PPARγ Transcription Deficiency Exacerbates High-Fat Diet-Induced Adipocyte Hypertrophy and Insulin Resistance in Mice

Background The transcriptional factor peroxisome proliferator–activated receptor γ (PPARγ) is an important therapeutic target for the treatment of type 2 diabetes. However, the role of the PPARγ transcriptional activity remains ambiguous in its metabolic regulation. Methods Based on the crystal structure of PPARγ bound with the DNA target of PPARγ response element (PPRE), Arg134, Arg135, and Arg138, three crucial DNA binding sites for PPARγ, were mutated to alanine (3RA), respectively. In vitro AlphaScreen assay and cell-based reporter assay validated that PPARγ 3RA mutant cannot bind with PPRE and lost transcriptional activity, while can still bind ligand (rosiglitazone) and cofactors (SRC1, SRC2, and NCoR). By using CRISPR/Cas9, we created mice that were heterozygous for PPARγ-3RA (PPARγ3RA/+). The phenotypes of chow diet and high-fat diet fed PPARγ3RA/+ mice were investigated, and the molecular mechanism were analyzed by assessing the PPARγ transcriptional activity. Results Homozygous PPARγ-3RA mutant mice are embryonically lethal. The mRNA levels of PPARγ target genes were significantly decreased in PPARγ3RA/+ mice. PPARγ3RA/+ mice showed more severe adipocyte hypertrophy, insulin resistance, and hepatic steatosis than wild type mice when fed with high-fat diet. These phenotypes were ameliorated after the transcription activity of PPARγ was restored by rosiglitazone, a PPARγ agonist. Conclusion The current report presents a novel mouse model for investigating the role of PPARγ transcription in physiological functions. The data demonstrate that the transcriptional activity plays an indispensable role for PPARγ in metabolic regulation.

Rosiglitazone (Avandia) and pioglitazone (Actos) belong to an anti-diabetic drug class that targets PPARg, called thiazolidiniones (TZDs) (Larsen et al., 2003;Karak et al., 2013). As full agonists of PPARg, TZDs induce the transcription and expression of hundreds of genes by activating PPARg. Some of these activated genes enhance insulin sensitivity, leading to the therapeutic effects; while activation of some other genes are thought to be the causes of adverse effects of TZDs including weight gain, fluid retention, congestive heart failure, and bone fractures (Ahmadian et al., 2013;Wright et al., 2014). These adverse effects caused by PPARg full agonists might override the glycemic benefits in T2D patients. In fact, rosiglitazone has ever been suspended by the European Medicines Agency and restricted by the U.S. FDA.
The concept of a disconnect between the agonism potency of PPARg agonists and their therapeutic property has been proposed years ago (Jones, 2010). Much evidence show that partial agonists of PPARg, also called selective PPARg modulators (SPPARM) with poor agonist activities, such as MRL24, INT-131, and MBX-102, exert good anti-diabetic property with fewer adverse effects (Acton et al., 2005;Gregoire et al., 2009;Taygerly et al., 2013). Actually, ligands that do not possess transcriptional agonism can potentially exhibit anti-diabetic property with little adverse effect by blocking Cdk5-mediated phosphorylation of PPARg (Choi et al., 2011). Therefore, the characterization of PPARg transcriptional activity in drug discovery remains unclear till now.
The controversy over TZD drugs as diabetic treatment has weakened confidence in developing drugs that target the PPAR family of nuclear receptors. Reports have demonstrated that heterozygous PPARg-deficient mice exhibit improved insulin sensitivity (Kubota et al., 1999;Miles et al., 2000), supporting the negative role of PPARg. However, there are also other reports suggesting that PPARg bearing mutations in DBD or LBD are associated with lipodystrophy (Barroso et al., 1999;Freedman et al., 2005;Agostini et al., 2006;Jeninga et al., 2007). Therefore, there is a dire need to create an applicable model to clarify the role of the transcriptional activity of PPARg in metabolism. Based on the crystal structure of the PPARg-RXRa complex bound to PPRE (Chandra et al., 2008), we identified three crucial residues (Arg134, Arg135, and Arg138) on PPARg that control the binding ability of PPARg with PPRE and the subsequent transcriptional activity of PPARg. Therefore, we created a transgenic mouse model containing the three point mutations (R134/135/138A, 3RA) to study the role of PPARg transcription in metabolism.

Protein Purification
Human PPARg containing domains from DBD to the C-terminus (CDE domains, residues 103-477) was expressed as an N-terminal 6×His fusion protein (H6-PPARg CDE) from the expression vector pET24a (Novagen, Germany). 3RA mutant plasmid was constructed by site-directed mutagenesis with forward primer: AGGATGCAAGGGTTTCTTCGCGGCAACAA TCGCATTGAAGCTTATCTATGACAG, and reverse primer: CTGTCATAGATAAGCTTCA ATGCGATTGTTGCCGCGAAGAAACCCTTGCATCCT, using Pfu DNA polymerase (Thermo Fisher Scientific, USA). BL21(DE3) cells transformed with the expression plasmids were grown in LB broth at 25°C to an OD 600 of approximately 1.0 and induced with 0.1 mmol/L isopropyl 1-thio-b-Dgalactopyranoside (IPTG) at 16°C. Cells were harvested and sonicated in 100 ml of extract buffer (20 mmol/L Tris pH8.0, 150 mmol/L NaCl, 10% glycerol, and 25 mmol/L imidazole) per 2 liters of cells. After sonication, the lysate was centrifuged at 20,000 rpm for 30 min, and the supernatant was loaded on a 5-ml NiSO 4 -loaded HiTrap HP column (GE Healthcare, PA, USA). The column was washed with extract buffer, and the protein was eluted with a gradient of 25 to 500 mmol/L imidazole. The PPARg CDE was further purified with a SP-Sepharose column (GE Healthcare, PA, USA).

AlphaScreen Assay
The binding of H6-PPARg CDE wild-type (WT) or H6-PPARg CDE 3RA mutant protein with biotin-labeled PPRE was determined by AlphaScreen assay using a hexahistidine detection kit from Perkin-Elmer. PPRE was prepared by annealing biotin-PPRE-F: AGGGGACCAGGACAAAGGTCA CGTTCGGGA and biotin-PPRE-R: TCCCGAACGTGAC CTTTGTCCTGGTCCCCT, both of which with 5' end biotinlabeled. The assay was performed in a buffer containing 50 mmol/L MOPS, 50 mmol/L NaF, 0.05 mmol/L CHAPS, and 0.1 mg/ml bovine serum albumin, all adjusted to a pH of 7.4. The binding assay was performed with 100 nM of protein with gradient doses of biotin-PPRE, or 1 nM of biotin-PPRE with gradient doses of H6-PPARg CDE with or without 1 µM of rosiglitazone. For the binding of PPARg CDE with cofactors peptide motifs in response to rosiglitazone, AlphaScreen assay was performed with 100 nM of PPARg CDE, 100 nM biotinlabeled peptides with gradient doses of rosiglitazone. The sequences of the peptides: SRC1-2, SPSSHSSLTERHKILHRL LQEGSP; SRC2-3, QEPVSPKKKENALLRYLLDKDDTKD; and NCoR-2, GHSFADPASNLGLEDIIRKALMGSF.

Dual Luciferase Report Assay
HEK-293T cells (ATCC, USA) were maintained in DMEM containing 10% fetal bovine serum (FBS) and were transiently transfected using Lipofectamine 2000 reagent (Thermo Fisher Scientific, USA). 24-well plates were plated 24 h prior to transfection (5 x 10 4 cells per well). 200 ng of pcDNA3.1-Flag-PPARg WT or 3RA mutant plasmid was co-transfected with 200 ng of PPRE-luc reporter plasmid into cells (Zheng et al., 2013). Renilla was co-transfected as an internal control. 1 µM of rosiglitazone or DMSO was added 5 h after transfection. Cells were harvested 24 h later for the luciferase assays. Luciferase activities were analyzed as the instruction of CheckMate ™ Mammalian Two-Hybrid System (Promega, USA).

Generation of PPARg 3RA/+ Mice
A PPARg BAC clone was screened and isolated from BAC library, mapped by restriction digests and sequenced. The arginine 134, 135, and 138 residues in exon 5 were all paralleled mutated to alanine (3RA) using overlap PCR, and the fragment was cloned into a targeting vector that contains exon 5 homology arm. Meanwhile, the vector contains cassette with a floxed pGK-neo r . The vector was then delivered to embryonic stem (ES) cells (C57BL/6) via electroporation, followed by G418 selection, PCR screening, and Southern blot confirmation. Targeted lines were expanded and electroporated with a Cre recombinants expression vector to delete the neo rcassette. Some correct targeted ES clones were selected for blastocyst microinjection, followed by chimera production in C57BL6 background. These mice were then interbred to obtain different genotypes littermate mice for experiments. Mice were maintained under environmentally controlled conditions with free access to diet and water. Animal experiments were conducted in the barrier facility of the Laboratory Animal Center, Xiamen University, approved by the Institutional Animal Use and Care Committee of Xiamen University, China. The methods were carried out in accordance with the approved guidelines.

Mice Treatment
8 week-old male PPARg 3RA/+ and WT littermates were fed with a high-fat diet (HFD, 60% kcal fat, D12492, Research Diets Inc, USA). The body weight of mice were weighed weekly and the food intake was assessed every 4 weeks. Blood samples were obtained by the tail-cut method for small samples every 4 weeks for detecting blood glucose and insulin levels. After 15 weeks of HFD, mice were euthanized after 6 h of fasting. For rosiglitazone treatment study, mice were divided into two groups after a 15week HFD, and intraperitoneally (i.p.) injected once daily with vehicle (40% of 2-hydroxypropyl-b-cyclodextrin, HBC, Sigma, USA) or 3 mg/kg of rosiglitazone for 6 days. Mice were euthanized after 6 h of fasting. For all mice research, part of liver and fat tissues was fixed in 4% paraformaldehyde for hematoxylin and eosin (H&E) staining by standard procedures. Other tissues were collected and frozen in liquid nitrogen for use. Serum was collected for the measurement of metabolic parameters. Animal experiments were conducted in the barrier facility of the Laboratory Animal Center, Xiamen University, approved by the Institutional Animal Use and Care Committee of Xiamen University, China.

GTT and ITT
Glucose tolerance test (GTT) and insulin tolerance test (ITT) were performed in mice before and after a 15-week HFD feeding. For the GTT, mice were fasted for 16 h with free access to water, and then orally gavaged with 1 g/kg body weight of glucose. Blood glucose level was assessed with the Accu-Check Performa (Roche Applied Science, Mannheim, Germany) at 0, 15, 30, 60, 90, and 120 min. For the ITT, mice were fasted for 6 h with free access to water, and then i.p. injected with 1 U/kg of recombinant human insulin (Novolin 30R; Novo Nordisk, Bagsvaerd, Denmark). Blood glucose level was measured at 0, 15, 30, 60, and 120 min after insulin injection.

Gene Expression
The protein level of PPARg in inguinal WAT (iWAT) was assessed by western blot using mouse monoclonal anti-PPARg (Santa Cruz, Cat. No. sc-7273, 1:1000) (Chakraborty et al., 2019;Jung et al., 2019) and mouse monoclonal anti-b-actin (Protein Tech, Cat. No. 60008-1-Ig, 1:2000). Total RNA was isolated from liver and fat tissues using Tissue RNA kit (Omega Bio-Tek, GA). The first strand cDNA was reverse-transcribed using TAKARA reverse transcription kit. Real-time quantitative PCR reactions were performed with SYBR Premix Ex TaqTM (TAKARA) on a CFX96 ™ Real-Time PCR Detection System (Bio-Rad). Relative mRNA expression levels were normalized to b-actin levels. The sequences of the primers used were listed in Supplementary  Table 1.

Statistical Analysis
Values were expressed as mean ± standard error of mean (SEM). Statistical differences were calculated by one-way ANOVA followed by the Dunn's test or Student's t test. Statistical significance was shown as *p<0.05, **p<0.01 or ***p<0.001.

R134/135/138A Mutations Abolish the PPRE-Binding Ability and the Transcriptional Activity of PPARg
To evaluate the role of the transcriptional function of PPARg on metabolism, we attempted to create a mouse model that is deficient in the transcriptional activity of PPARg while the DBD-independent actions of PPARg are intact. Because PPARg needs to bind to PPRE to activate the downstream transcription, we searched for crucial sites in PPARg to destroy its binding on PPRE. Based on the crystal structure of the PPARg-RXRa complex bound to PPRE (Chandra et al., 2008), we found that the Arg134, Arg135, and Arg138 residues in PPARg form six hydrogen bonds in the major groove of the PPRE double helix ( Figures 1A-C). However, if the three arginine residues were mutated into alanine, the six hydrogen bonds will not form ( Figure 1D). The absence of the hydrogen bonds is predicted to abolish the binding between PPARg and PPRE while sparing other functions crucial for the transcriptional activity of PPARg including the zinc finger structure of PPARg DBD and the ligand-binding LBD (Chandra et al., 2008).
Therefore, we mutated the three arginine residues at Arg134, Arg135, and Arg138 of PPARg to alanine (named PPARg-3RA). The binding ability of PPAR-3RA with PPRE was studied by using AlphaScreen assay. 6 × His tag-PPARg WT or PPARg-3RA that contains the domains ranging from DBD, hinge, and the C-terminus LBD of PPARg (H6-PPARg CDE) was expressed in BL21 (DE3) and purified for the assay. As expected, the binding signal of the H6-PPARg CDE WT increased in a PPRE concentration-dependent manner ( Figure 1E). In contrast, H6-PPARg CDE 3RA did not show any binding signal with increasing concentration of PPRE. We obtained similar results when a gradient concentration of H6-PPARg CDE WT or H6-PPARg CDE 3RA was used to bind PPRE ( Figure 1F). Additionally, the same result was produced with the treatment of rosiglitazone, an agonist for PPARg ( Figures 1E, F), due to the ligand-independent nature of the binding ability of DBD domain to DNA, which is distinct from LBD (Rastinejad et al., 2013). These data confirmed that the PPARg-3RA mutant lost the binding ability toward PPRE. Next, we tested the transcriptional activity of the PPARg-3RA mutant. WT or 3RA mutant pcDNA3.1-Flag-PPARg was co-transfected with PPREluc plasmid into HEK-293T cells for luciferase reporter assay. The result showed that the transcriptional activity of PPARg dramatically decreased after 3RA mutation. Rosiglitazone treatment significantly induced the transcriptional activity of WT PPARg, but failed to activate PPARg-3RA ( Figure 1G).
To test if the 3RA mutations affect the binding ability of PPARg with ligands and cofactors, we performed an AlphaScreen assay using PPARg CDE WT or 3RA protein and biotin-labeled cofactors peptides in response to rosiglitazone. The results showed that both WT and 3RA mutant of PPARg CDE can recruit co-activators SRC1 and SRC2 and release corepressor NCoR in response to rosiglitazone ( Figures 1H-J). Our data demonstrate that although PPARg 3RA mutant cannot bind with PPRE ( Figures 1E, F), the mutant is still able to bind ligands and cofactors such as SRC1, SRC2, and NCoR.
Together, these results demonstrate that the PPARg-3RA lost the binding ability toward PPRE and further the DBD-dependent ligand-regulated transcriptional activity of PPARg, while maintains the ability to bind ligands such as rosiglitazone and cofactors, such as SRC1, SRC2, and NCoR.

Decreased Transcriptional Ability of PPARg in Heterozygous PPARg 3RA/+ Mice
We mutated codons for amino acids 134, 135, and 138 of the mouse PPARg gene from CGA (arginine), AGA (arginine) and CGA (arginine) to GCA (alanine), respectively, via gene targeting in mouse ES cells and created genetically modified mouse carrying the 3RA mutations (Figures 2A-C). In 74 progenies born from PPARg 3RA/+ intercrosses, no PPARg 3RA/3RA homozygous mice were obtained. WT and PPARg 3RA/+ littermates were born at the expected Mendelian ratio (26:48 ≈ 1:2) ( Figure 2D), indicating that the PPARg 3RA mutations cause embryonic lethality due to the loss of PPARg transcriptional activity. These results suggest that the transcriptional function of PPARg is essential for embryonic development.
To confirm the decreased transcriptional ability of PPARg in heterozygous PPARg 3RA/+ mice, we analyzed gene expression in iWAT from WT and PPARg 3RA/+ mice. The expression level of total PPARg was higher in iWAT from PPARg 3RA/+ than WT mice ( Figure 2E) which might be the compensatory expression due to the loss of transcriptional activity for PPARg 3RA mutation in vivo. Our in vitro reporter assay indicated that co-existence of PPARg 3RA mutant significantly reduced the transcriptional activity of WT PPARg ( Figure 2F), further supporting the decreased transcriptional activity in heterozygous PPARg 3RA/+ mice. As expected, the mRNA levels of genes that are directly downstream of PPARg, such as FAT/CD36, PEPCK, and AQPap, were significantly lower in iWAT of PPARg 3RA/+ mice compared to those of WT littermates ( Figure 2G). The differential mRNA expression level of PPARg downstream genes in WT and PPARg 3RA/+ mice confirmed the impaired transcriptional activity of PPARg in the PPARg 3RA/+ mice.

PPARg 3RA Mutations in Mice Exacerbate HFD-Induced Obesity and Adipocyte Hypertrophy
Under chow diet, the WT and PPARg 3RA/+ littermates showed similar phenotypes in body weight, liver/body weight ratio, fat/ body weight ratio, as well as the histological analysis of the brown adipose tissue (BAT), WATs and the liver tissue ( Figure  3A and Supplementary Figure 2). The similarity in these parameters suggests that the transcriptional activity of one PPARg allele is enough for maintaining basic metabolism in  the absence of external stimuli. Interestingly, when fed HFD, PPARg 3RA/+ mice gained significantly more body weight than WT mice did ( Figure 3A) even though food intake was similar (Supplementary Figure 3). After 15 weeks of HFD feeding, the WATs and livers of PPARg 3RA/+ mice weighed significantly more than those of WT mice, while the BAT of PPARg 3RA/+ mice weighed significantly less than that of WT mice ( Figures  3B, C). Histological analysis by H&E staining showed larger adipocytes size in the sections of iWAT, gonadal WAT (gWAT), and BAT from PPARg 3RA/+ mice than those of WT mice ( Figures 3D, E). These results indicate that heterozygous PPARg deficiency leads to more severe hypertrophy in white adipocytes and more whitening in brown adipocytes in mice under HFD. Notably, there was also visible inflammatory infiltration in iWAT of PPARg 3RA/+ mice ( Figure 3D). Additionally, histological examination showed that heterozygous PPARg deficiency leads to more lipid accumulation in the liver of mice under HFD ( Figure 3F), which was further confirmed by the biochemical analysis of the hepatic triglycerides level ( Figure 3G). As the results shown, the fasting plasma levels of total cholesterol (TCHO) ( Figure  3H), triglyceride (TG) ( Figure 3I), LDL-C, and FFA ( Figures  3J, K) of PPARg 3RA/+ were all significantly higher than those of WT littermates, whereas the level of HDL-C ( Figure 3L) was significantly lower. These results demonstrate that PPARg 3RA mutations exacerbated HFD-induced obesity and adipocyte hypertrophy in mice.

PPARg 3RA Mutations in Mice Exacerbate HFD-Induced Insulin Resistance
Chow diet-fed PPARg 3RA/+ and WT littermates showed similar fasting blood glucose level, glucose tolerance, and insulin tolerance ( Figures 4A, B). When fed HFD, PPARg 3RA/+ and WT mice maintained similar fasting blood glucose levels ( Figure  4C). However PPARg 3RA/+ mice showed significantly higher level of fasting blood insulin from 8 weeks after HFD-feeding ( Figure  4D). Furthermore, PPARg 3RA/+ mice showed impaired glucose tolerance and insulin tolerance compared to their WT counterparts, suggesting that PPARg 3RA mutations in mice exacerbate HFD-induced insulin resistance ( Figures 4E, F). Taken together, our data demonstrate that the decreased transcriptional activity of PPARg in PPARg 3RA/+ mice led to impairment of lipid and glucose metabolism under HFD.

Metabolic Disorders in HFD-Fed
PPARg 3RA/+ Mice Were Improved by Rosiglitazone Treatment Next, we aimed to investigate whether or not the metabolic disorders induced by HFD in PPARg 3RA/+ mice could be reversed by increasing the transcriptional activity of PPARg. Rosiglitazone was administered at 3 mg/kg once daily for 6 days to PPARg 3RA/+ and WT littermates that had been fed HFD for 15 weeks. Metabolic parameters were studied after the treatment. As shown in Figure 5, rosiglitazone treatment significantly reduced or showed the tendency to reduce the levels of cholesterol, triglyceride, FFA, LDL-C, and glucose in the serum, while increased the level of HDL-C in the serum of both PPARg 3RA/+ and WT littermates ( Figures 5A-F). Histological examination further showed that after HFD feeding, both WT and PPARg 3RA/+ mice administrated with rosiglitazone showed less fat vacuoles in BAT and smaller adipocyte size in WAT ( Figures 5G, H). Notably, rosiglitazone administration significantly improved the inflammation in WAT of HFD-fed PPARg 3RA/+ mice ( Figure 5G). Additionally, rosiglitazone treatment not only efficaciously improved the hepatic steatosis in WT mice, but also in PPARg 3RA/+ mice ( Figure 5I). We further investigated the mRNA levels of PPARg target genes in the adipose tissue. As the results showed, the mRNA levels of the target genes of PPARg we tested were significantly decreased in PPARg 3RA/+ mice compared to those in WT mice ( Figure 5J). Rosiglitazone treatment significantly induced the expression PPARg target genes in WT mice; in PPARg 3RA/+ mice, rosiglitazone treatment restored the expression level of PPARg target genes in PPARg 3RA/+ mice to levels similar to those of vehicle-treated WT mice ( Figure 5J). Taken the rosiglitazoneinduced metabolic improvement and gene expression restoration together, our results suggest that increasing PPARg transcriptional activity could overcome the HFD-induced obesity and adipocyte hypertrophy in PPARg 3RA/+ mice.

DISCUSSION
In this study, based on the structural analysis of PPARg/PPRE complex (Chandra et al., 2008), the roles of arginine at 134, 135, and 138 residues of PPARg in binding PPRE, ligand (rosiglitazone) and cofactors (SRC1, SRC2, and NCoR) were verified by in vitro biochemical AlphaScreen and cell-based reporter assays. PPARg may also possess regulatory mechanisms independent of DBD, such as PPARg Ser273 phosphorylation mediated by CDK5 (Choi et al., 2011). Because previous reports have demonstrated that the regulation of PPARg Ser273 phosphorylation can be detected by in vitro kinase assay in a reaction system including PPARg LBD, CDK5, and ATP (Choi et al., 2011;Zheng et al., 2013), and the results from in vitro assay are consistent with that from in vivo assay, these results suggest that the post-translational phosphorylation of PPARg at Ser273 by CDK5 is not rely on the existence of PPARg DBD. Thus, the phosphorylation mediated by CDK5 may also be preserved in the 3RA mutant, although this has not been confirmed experimentally. Together, these data suggest that DBD-independent PPARg regulations are intact in the PPARg 3RA mutant. Therefore, we created a knock-in mouse model containing the PPARg 3RA mutations. Homozygous PPARg 3RA/ Values are means ± SEM, n=6 per group, *p < 0.05, **p < 0.01 and ***p < 0.001 by one-way ANOVA followed by the Dunn's test.
3RA leads to embryonic death, suggesting the necessary role of the transcriptional activity of PPARg for development. Chow diet fed PPARg 3RA/+ mice showed decreased transcriptional activity of PPARg, while maintained comparable phenotypes with the WT littermate mice, suggesting that the one PPARg alleles is sufficient to maintain the organismal metabolic network without stimuli. However, due to impaired transcriptional activity, PPARg 3RA/+ mice cannot sustain the burden of HFD stimuli, and appeared more severe insulin resistance and obesity. Accordingly, PPARg agonist treatment rescued the activity of PPARg, and restored the metabolic disorders in HFD-fed PPARg 3RA/+ mice. These results would indicate the important role of the transcriptional activity of PPARg in protecting mice from HFD-stimulated metabolic disorders.
PPARg plays crucial roles in maintaining the homeostasis of glucose and lipid metabolism. Over activating or debilitating its downstream signaling may cause the imbalance of the homeostasis (Rubenstrunk et al., 2007). For example, a water/ glycerol transporting protein AQP7 regulates adipocyte glycerol efflux and influences lipid and glucose homeostasis. The deletion of AQP7 gene in mice leads to obesity and T2D (Rodriguez et al., 2006). However, it has also been reported that either increased or decreased AQP7 expression may lead to impaired glycerol dynamics and adipocyte hypertrophy (Oikonomou et al., 2018). Another example is that GLUT-4 is necessary for the insulin-regulated glucose uptake into muscle and fat cells which keeps the glucose homeostasis (Wu et al., 1998). However, if GLUT4 is over-expressed, it will send excess glucose into adipose tissue, leading to increased adipose cell hypertrophy and obesity (Shepherd et al., 1993). Also, overexpression of SCD1 in humans may be involved in the development of hypertriglyceridemia, atherosclerosis, and diabetes (Mar-Heyming et al., 2008). While inhibiting SCD1 function may also result in the accumulation of fatty acid metabolites that are deleterious to insulin signaling, and accordingly, the development of fatty acid-induced insulin resistance (Pinnamaneni et al., 2006). Thus, disorders of these genes will result in an imbalance of nutrients distribution and lead to obesity and diabetes. Rosiglitazone induces the expression of PPARg target genes, which may provide a potential lighthouse to explain the adverse effects of long-term administration of TZD drugs (Rubenstrunk et al., 2007), as well as the metabolic improvement in HFD-fed PPARg 3RA/+ mice.
A number of laboratories have reported metabolic changes observed in heterozygous PPARg-deficient mice (Barak et al., 1999;Kubota et al., 1999;Miles et al., 2000). Contrary to our finding, Kubota and colleagues reported that heterozygous PPARg-deficient mice were protected from the development of insulin resistance caused by adipocyte hypertrophy after HFDfeeding. After administration of pioglitazone, the mice showed worsened phenotypes. Similarly, another group has reported improved insulin-sensitivity in an independently generated heterozygous PPARg-deficient mouse model (Miles et al., 2000). These reports seem to approbate the negative roles of PPARg transcriptional activity in metabolic regulation. However, PPARg is required for adipose tissue development. Barak et al. found that the absence of PPARg in mice leads to complete lipodystrophy (Barak et al., 1999), indicating the necessary role of PPARg in lipid metabolism. It should be noted that besides transcriptional regulation, all of the five domains of PPARg are involved in modulating the PPARg signaling cascades (Wang S. B. et al., 2016). In this process, except transcriptional regulation by binding to PPREs, cofactors binding, and post-translational modifications including phosphorylation, acetylation, sumoylation, and ubiquitination throughout the full length of PPARg also contribute to the functions regulated by PPARg (Jin and Li, 2010;Ahmadian et al., 2013). However, both of the heterozygeous PPARg-deficient mouse models by groups of Kubota and Barak eliminate most of the domains of PPARg from DBD to the C-terminus (Barak et al., 1999), and thus the PPARg in these models lost not only the transcription activity, but also other functional regulations. On the contrary, our PPARg 3RA model is only mutated at three residues in DBD which contribute to PPARg transcription deficiency, therefore may represent a suitable tool for the research of the role of transcriptional function of PPARg in metabolism. The different phenotypes between PPARg-deficient mice and PPARg 3RA mutant mice further suggest that PPARg needs to coordinate its transcriptional activity and its non-transcriptional regulatory actions for metabolic regulation.
Among the domains in nuclear receptors, the sequence of DBD shows the highest evolutionary conservation (Jin and Li, 2010;Helsen et al., 2012). Importantly, the three arginine residues we selected for mutation are conserved from birds to mammals including rodents and humans (Supplementary Figure 4), suggesting the conserved function or the PPARg 3RA mutant in evolution, including humans. Additionally, the PPRE for PPARg binding is also conserved with a direct repeats of hexameric sequence AGGTCA in different target genes, although each gene has distinct flanking sequence for its selective regulation (Khorasanizadeh and Rastinejad, 2001). Therefore, our PPARg 3RA mutant model is a suitable tool for the research of PPARg transcription in evolution.
In conclusion, we provide an alternative mouse model for further research on the transcriptional activity of PPARg, and also for the drug discovery by targeting PPARg. It should be noted that more detailed investigation about the DBDindependent PPARg actions will further improve the significance of this mouse model. Considering the embryonic death of the PPARg-3RA mice, future research will focus on creating homozygous conditional knockout mouse model with tissue specific PPARg 3RA mutations to completely investigate the role of PPARg transcriptional activity in specific tissues, particularly the adipose tissues and liver.

DATA AVAILABILITY STATEMENT
The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation, to any qualified researcher.

ETHICS STATEMENT
The animal study was reviewed and approved by the Laboratory Animal Center, Xiamen University.

AUTHOR CONTRIBUTIONS
YLi, LJ and YH designed the experiment, LJ and FG wrote and revised the manuscript. FG, SX, and YZ performed experiments. YLi and JT discussed and revised the manuscript. XZ assisted with the mice experiments. YLu and FG contributed to the structural analysis. All authors contributed to the article and approved the submitted version.