The phosphorylation regulation of PPARγ is one of the main ways to regulate its activity (Montanari et al., 2020). With different stimuli, PPARγ could be phosphorylated at different sites and resulting diverse biological effects (Choi et al., 2011). Cyclin-dependent kinase (CDK) (Choi et al., 2010, 2011; Laghezza et al., 2018) and mitogen-activated protein kinase (MAPK) (Yin et al., 2006; Yang et al., 2013; Ge et al., 2018) are involving in the phosphorylation of PPARγ, and the main sites include Ser273 (245 in isoform 1) (Dias et al., 2020; Hall et al., 2020) and Ser112 (Ser82 in isoform 1) (Figure 2A; Ge et al., 2018). CDK5-mediated phosphorylation of PPARγ S273 results in a decrease in its transcriptional activity and adipogenesis. One of TZDs’ major side effects is due to its activation of PPARγ in adipose tissue, and high-fat diet increased CDK5-mediated of PPARγ phosphorylation, which was negatively associated with TZDs’ insulin sensitization in humans. Meanwhile PPARγ phosphorylation can up-regulate lipid uptake of CD36 and SR-A1 related proteins, inhibit cholesterol efflux ABCA1 and ABCG1 related proteins, induce the expression of TNF-α, IL-1β and other inflammatory factors, and promote the formation of foam cells to accelerate the process of atherosclerosis (Choi et al., 2010, 2011; Banks et al., 2015). PPARγ phosphorylation by CDK5 may contribute to its dissociation with PGC1α and TIF2 coactivators but interaction with SMRT and NCoR corepressors (Dias et al., 2020). NCoR can regulate the phosphorylation of PPARγ on Ser 273 by stabilizing CDK5 (Li et al., 2011). Mice with fat cell NCoR knockout (NCoR^–/–) improves glucose tolerance and insulin sensitivity and reduces macrophage infiltration and inflammation (Ribeiro Filho et al., 2019). As from CDK5 studies, compounds can be designed to alter specific PTMs of PPARγ to partly prevent disturbed fat metabolism while retaining anti-diabetic potency. CDK9/CDK7-mediated phosphorylation of S112 can increase the transcriptional activity of PPARγ and promote adipocyte differentiation (Iankova et al., 2006; Li et al., 2017; Ge et al., 2018).

MAPK can phosphorylate PPARγ in the AF1 region (PPARγ2 Ser112, PPARγ1 Ser82), inhibiting the ligand binding and changing the recruitment of co-factors, and then change the transcriptional activity (Yin et al., 2006; Yang et al., 2013; Ge et al., 2018). MAPK-mediated phosphorylation of PPARγ, which promotes the formation of foam cells by macrophages exposed to ox-LDL (Yin et al., 2006). Growth factors can phosphorylate PPARγ by the MAPK signaling pathway and reduce the transcriptional activity of PPARγ (Yin et al., 2006), such as epidermal growth factor (EGF) and platelet-derived growth factor (PDGF) (Osman and Segar, 2016). In addition, Choi et al. (2015) recently described the phosphorylation of Y78, is also important for the cytokine and chemokine gene expression’s regulation. PPARγ phosphorylation can alter its transcriptional activity, and the blockage of PPARγ phosphorylation is related to improve insulin sensitization (Choi et al., 2014). However, PPARγ phosphorylation mediated by different enzymes, conformational changes at different sites can cause the recruitment response of different cofactors, its role in atherosclerosis and its mechanism need to be further studied, and also provide ideas for the drug design of PPAR ligand.

Acetylation of PPARγ is a ligand- independent activation of PPARγ. Qiang et al. (2012) showed that five lysine residues (K98, K107, K218, K268, and K293) could be acetylated, of which two K268/K293 could be deacetylated by TZD rosiglitazone (Rosi) via activation of the NAD (Nicotinamide adenine dinucleotide)-dependent deacetylase sirtuin-1 (SIRT1) deacetylase (Figure 2B; Kraakman et al., 2018). PPARγ is acetylated by p300 or CBP (Kim et al., 2006), and it may play an important role in lipid synthesis (Tian et al., 2014). Acetylation of PPARγ at K268/K293 increases atherosclerosis through upregulating ABCA1, ABCG1, and NcoR but inhibiting PRDM16, while deacetylation of PPARγ at K268 and K293 alleviates atherosclerosis.

PPARγ deacetylation on K268 and K293 induces brown remodeling of white adipose tissue and reduces the adverse effects of TZDs while maintaining insulin sensitization (Qiang et al., 2012; Liu L. et al., 2020). Deacetylation of PPARγ can selective regulation the target genes, it could inhibit aP2, Cd36, upregulation genes ucp1 and adipsin on lipid oxidative genes cpt1a (Kraakman et al., 2018). PPARγ deacetylation improves endothelial function with diabetes treatment. The aortic arch lesion size was reduced in 2KR (K268 and K293) LDLr^–/– mice, the expression of iNOS, Nox2, and IL-6 in endothelial cells were decreased, while the side effects of TZD, including fluid retention and bone loss were reduced (Liu L. et al., 2020). Deacetylation of PPARγ inhibits the cholesterol efflux PPARγ/LXRα/ABCA1 pathway (Cao et al., 2014; Yang et al., 2015), increased production of proinflammatory M1 macrophages and promotes the development of inflammatory response (Chen et al., 2008, 2010), leading to the onset and development of atherosclerosis. From a mechanistic perspective, deacetylated PPARγ preferentially interacts with PRDM16 and disrupts the binding of the transcriptional corepressor NCoR (Qiang et al., 2012). Therefore, manipulating PPARγ acetylation is a promising therapeutic strategy to anti-atherosclerosis.

PPARγ sumoylation with SUMO1 modification of K107 (K77 in PPARγ1) (Figure 2C), and ubiquitin carrier protein 9 (Ubc9) and PIAS1 (protein inhibitor of activated STAT1) are involved as PPARγ specific E2 binding enzymes and E3 ligases, respectively (Lim et al., 2009). Sumoylation of PPARγ K107 inhibits its transcriptional activity, and is enhanced by the K107 mutation (K107R). PPARγ sumoylation at K107 position strongly inhibited VSMCs proliferation and migration (Katafuchi et al., 2018), and reduced neointimal formation after balloon injury (Lim et al., 2009). Desumoylation at K107 in PPARγ may inhibit serum-stimulated VSMCs proliferation (Lim et al., 2009; Osman and Segar, 2016), might play an important role against atherosclerosis. Moreover, K107R improves insulin sensitivity without body weight gain or adiposity (Wadosky and Willis, 2012; Pourcet et al., 2013). However, some studies have shown that K107 sumoylation plays an important role in the anti-inflammatory response triggered by apoptotic cells, possibly by stabilizing the co-inhibitor NCoR on the target gene (Jennewein et al., 2008; Lu et al., 2013). The sumoylation modification of PPARγ1 inhibits the M2 polarization of macrophages by inhibiting the transcription of Arg-1 (Haschemi et al., 2011). Pascual et al. (2005) shows that the PPAR agonist TZD can exert anti-diabetic and anti-atherosclerotic effects through the NF-kB inflammatory pathway, the first response is associated with TZD-mediated SUMO1 modification of K365 (K395 in PPARγ2) followed by targeted regulation of the PPAR co-cofactor NCoR. The exact biological role of the two modifications in anti-inflammatory responses, especially the potential functional overlap, remains to be determined. Nevertheless, targeting PPARγ sumoylation may provide a novel mechanism for anti-atherosclerosis.

Ubiquitination modification cannot only regulate the proteasome-mediated degradation of target proteins, but also serve as a “scaffold” to recruit other proteins to form signal complexes. PPARγ undergoes conformational changes after binding to the ligand. On the one hand, Makorin RING finger protein 1 (MKRN1) (Kim et al., 2014) and Seven in absentia homolog 2 (SIAH2) (Kilroy et al., 2012) services as PPARγ E3 ligases, targeting PPARγ for proteasomal degradation. Ubiquitination of PPARγ on K184 and K185 inhibits its activity in mature 3T3-L1 adipocytes (Kilroy et al., 2012; Kim et al., 2014). On the other hand, it can also recruit the binding of ubiquitination-related enzymes and induce proteasome-dependent degradation, thereby negatively regulating the transcriptional activity of PPARγ. Rosiglitazone reduces the inflammatory response in diabetic plaques, less ubiquitin, proteasome 20S, TNF-α, and NF-κB, ubiquitin-proteasome activity with diabetic plaque NF-κB-mediated inflammatory response is involved (Marfella et al., 2006). This study strengthens the earlier findings on PPARγ regulation through modulation of its stability. Currently, the use of ubiquitination modifications for the regulation of PPARγ transcriptional activity is still controversial (Watanabe et al., 2015). It has been reported that the ubiquitin-proteasome pathway can mediate the protein renewal of PPARγ (Li et al., 2016), a process that is required for the efficient transcription of its downstream genes, while the ubiquitin activase inhibitor E1 inhibitor, and the proteasome inhibitor MG-132 can cause a decrease in PPARγ transcriptional activity.

In fact, partial PPARγ agonist is different from the classic TZDs, with rosiglitazone as the “full agonist.” It is generally believed that 20–60% of rosiglitazone efficacy is a partial activator (Table 1), such as GQ-177 (Silva et al., 2016), S 26948 (Carmona et al., 2007), WSF-7 (Zhang et al., 2020) lobeglitazone (Lim et al., 2015), and INT131 (Xie et al., 2017). LDLr^–/– mice treated with GQ-177 can significant decrease the VLDL, LDL fractions and increase mean HDL, Glut4 levels, increased the expression of apoA1, CD36, ABCA1, SR-B1, and ABCG5 in hepatic, contrary to rosiglitazone, GQ-177 did not affect fat accumulation and bone mineral density (Silva et al., 2016). It was shown that TZDs facilitated the transport of BM-derived circulating progenitor cells to adipose tissue and their differentiation into multilocular adipose cells (Crossno et al., 2006). Meanwhile, Hu et al. (2021) found RANKL from bone marrow adipose lineage cells promoted osteoclast formation and bone loss. S26948 improves lipid parameters (LDLs, VLDL) and reduces atherosclerotic lesions in ob/ob male C57BL/6 mice (Carmona et al., 2007). WSF-7 upregulated PPARγ-responsive genes, such as adiponectin and Glut4, inhibits PPARγ phosphorylation at Ser273 by obesity and enhances insulin sensitivity in 3T3-L1 Adipocytes (Zhang et al., 2020). Lobeglitazone inhibits the VSMCs proliferation and migration, reduces the vascular cells adhesion, NF-kB p65 translocation, and improves circulating factors related to atherosclerosis, then reduced neointimal formation significantly in balloon injury rat carotid arteries in ApoE^–/– mice (Lim et al., 2015; Song et al., 2021). In the presence of pro-inflammatory stimulation, Lobeglitazone effectively inhibited expression of pro-inflammatory gene expression in macrophages and adipocytes (Sohn et al., 2018). The latest study has found that macrophages targeted PPARγ activator Lobeglitazone could rapidly stabilize a coronary artery-sized inflammatory plaque (Song et al., 2021). Both non-clinical and clinical studies have demonstrated that INT131 have the potential to separate insulin-sensitizing actions and undesirable side effects in Patients With Type 2 Diabetes (DePaoli et al., 2014), and it also has the potential to decrease free fatty acids, increase HDL-C (Dunn et al., 2011). However, there is still no relevant study on the effect of INT131 on atherosclerosis.

Full agonist-TZDs forms a key hydrogen bond with the side chain of Y473 on helix 12, mainly interacts with residues from arm I in the ligand binding pocket, to enhance the binding affinity of coactivators/weaken corepressors, inducing transcriptional activation (Brust et al., 2018). Unlike full agonist TZDs, some partial agonists have different binding mode, such as GQ-177 interacts through hydrophobic contacts with residues from arm II (Barros et al., 2010), lobeglitazone makes additional hydrophobic contacts with the Ω-pocket. There are also some partial agonists having similar binding sites to TZDs, but the recruiting cofactors were different. For example, INT131 activates PPARγ, but does not recruit the cofactor MED1 (key factor in regulating adipogenesis) (Lee et al., 2012; Xie et al., 2017), S26948 is unable to recruit DRIP205 or PCG-1α (Key genes in regulating gluconeogenesis), so that it selectively reduces blood glucose without the obvious adverse effects of adipogenesis (Carmona et al., 2007; Sohn et al., 2009). Compared to rosiglitazone, Lobeglitazone strongly blocks the phosphorylation of PPARγ at Ser245, but the pharmacological effects of this translational modification change need further studied (Jang et al., 2018). Although the exact mechanism beyond these effects remains to be determined, partial agonist might represent a new class of therapeutic molecules for the treatment of atherosclerosis.

Dual PPARα/γ agonist has been focusing on the activation of both PPARα and PPARγ, which may provide a wider range of metabolic benefits. Studies have shown that most of the Dual PPARα/γ agonist play an active role in anti-atherosclerosis, such as GQ-11 (Silva, 2018), P633H (Chen et al., 2009), C333H (Xu et al., 2006), LT175 (Gilardi et al., 2014), and Tesaglitazar (Zadelaar et al., 2006; Chira et al., 2007; Table 1). GQ-11 improved insulin sensitivity and enhanced Glut4 expression in the adipose tissue, meanwhile the levels of MCP-1were reduced and the levels of IL-10 were increased. Furthermore, it also upregulation of Apoa1 and ABCA1 gene expression, then reduced triglycerides and VLDL cholesterol and increased HDL cholesterol in LDLr^–/– mice (Silva et al., 2018). P633H can be accompanied by the upregulation of ACO and aP2 expression in db/db and KK-A^y mice, and is targeted to regulate PPARα in the liver and PPARγ in adipose tissue, respectively (Chen et al., 2009). C333H efficiently reduced blood lipid and glucose concentration in db/db mice (Xu et al., 2006). LT175 activates PPARγ in adipocytes, increases the expression of PPARγ target gene Glut4 and Adipoq in 3T3-L1 adipocytes and in a mouse model. Moreover, LT175 can also activate PPARα in the liver, trigger triglyceride and fatty acid catabolism, and achieve to eliminate the side effects of some conventional PPAR agonists (Gilardi et al., 2014). However, such effects are not clear in human situation. Tesaglitazar reduces atherosclerosis, reduces macrophage inflammation, number of adhesion monocytes and nuclear factor activity of the vessel wall (Zadelaar et al., 2006). However, some Dual PPARα/γ agonist such as compound 3q may accelerate atherosclerosis, it may related to the increase expression of the vascular endothelial activation and inflammation markers, such as P-selectin, MCP-1, VCAM-1, and CD36, that are also associated with plaque complexity (Calkin et al., 2007). Dual agonists such as Tesaglitazar also showed better insulin sensitization effects as well as the prevention of atherosclerosis progression in clinical studies, but due to adverse side effects, including heart failure and myocardial ischemia, it has been discontinued in phase III clinical trials (Yamaguchi et al., 2014).

The PPARγ agonists shown to increase adipogenesis and body weight, whereas PPARα agonists counteract these effects by decreasing food intake and fat deposits. However, both its PPAR binding mode and its downstream targeting will change accordingly. LT175 impaired the recruitment of CBP coactivator, Tesaglitazar is accounted for by inhibition of both expression and acetylation/deactivation of cardiac PGC1α both in healthy C57BL/6 and diabetic db/db mice. Consistent with other partial agonists, GQ-11 only hydrogen bond with the PPARγ residue Ser289 at helix 3, which could reflect in weak PPARγ agonistic activities, and also interacts and weakly activates PPARα (Silva, 2018). Most of the dual PPARα/γ agonists, although they can improve insulin resistance as the full agonists, and do not have similar weight gain, negative bone effects, but it will appear adverse effects on the urothelial, renal, and cardiovascular system (Kaul et al., 2019; Chen et al., 2021). For the adverse side effects of PPARγ agonists in fluid retention, most studies have shown that this effect was due to increased reabsorption of sodium and water by the renal tubules, but the role of specific renal unit segments and sodium carriers was unclear. PPARγ-induced EGF receptors and non-genomic trans-activation of downstream extracellular signal-modulating kinases (ERKs) may augment sodium reabsorption in the proximal tubule (Beltowski et al., 2013). TZDs-like compounds significantly inhibited PPARγ phosphorylation in Ser112 while telmisartan did not (Kolli et al., 2014). Thus, telmisartan did not have a significant effect on osteoclast differentiation and osteogenesis. Dual PPARα/γ Tesaglitazar activation inhibits SIRT1-PGC1α axis and causes cardiac dysfunction. It was showed that this cardiac dysfunction was associated with reduced PGC1α expression. These effects are related to competition between PPARα and PPARγ to regulate Ppargc1a gene expression and to reduce cardiac SIRT1 expression (Kalliora et al., 2019). PGC1α is a regulator of mitochondrial function in thermogenic tissues, such as brown fat. Lehman et al. (2000) also found cardiac-specific overexpression of PGC1α in mice lead to uncontrolled mitochondrial proliferation in cardiomyocytes, resulting in loss of sarcomere structure and dilated cardiomyopathy. Inhibiting PGC1α on dual PPARα/γ activation is potentially as a key event that mediates the cardiotoxic effect, which would provide a guide for design of future PPAR agonists.

Non-agonist (Antagonists) PPARγ ligand, also known as PPAR modulators, exhibit high affinity but do not activate PPARγ. The Antagonists consists of two main categories, and one is known as PPARγ phosphorylation inhibitors, such as SR1664 (Cariou et al., 2012; Dias et al., 2020) and UHC1 (Choi et al., 2014). Taken SR1664 as an example, it has basically no transcriptional activation effect on PPARγ, but has a high affinity with PPARγ and belongs to a phosphorylation inhibitor, blocking the CDK5-mediated phosphorylation of PPARγ in Ser273 (Laghezza et al., 2018; Dias et al., 2020). SR1664 did not stimulate lipid accumulation or adipogenesis gene expression (such as aP2, Glut4, Lpl, CD36) in differentiating fat cells. UHC1 blocking CKD5-mediated PPARγ phosphorylation at position K395 in LBD, reducing macrophages inflammatory factor LPS-induced nitric oxide (NO) production both in vitro and in HFD-fed mice (Choi et al., 2014; Ribeiro Filho et al., 2019). Comparison with conventional full agonists, these antagonists do not stabilize helix 12 and display negligible changes in activation, but make unfavorable interactions with F282 on helix 3 (Asteian et al., 2015). Another class of PPAR modulators is relatively special, with structurally TZDs analogs, but with little effect on PPARγ, represent the compound as MSDC. It have been showed a potential therapeutic avenue for treating non-alcoholic steatohepatitis, improve the insulin resistance effect, regulate the lipid metabolism (Colca et al., 2018). However, the current study of atherosclerosis has not been reported. Its target of action was reported as a line Mitochondrial pyruvate carrier 2 (MPC2), which is the mitochondrial target of thiazolidinediones (mTOT) (Vigueira et al., 2017). Although not directly stimulated to PPARγ, but still has the potency of insulin sensitization.