cellular and molecular mechanisms mediating PPARγ action will help determine if modulating PPARγ activity, for which clinical pharmacologic agonists already exist, might modify the course of pregnancy disorders associated with placental dysfunction.


Introduction
Placental dysfunction, caused by abnormalities in trophoblast differentiation and function, is associated with many pregnancy disorders, including intrauterine growth restriction and preeclampsia [reviewed in 1]. Peroxisome proliferator-activated receptors (PPARs), which are transcription factors and members of the ligand-activated nuclear hormone receptor superfamily, play major roles in diverse aspects of energy metabolism, inflammation, and development [2][3][4][5]. Following ligand binding, PPARs form heterodimers with retinoid X receptors (RXRs), and bind to PPAR-response elements (PPREs) of target genes to activate transcription [2]. PPARc has received much attention with regard to its role in adipogenesis and energy metabolism, including its highly efficacious synthetic ligands, the thiazolidinediones, which are routinely used clinically for treatment of type 2 diabetes [3,4]. However, genetically null mice revealed a surprising role for PPARc: namely its essential role in placental development [6,7]. Since this discovery, additional studies have focused primarily on the effect of PPARc ligands in human trophoblast, revealing effects on both syncytiotrophoblast function [8] and trophoblast invasion [9]. Further studies in mice have identified some transcriptional targets of PPARc in trophoblast [10] and shown that treatment of pregnant mothers with PPARc agonists led to fetal and placental growth restriction in a PPARc-dependent manner [11]. However, specific cellular and molecular mechanisms by which this important transcription factor exerts its predominate role in trophoblast differentiation and placental development have not been elucidated.
Given the essential role of PPARc in adipocyte differentiation [12,13], and the abnormalities observed in specific layers of PPARc-null placentas, we hypothesized that PPARc also regulates critical aspects of normal trophoblast differentiation. In order to test this hypothesis, we generated both wild-type (WT) and PPARc-null trophoblast stem (TS) cells, from E3.5 murine blastocysts using previously published methods [14]. Murine TS cells are an invaluable tool for characterizing the role of specific genes in trophoblast differentiation and function [15][16][17]. In this study, we characterized cellular properties and the differentiation capacity of PPARc-null TS cells and found a predominate role for this protein in both trophoblast proliferation and labyrinthine trophoblast differentiation.

Results
PPARc 2/2 TS Cells Grow Slower than Their Wild-Type Counterparts PPARc +/+ and PPARc 2/2 TS cells were generated from E3.5 blastocysts from two heterozygous matings. Two PPARc +/+ , one PPARc +/2 , and three PPARc 2/2 TS cell lines were obtained from 14 explanted blastocysts (43% overall yield). Both PPARc +/+ and two PPARc 2/2 TS cell lines were chosen for further study. Both wild-type (WT) lines were analyzed in most experiments and did not show statistically significant differences. Results from one WT line are shown for simplicity. TS cell genotypes, determined by PCR of genomic DNA, were confirmed at the phenotypic expression level using RT-PCR of total cell RNA with primers surrounding exon 3, which is deleted in the null allele ( Figure 1A), and by Western blot ( Figure 1D). Both WT and null lines expressed markers of trophectoderm (Cdx-2, Eomes, Errb; see Table 1 for definitions of gene abbreviations), and lacked inner cell mass (ICM)/embryonic stem (ES) cell markers (Oct3/4) ( Figure 1B, and data not shown). In WT TS cell cultures, PPARc mRNA and protein were detected at very low levels in the undifferentiated state, but were strongly upregulated upon differentiation, with maximal expression ob-served at day 4 ( Figures 1C and 1D). In contrast PPARc protein was not detected in PPARc 2/2 TS cell cultures at any stage of differentiation ( Figure 1D). In addition, while levels of PPARa were comparable between WT and null TS cells, PPARb/d was upregulated (1.5-2.5 fold by densitometry, normalized to bactin expression) in differentiated null cells compared to PPARc +/+ cells ( Figure 1D); these changes in PPARa and PPARb/d protein levels were reflective of changes in their respective mRNA levels (data not shown).
Morphologically, both WT and null cells formed typical undifferentiated colonies when grown in TSMFH media (standard TS cell growth media; see Materials and Methods), and differentiated to form trophoblast giant cells (TGC's) when switched to TS media (differentiation media: withdrawal of conditioned media, FGF4 and heparin) (see below). However, both PPARc 2/2 lines in TSMFH media grew more slowly (1.5-3-fold) than WT TS cells (Figure 2A) such that expansion of PPARc 2/2 , but not PPARc +/+ cultures was similar under growth and differentiation conditions ( Figure 2B). To examine stem cells more directly, and to assess whether slow overall growth in PPARc 2/2 cultures was due to accumulation and carry-over of post-mitotic giant cells during passage, we lightly trypsinized both PPARc +/+ and PPARc 2/2 cultures (selectively releasing undifferentiated cells), plated the resultant nonadherent, stem cell enriched populations at equivalent densities (10 5 cells per well in a 6-well plate) and assessed growth and morphologic differentiation in TSMFH media over a 3-day period ( Figure 3). The ratio of giant cells to undifferentiated (stem) cells was initially similar in PPARc +/+ and PPARc 2/2 cultures, but increased significantly over time only in the latter ( Figure 3B). This indicates that absence of PPARc leads to premature differentiation of TS cells into, and not carry-over of, giant cells even in the presence of MEF-conditioned medium, FGF4 and heparin. Differentiation of syncytiotrophoblast under these conditions was negligible in both PPARc +/+ and PPARc 2/2 cultures, with no significant difference between the two cell types ( Figure 3C).

PPARc 2/2 TS Cells Exhibit Altered Trophoblast Lineage Differentiation In Vitro
We differentiated both WT and null cells for 7-days and assessed labyrinthine (syncytiotrophoblastic) and invasive trophoblast/giant cell differentiation both by morphology and qRT-PCR. Morphologic differentiation was assessed by double staining with phalloidin and DAPI to distinguish giant cells, with one or two large nuclei (each at least double the size of a stem cell nucleus) from multi-nucleated syncytiotrophoblast (three or more nuclei each equivalent in size to stem cell nuclei). Morphologic analyses revealed that syncytiotrophoblasts were significantly less abundant in differentiated PPARc 2/2 than in PPARc +/+ TS cells (4.760.7% vs. 7.361.0%, respectively, P = 0.023) ( Figure 4).
In preliminary experiments, differentiating WT and PPARc 2/2 cells were analyzed daily for 7 days by qRT-PCR to determine the time of maximal expression of each lineage marker. The results were similar to previously published data, indicating maximal expression of Gcm1 at day 1, spongiotrophoblast and labyrinthine markers at day 4, and giant cell markers and muc1 at day 7 of differentiation [10,17,18 , and data not shown]. For the remaining experiments, these time points were used to determine differences in marker expression, although additional days of differentiation were also frequently examined to avoid misinterpretation due to minor alteration in expression kinetics. Compared to PPARc +/+

PPARc 2/2 TS Cells Differentiate Preferentially into Invasive Subtypes of Trophoblast Giant Cells
Recently, Simmons et al. [20] characterized diverse subtypes of trophoblast giant cells (TGC) based on expression of specific markers, including Pl-I, Pl-II, Plf, and Ctsq, and showed that formation of these subtypes by TS cells is differentially regulated in vitro. Specifically, Ctsq is a marker of sinusoidal TGC, likely derived from chorion-proximal Tpbpa 2 precursors in ectoplacental cone (EPC) and participating in labyrinth formation, while Plf is present in canal-type, spiral arteryassociated, and parietal TGC, likely derived primarily from chorion-distal Tpbpa + EPC [19,20]. In addition to upregulation of Pl-I and Pl-II (30-fold and 3.5-fold, respectively, Figure 5), we also found that Plf is upregulated ,4-fold while Ctsq is downregulated ,3-fold in differentiated PPARc 2/2 compared to WT TS cells ( Figure 7A). This suggests that PPARc 2/2 TS cells preferentially differentiate to invasive trophoblast sub-types, including canal-type and parietal TGC, and do not readily form sinusoidal TGC, an important constituent of chorion and fully differentiated labyrinth [19]. To determine whether the functional capacity of PPARc 2/2 TS cells mirrors their altered expression profile of TGC genes, we compared the behavior of PPARc +/+ and PPARc 2/2 TS cells, differentiated for 7 days, in a Matrigel invasion assay. As previously described [21], the majority of cells invading Matrigel were morphologically consistent with TGC (arrowheads in Figure 7B). In addition, although similar numbers of WT and PPARc 2/2 differentiated TGC were applied to the membrane (see Figure 4; ,92,700 TGC in wild-type vs. ,95,200 TGC in null cells after 7 days of differentiation), over two-fold more PPARc 2/2 TGC invaded to the far side of the Matrigel membrane compared to their WT counterparts ( Figure 7C), consistent with an enhanced invasive capacity of individual TGC in the absence of PPARc.

Rosiglitazone Inhibits Trophoblast Giant-Cell Differentiation and Invasion in a PPARc-Dependent Manner
To determine the effects of pharmacologic activation of PPARc on trophoblast, we tested the ability of rosiglitazone, a specific PPARc agonist, to alter TS cell proliferation and differentiation in vitro. While TS cell proliferation was not altered (data not shown), expression of differentiation markers (Figures 8  and 9) was differentially regulated by treatment of PPARc +/+ TS cells, but not PPARc 2/2 TS cells, with 1 mM rosiglitazone. Specifically, while rosiglitazone decreased expression of the spongiotrophoblast marker Tpbpa (2.2 fold, P = 0.0287) and giant cell markers Pl-I (5 fold, P = 0.0025) and Pl-II (11 fold, P = 0.0183;), the labyrinthine markers syncytin A and Tfeb were not affected (Figures 8 and 9). However, muc1, a known direct transcriptional target of PPARc in trophoblast [10], was increased 16 fold (P = 0.0089) and Gcm1 was increased 1.2 fold (P = 0.03) by rosiglitazone ( Figure 9). The effects of rosiglitazone were PPARc-dependent, as all were absent in PPARc 2/2 TS cells (Figures 8 and 9). In addition, downregulation of giant cell markers was not accompanied by changes in the morphologic assessment of differentiation, as the percentage of giant cells and syncytiotrophoblast remained the same with or without rosiglitazone (data not shown). Finally, rosiglitazone decreased invasion of differentiated TS cells by 36% in a PPARc-dependent manner ( Figure 7C; P = 0.0342).

Introducing PPARc in Genetically Deficient TS Cells Rescues Syncytiotrophoblastic and Labyrinthine Differentiation
Since PPARc-null TS cells preferentially and prematurely differentiate into TGC at the expense of labyrinthine cell types, and since PPARc activation by rosiglitazone downregulates markers of TGC while upregulating some labyrinthine genes, we hypothesized that PPARc promotes labyrinthine/ syncytiotrophoblastic differentiation. To directly test this hypothesis, we generated an adenoviral construct expressing a murine PPARc1-HA fusion protein and co-expressing GFP from an internal ribosome entry site. Western blot verified formation of an appropriately-sized HA-tagged protein (data not shown). PPARc-null TS cells were infected with the highly purified adenovirus; virus expressing only the backbone GFP construct, without the PPARc1 gene, was used as control. Both GFP and HA-PPARc1 were detectable in TS cell cultures seven days after infection, corresponding to day 6 of differentiation.
PPARc-expressing virus slowed growth of PPARc-null TS cells in TSMFH medium ( Figure 10A). Also, reintroduction of PPARc into null cells decreased Matrigel invasion by 44% ( Figure 10B; P = 0.0487). When induced to differentiate, PPARc-virus-infected null cells formed almost twice as many multi-nucleated syncytiotrophoblast, as control virus-infected cells ( Figure 10C). In addition, by qRT-PCR analysis, Gcm1, synA, and muc1 expression was induced 1.3-1.5 fold in differentiated PPARc-virus-infected null TS cells compared to control virus-infected cells (Figure 11, P,0.05 for each). In contrast, Tfeb and all giant cell markers were unaffected ( Figure 11, and data not shown). Both WT and PPARc-null TS cells when differentiated in media containing DMSO, the carrier used for rosiglitazone, routinely showed more syncytialized cells than did TS cells cultured without DMSO (compare Figure 10, with DMSO, to Figure 4, without DMSO). The additional presence of PPARc increased percentage of syncytiotrophoblast well above and beyond this DMSO effect; however, rosiglitazone did not further alter differentiation of PPARc-virus-infected null TS cells as assessed either by morphology or by qRT-PCR ( Figure 10C, and data not shown).

Discussion
PPARc is expressed at high levels in all placental layers of the mouse, starting as early as E8.5 (8.5 days post-coitum) [6 ; Milstone et al., manuscript in preparation]; however, its role in murine trophoblast differentiation has not been extensively studied. PPARc-null mice die at midgestation (E10.5-E11.5) due to abnormalities of the placenta, characterized by small labyrinth, reduced spongiotrophoblast, and expanded giant cell layers [6,7]. To better understand the role of PPARc in placental development and function, we generated both wild-type and PPARc-null trophoblast stem (TS) cells and characterized cellular properties in vitro. We determined that PPARc-null TS cells show defective differentiation of all three lineages contributing to labyrinth [19], with decreased expression of Ctsq (marking sinusoidal TGC after E12.5), syncytin A (marking syncytiotrophoblast layer I/SynT-I), and Gcm1 (marking syncytiotrophoblast layer II/SynT-II) (Figures 6 and  7A). Instead of forming labyrinth, PPARc-null TS cells differentiate prematurely and preferentially into trophoblast giant cells. These observations are consistent with the effects of PPARc genetic deficiency on placental development in vivo, particularly the extremely small labyrinth [6,7], and suggest that PPARc contributes directly to both maintenance of Muc1 and PGAR (PPARc angiopoietin related/FIAF/ ANGPTL4) are the only known direct transcriptional targets of PPARc in placenta [10,22]. Our data show that Gcm1 mRNA expression is induced both by rosiglitazone treatment of WT (but not null) TS cells (1.2 fold increase, Figure 9) and by reintroduction of PPARc into null TS cells (1.5 fold increase, Figure 11). This induction may represent direct transcriptional regulation by PPARc, as the proximal Gcm1 promoter contains at least one putative ''atypical'' PPARc-response element (PPRE) [23,24 , and data not shown]. Luciferase reporter gene assays are currently in progress to test PPARc1 regulation of transcriptional initiation at the proximal murine Gcm1 promoter [23].
The effect of rosiglitazone on downregulation of spongiotrophoblast/giant cell markers is consistent with previously published in vivo data, showing that treatment of pregnant mice with rosiglitazone downregulates placental Tpbpa expression and reduces the spongiotrophoblast compartment [11]. We also observed that rosiglitazone upregulates differentiated TS cell expression of muc1 (Figure 9), a known direct transcriptional target of PPARc in labyrinthine trophoblast in vivo [10]. However, while it induced Gcm1 expression, rosiglitazone did not alter the additional labyrinthine markers syncytin A or Tfeb, or cause morphologic evidence of syncytiotrophoblastic differentiation in our experiments (Figures 9, 10C, and data not shown). This induction of some but not all mediators and consequences of syncytiotrophoblast formation and labyrinthine differentiation, processes we hypothesize are PPARc-dependent, could reflect near-maximal PPARc activation by mediators already present in differentiating TS cell cultures and/or different roles of PPARc in formation of SynT-I and SynT-II cells. In addition, although wild type TS cells contribute robustly to all lineages in vivo, they show limited (only 5-10%) formation of labyrinthine cell types under the ''standard'' in vitro differentiation conditions [14,18 , and this paper], [ Figure 4]. In contrast, HIF1b/ARNT-deficient TS cells express high levels of Gcm1 and preferentially differentiate to syncytiotrophoblast in vitro [17]. Interestingly, differentiated HIF1b/ARNT 2/2 TS cells also contain high levels of acetylated histone-4, and inhibition of histone deacetylase in wild type TS cells phenocopies the HIF1b/ ARNT 2/2 mutation [17]. These data suggest that the Gcm1 promoter may be less accessible to transacting factors in cultured TS cells, due to a condensed chromatin structure, thereby limiting induction by potential transcriptional regulators, such as PPARc. Epigenetic effects may thus play a prominent role in PPARc-dependent trophoblast differentiation and placental development in vivo but may be incompletely recapitulated by TS cells differentiating in vitro. Differentiated PPARc 2/2 TS cells showed a 36% decrease in syncytiotrophoblast, assessed morphologically, compared to WT cells (4.7% compared to 7.3%; Figure 4). Syncytiotrophoblast differentiation was also restored (increased 84%) in PPARc 2/2 TS cells by reintroducing PPARc using adenovirus, and was not further altered by rosiglitazone ( Figure 10). However, restoring PPARc expression in null cells only modestly altered labyrinthine marker expression ( Figure 11) and did not alter giant cell markers (data not shown). This apparent quantitative discrepancy between morphologic and molecular differentiation may reflect, in part, that only 20-50% of PPARc 2/2 TS cells were infected by adenovirus (determined by GFP fluorescence), even at the highest MOI, and only 50% of these infected cells also showed nuclear localization of exogenous PPARc (determined by immunofluorescent detection of HA tag) ( Figure 10C and data not shown). Because syncytiotrophoblast cells were reported as a percent of the cells showing nuclear localized exogenous PPARc, while mRNA was (necessarily) quantified in the entire culture, the expression data underestimates the functional effects of exogenous PPARc, and the magnitude of the discrepancy is proportional to the fraction of uninfected cells. In fact, if all infected cells were counted (those expressing HA-PPARc in the nucleus, the cytoplasm or both), the percentage of syncytiotrophoblast increased from 8.3% to 11.5% (38% increase), not to 15.3% (84% increase). However, assessed either way, the effect of exogenous PPARc on syncytiotrophoblast differentiation remained statistically significant (P,0.05). In addition, nuclear HA-PPARc was present in all syncytiotrophoblast differentiated from PPARc 2/2 TS cells rescued by Ad-HA-PPARc-GFP. This suggests that PPARc supports syncytiotrophoblast differentiation by regulating activity of a nuclear/transcription factor(s). This could include upregulation of a factor inducing labyrinthine differentiation (such as Gcm1) or inhibition of factors favoring the spongiotrophoblast/giant cell lineage. Finally, enhanced giant cell differentiation of PPARc 2/2 TS cells may also, in part reflect unopposed action, and perhaps upregulation of PPARb/d (see Figure 1D), which is known to induce giant cell differentiation both in vivo and in vitro [25].
PPARc 2/2 TS cells grew more slowly and showed enhanced TGC, but not syncytiotrophoblast differentiation in TSMFH medium relative to their WT counterparts. This suggests that PPARc 2/2 TSC preferentially differentiate to TGC even under stem cell conditions (in TSMFH media). Unlike its effect on syncytiotrophoblast differentiation, reintroducing PPARc into null cells exacerbated, rather than reversed, the growth abnormality ( Figure 10A). This divergent behavior might reflect differential sensitivity of growth and differentiation to PPARc activity. CMV promoter-driven expression from Ad-HA-PPARc-GFP was 12-fold higher than endogenous PPARc levels observed in WT TS cell cultures, and was equivalent to levels normally achieved only after 4 days of differentiation ( Figure 1C and data not shown). This suggests that low level PPARc expression stabilizes the TS cell undifferentiated state under growth conditions, while high level expression favors syncytiotrophoblast formation when differentiation conditions predom-inate ( Figure 12); artificially imposing high level expression in the absence of differentiation conditions may suppress TS cell proliferation directly.
In contrast to these defects in labyrinthine trophoblast development, elevated expression of Plf, Pl-I, and Pl-II, suggests enhanced differentiation of PPARc 2/2 TS cells into parietal, canal and spiral artery associated TGCs [19], each of which are involved in establishing the maternal-fetal interface, analogous to invasive-type trophoblast in human placenta. Because PPARc has been implicated in human trophoblast invasion [8], we analyzed PPARc-null TS cells in Matrigel invasion assays ( Figure 7B). PPARc-null TGC showed enhanced trophoblast invasion, on a per cell basis, which was reversed by reintroduction of PPARc ( Figure 10B). In addition, rosiglitazone inhibited trophoblast invasion, in a PPARcdependent manner ( Figure 7C). This data is consistent with previous studies showing inhibition of human cytotrophoblast invasion by PPARc agonists [9].
In summary, the characterization of PPARc-null TS cells has shed more light on the role of PPARc in trophoblast differentiation and placental development. The data are consistent with the in vivo defects of PPARc-deficient placentas and confirm the utility of cultured TS cells in studies of placental development. We have shown that PPARc regulates trophoblast proliferation, invasion, and labyrinthine differentiation, and we hypothesize that these effects are mediated, at least partially, through Gcm1. More extensive gene expression analysis, using whole mouse genome expression arrays, will further identify potential downstream targets of PPARc in trophoblast differentiation. These studies will help elucidate the role of PPARc in placental development and determine the utility of specific agonists in treatment of placenta-related pregnancy disorders.

Ethics Statement
All animal work was approved by, and conducted according to the guidelines of, the Institutional Animal Care and Use Committee at Brigham and Women's Hospital.

Morphologic Analysis of Differentiation
TS cells were cultured as described above, on glass coverslips. Cells were fixed in 2% paraformaldehyde in PBS for 15 minutes and permeabilized in 0.02% Triton X-100 in PBS for 2 minutes. Mouse anti-HA tag antibody (Abcam) and Alexa dye-conjugated phalloidin and secondary antibodies (Molecular Probes) were used per manufacturer's instructions. Coverslips were mounted using Vectashield mounting media containing DAPI (Vector labs). Trophoblast giant cells were identified by large single or double nuclei, each at least twice the size (judged as estimated cross-sectional area) of an undifferentiated TS cell nucleus; syncytiotrophoblast were identified as cells with 3 or more nuclei, each of which were similar in size to an undifferentiated TS cell nucleus.

Proliferation and Invasion Assays
Cell proliferation was assayed using the MTT Cell Proliferation Kit I (Roche), with cells plated in 96-well plates at an initial density of 5610 3 cells per well. Five replicate wells were assayed for each condition and/or time point.
Matrigel invasion assays were done as described previously [21] with the following modifications. Cells were differentiated for 7 days in TS media; on day 7, cells were trypsinized and plated onto the Matrigel at a concentration of 10 5 cells per insert. The media in the lower chamber was changed every day. After 3 days, the Transwell inserts were fixed for 15 minutes with 3.7% formaldehyde in PBS and washed with PBS. Cells that remained at the upper surface of the filters were scraped off and filters were stained

RNA Isolation and Quantitative Reverse Transcription-PCR (qRT-PCR)
DNase-treated total RNA was isolated using NucleoSpin RNA II kit (Clontech). cDNA was prepared from 1 mg RNA using iScript (Bio-Rad) in a 20 ml reaction, and diluted 1:5 with nuclease-free water. PCR was performed using 4 ml of the diluted cDNA, along with 625 nM of each primer and POWER SYBR Green PCR master mix (Applied Biosystems) in a total reaction volume of 20 ml. qRT-PCR was performed using a System 7300 instrument (Applied Biosystems) and a one-step program: 95uC, 10 min; 95uC, 30 s, 60uC, 1 min, for 40 cycles. Unless otherwise stated, each experiment was performed in triplicate, and all results were normalized against 18S rRNA. Relative mRNA expression levels, compared to 18S rRNA, were determined by the DDC T method [28]. Fold change in normalized expression of individual genes in experimental samples was determined by comparison to expression in undifferentiated PPARc +/+ cells. All primer pairs (Table 1) were checked for specificity using BLAST analysis and were checked by both agarose gel electrophoresis and thermal dissociation curves to ensure amplification of a single product with the appropriate size and melting temperature.

Whole Cell Lysate Preparation and Western Blotting
Cells were washed with cold 1X PBS and lysates were prepared by scraping cells in boiled 1X Laemmli sample buffer (Bio-Rad), and passing the lysate through a 25G needle to sheer DNA and decrease viscosity. Protein determination was done using a Coomassie-binding assay [29]. For Western blot, 20 mg of total protein was loaded in each lane. After transfer to PVDF, the membranes were blocked with 5% non-fat dried milk in TBS-Tween for 1 hour at room temperature. Primary and secondary antibody incubations were similarly performed, with antibodies diluted in blocking buffer per the manufacturer's recommendation. Antibodies used included mouse anti-PPARc monoclonal antibody and rabbit polyclonal antibodies against PPARa and PPARb (all three from Santa Cruz Biotechnology, Inc.), mouse anti-actin monoclonal antibody (Abcam) used as loading control, and HRP-conjugated secondary antibodies (Jackson Immunochemicals). After treatment with ECL reagent (Pierce), membranes were exposed to autoradiography film (Kodak). Expression was quantified by densitometric scanning followed by normalizing PPARc, PPARa and PPARb/d expression to that of b-actin.

Generation of Adenovirus and Transduction of TS Cells
Full length murine PPARc1 cDNA was obtained by RT-PCR from E10.5 placental RNA, confirmed by sequence analysis, and cloned into pShuttle-IRES-hrGFP-2 adenoviral vector (Stratagene), in order to generate a hemagglutinin (HA)-tagged  . Labyrinthine marker assessment in null TS cells infected with either GFP-or PPARc-expressing adenovirus. Gcm1, muc1, and SynA, but not Tfeb, are increased in PPARc-virus-infected null cells. White columns indicate values for an aliquot of the same undifferentiated PPARc 2/2 TS cells used for transduction but not exposed to virus. *indicates P,0.05 compared to pAd-GFP-infected cells. doi:10.1371/journal.pone.0008055.g011 Figure 12. Model of PPARc's role in trophoblast differentiation. Note that PPARc contributes to both maintenance of undifferentiated trophoblast (by inhibiting premature differentiation) and differentiation towards the labyrinthine lineages. Errb + cells denote undifferentiated TS cells; FGF4/CM refers to the presence of FGF4 and MEF-conditioned media, both of which are required for maintenance of the undifferentiated state. The spongiotrophoblast (Tpbpa + giant cell precursor) and trophoblast giant cell (Pl-I + , Pl-II + , Plf + ) lineages are distinguished from labyrinthine precursor (Tpbpa 2 ) cells as well as fully differentiated labyrinthine cells (synA + SynT-I, Gcm1 + SynT-II, and Ctsq + mononuclear cells). doi:10.1371/journal.pone.0008055.g012 protein (HA-PPARc) co-expressed with GFP. This construct was used to generate adenovirus using the AdEasy XL system (Invitrogen). Large-scale production of adenovirus was accomplished by infecting subconfluent 293A cells (Invitrogen) with the crude viral lysate in 15-cm tissue culture dishes. Virus was purified by CsCl gradient centrifugation, desalted by chromatography on PD-10 columns (Amersham), and stored in 10% glycerol in PBS at 280uC. Titers were determined by plaque assay using 293A cells, according to Invitrogen's Virapower Adenoviral Expression System protocol. TS cells were infected with this purified virus, at an MOI of 10-50, in growth media (TSMFH media: 70% MEF-conditioned medium, with FGF4 and heparin); the day after infection, cells were switched to differentiation media (TS media only) and cultured an additional 1-7 days before harvesting. Expression of HA-PPARc from this adenovirus was confirmed by both immunofluorescence and Western blot using an anti-HA tag antibody (Abcam). On average, this protocol resulted in infection of 20-50% of TS cells.

Statistical Analysis
Unless otherwise stated, data presented are mean 6 standard deviation of three replicate wells of a single prep of cells; the results shown are representative of 3-5 separate experiments each performed on different days and using a different preparation of cells. Student's t-test was performed and a P value of 0.05 was taken to indicate a statistically significant difference between the populations sampled. Standard deviations are illustrated in all figures, but in some cases are too small to observe.