LXRα Phosphorylation in Cardiometabolic Disease: Insight From Mouse Models

Abstract

Posttranslational modifications, such as phosphorylation, are a powerful means by which the activity and function of nuclear receptors such as LXRα can be altered. However, despite the established importance of nuclear receptors in maintaining metabolic homeostasis, our understanding of how phosphorylation affects metabolic diseases is limited. The physiological consequences of LXRα phosphorylation have, until recently, been studied only in vitro or nonspecifically in animal models by pharmacologically or genetically altering the enzymes enhancing or inhibiting these modifications. Here we review recent reports on the physiological consequences of modifying LXRα phosphorylation at serine 196 (S196) in cardiometabolic disease, including nonalcoholic fatty liver disease, atherosclerosis, and obesity. A unifying theme from these studies is that LXRα S196 phosphorylation rewires the LXR-modulated transcriptome, which in turn alters physiological response to environmental signals, and that this is largely distinct from the LXR-ligand–dependent action.

Liver X receptors (LXRs) are ligand-activated transcription factors that control lipid and glucose homeostasis, as well as inflammation (1, 2). LXRs consist of 2 subtypes, LXRα and LXRβ (NR1H3 and NR1H2, respectively), which exhibit 75% sequence homology (3). LXRα is expressed in tissues involved in lipid metabolism, such as adipose tissue and liver, as well as in bone marrow and immune cells, whereas LXRβ is ubiquitously expressed (4). Like other nuclear receptors, LXR is organized with an N-terminal activation function, a central DNA-binding domain followed by a flexible hinge region, and a C-terminal the ligand-binding domain (Fig. 1A). LXRs forms obligatory heterodimers with 9-cis retinoid X receptor (RXR) and are activated by desmosterol and oxysterols (2, 5), which are oxidized cholesterol derivatives. LXR can also homodimerize (6); however, the biological significance of this interaction has not been demonstrated. In the absence of a ligand, the LXR/RXR heterodimer occupies LXR response elements (LXREs) on the regulatory regions of target genes (7). Unliganded LXR/RXR interacts with co-repressors, such as the nuclear receptor co-repressor (NCoR) and silencing mediator of retinoic acid and thyroid hormone receptor, thereby silencing gene expression (8). This is also known as ligand-independent active repression. Once the ligand binds to nuclear receptors, their ligand-binding domain undergoes a structural shift to a more stable position, which inhibits co-repressor interaction and creates a high-affinity site for coactivator proteins (9). These coactivator complexes are responsible for the modification of chromatin structure, thereby facilitating the assembly of the general transcriptional machinery to the gene promoter and inducing ligand-dependent transactivation of gene expression (10). LXRs can also repress gene expression through a mechanism known as transrepression (11, 12), by which LXRs antagonize the activity of other signal-dependent transcription factors, such as the nuclear factor-κβ and activator protein 1 (13, 14). It appears this is partly the mechanism through which LXRs exert their anti-inflammatory functions (15, 16). LXRs must be present in the nucleus to regulate target gene transcription, and the LXRα subtype has shown nuclear localization both in the absence and presence of its ligands (17, 18).

Figure 1.

Liver X receptor (LXR)α serine 196 (S196) phosphorylation of the hinge region. A, Schematic of human LXRα showing the N-terminal activation function (AF), DNA-binding domain (DBD), the hinge region, the ligand-binding domain (LBD), and conservation of the LXRα phosphorylation site in mouse on the S196. Phosphorylated serine residues are in red. This site is not conserved in LXRβ. B and C, Molecular modeling of LXRα S198 phosphorylation on the DNA-bound LXRα–retinoid X receptor (RXR)α heterodimer reveals changes in LXRα S198 residue orientation on phosphorylation: conformations of the human LXRα-RXRα heterodimer bound to DNA (multicolored cylinder with central plates) B, in the S198 phosphorylated form (red arrow) and C, in the S198 nonphosphorylated form (green arrow). Residues 181 to 195 loop orientation in the LXRα S198 nonphosphorylated state is altered in the phosphorylated state (C, black arrow). RXRα is shown in blue, and LXRα is depicted in gray. Adapted from (27) with permission.

Most proteins are subjected to posttranslational modifications, which are covalent modifications catalyzed by specific enzymes that result in changes in protein function (19, 20). Posttranslational modifications of nuclear receptors, such as phosphorylation, acetylation, sumoylation, methylation, and O-GlcNAcylation, play important regulatory roles and allow rapid and reversible molecular changes in the nuclear receptor response (21). Phosphorylation (22-24), O-GlcNAcylation (25), and acetylation (26) of LXRα have been shown to influence LXRα-mediated transcriptional activation of target genes through changes in LXRα stability, transcriptional activity (either increasing or decreasing LXRα transcriptional potential), and/or recruitment of transcription factors at specific LXR target genes.

We have demonstrated that altering LXRα phosphorylation modifies its target gene repertoire in vitro in LXRα-overexpressing macrophage-like cell lines (22, 27), and more recently in vivo in the context of cardiometabolic disease models of atherosclerosis progression (28), atherosclerosis regression, and diabetes (29), obesity (30) and nonalcoholic fatty liver disease (NAFLD) (31). In this review, we discuss these recent findings and the potential of targeting LXRα phosphorylation for chronic inflammatory diseases.

Liver X Receptor α Phosphorylation at Serine 198 and Its Kinases

LXRα is phosphorylated at serine 198 (S198) in human LXRα, and at the cognate serine 196 (S196) site in the mouse (22-24). Phosphorylation of this site is observed in mouse and human macrophages in vitro and in atherosclerotic plaque macrophages in apolipoprotein E knockout (Apoe KO) mice (22, 27), as well as in the liver (24, 31). This site is located in the hinge region of LXRα, and is conserved between humans, rats, and mice but absent in LXRβ (Fig. 1A). Phosphorylation at this site is increased by both endogenous and synthetic LXR ligands (24[S],25-epoxycholesterol and T0901317 or GW3965, respectively), and has been reported to be a substrate of casein kinase 2 (CK2) (22) and protein kinase A (24). Phosphorylation at S198 also affects LXRα transcriptional activity in a gene-selective manner. Molecular modeling of the DNA-bound LXRα-RXRα heterodimer (32) reveals alterations in the orientation of the LXRα S198 residue on its phosphorylation. Phosphorylated S198 is exposed on the surface of the complex, consistent with a site for protein interactions (Fig. 1B), whereas the unphosphorylated S198 is buried. However, a segment containing the unphosphorylated S198 adopts a short alpha helical stretch, suggesting an additional surface for protein interaction in this S198 nonphosphorylated state (27) (Fig. 1C). This is consistent with the phosphorylated and unphosphorylated LXRα imparting distinct structural characteristics to the hinge domain to influence cofactor recruitment, and modulating LXRα activity in a gene-specific manner. Given the structural impact and functions effects of LXRα phosphorylation at this site in cell-based models, we examined the relevance of LXRα phosphorylation in mouse models of cardiometabolic disease, which is the focus of this review.

Liver X Receptor α S196 Phosphorylation–Deficient Knockin Mice

To determine the effect of mouse LXRα S196 phosphorylation (pS196) in vivo, we developed a global LXRα S196A knockin mouse carrying a serine-to-alanine mutation (S196A) introduced by site-directed mutagenesis. This mutation prevents phosphorylation at this site but does not significantly affect the level of LXRα expression. The LXRα S196A mice do not present any apparent morphological phenotype, have similar developmental growth compared to matched wild-type (WT) mice, and exhibit equivalent metabolic parameters (31). These mice have been used to determine the effect of LXRα phosphorylation in vivo on various diseases, such as NAFLD (31), atherosclerosis (28), diabetes (29), and obesity (30). Their phenotypes and mechanisms are discussed later and summarized graphically in Fig. 2 and the models, sex of the animals, diets used, and phenotypes are detailed in Table 1.

Figure 2.

Liver X receptor (LXR)α serine-to-alanine mutation at Ser196 (S196A) mouse models in cardiometabolic diseases. Top, left, The full body knock-in of LXRα S196A showed reduced progression from steatosis to steatohepatitis in a model of diet-induced nonalcoholic fatty liver disease (NAFLD) (31). Top, right, A myeloid specific knock-in of LXRα S196A in an atherosclerotic background demonstrated that LXRα S196A increased atherosclerosis via enhanced macrophage proliferation. Bottom, left, and bottom, right, bone marrow transplant models in which all hematopoietic cells express LXRα S196A revealed that LXRα S196A decreased atherosclerosis progression and obesity (30) and reduced atherosclerosis during regression in a diabetic context (29).

Table 1.

Models used in assessing liver X receptor α S196 phosphorylation in cardiometabolic diseases

Disease	Model	CRE	S196A expression	Genetic background	Diet	Sex	Phenotype	Reference
Atherosclerosis	LXRαS196AFL/FL	LYSM-CRE	Monocyte/Macrophage	C57Bl6; Ldlr–/–	WDa	Male	Increased atherosclerosis and macrophages proliferation	Gage et al (28)
NAFLD	LXRαS196AFL/FL	PGK1-CRE	Whole body	C57Bl6	HFHCb	Female	Retards NAFLD progression	Becares et al, (31)
Atherosclerosis and diabetes	LXRαS196AFL/FL	ROSA-CRE-ERT2: BMT	All hematopoietic cells	C57Bl6; Ldlr–/–	WDc	Male	Decreased macrophage adhesion in diabetic plaques	Shrestha et al, (29)
Atherosclerosis and obesity	LXRαS196AFL/FL	ROSA-CRE-ERT2: BMT	All hematopoietic cells	C57Bl6; Ldlr–/–	WDc	Male	Decreased atherosclerosis and obesity	Voisin et al, (30)

Abbreviations: BMT, bone marrow transplant; FL, floxed allele; HFHC, high fat high cholesterol; Ldlr, low-density lipoprotein receptor; LXR, liver X receptor; NAFLD, nonalcoholic fatty liver disease; S196A, serine-to-alanine mutation at Ser196; WD, Western diet.

^aWD = 20% (wt/wt) fat, 0.15% cholesterol.

^bHFHC = 17.2% cocoa butter, 2.8% soybean oil, 1.25% cholesterol, 0.5% sodium cholate.

^cWD = 21% (wt/wt) fat, 0.3% cholesterol.

Liver X Receptor α (LXRα) Phosphorylation in Nonalcoholic Fatty Liver Disease (NAFLD)—LXRα Serine-to-Alanine Mutation at Ser196 Retards NAFLD Progression

NAFLD is a progressive liver disease that ranges from simple fatty liver (steatosis) to nonalcoholic steatohepatitis and can progress to fibrosis, cirrhosis, and hepatocellular carcinoma (33). NAFLD is the major cause of chronic liver disease in the Western world, and is predicted to become the main cause for liver transplantation by 2030 (34). NAFLD may progress from benign steatosis to inflammatory and fibrotic steatohepatitis. LXRs contribute to this pathology by promoting fatty acid and triglyceride accumulation (35). Although deletion of LXRs in mouse models have shed light onto the LXR-mediated pathways controlling steatohepatitis (36, 37), the impact of LXRα phosphorylation on NAFLD progression remained unclear, and we sought to address this gap.

To this end, Becares et al (31) used a global knockin mouse carrying an LXRα homozygous serine-to-alanine mutation at Ser196 (S196A) that impairs LXRα phosphorylation without affecting overall hepatic LXRα levels (LXRα S196A mouse). Because cholesterol induces LXRα phosphorylation (22), it was hypothesized that LXRα S196A mice may respond differently to a high-fat, high-cholesterol (HFHC) diet when compared to chow.

Phenotype

LXRα S196A mice fed an HFHC diet showed higher levels of hepatic nonesterified fatty acids and triglycerides compared to WT mice and exhibited enhanced hepatic steatosis. Livers of the LXRα S196A mice featured increased numbers of larger lipid droplets, associated with enhanced expression of lipid droplet genes. Because plasma nonesterified fatty acids, triglycerides, and insulin levels did not differ between genotypes, the increased hepatic fat accumulation in S196A mice was likely a result of enhanced de novo lipogenesis. In fact, S196A livers showed an increase in the total amount of saturated as well as unsaturated fatty acids, specifically ω9 and certain ω6 fatty-acid species. Thus, LXRα S196A expressed globally induces hepatic steatosis and alters fatty-acid profiles in response to a HFHC diet.

Despite the increased steatosis, HFHC-fed LXRα S196A mice displayed less liver inflammation and collagen deposition than their WT counterparts. This was associated with reduced expression of proinflammatory and profibrotic mediators. Expression of factors involved in the activation of endoplasmic reticulum (ER) stress were reduced in LXRα S196A mice, suggesting these animals are protected from lipotoxicity through a reduction in ER stress. In addition, HFHC-fed LXRα S196A mice were protected from plasma and hepatic cholesterol accumulation. This was accompanied by a 20% reduction in liver weight in LXRα S196A compared to controls and a decrease in the expression of the cholesterol efflux transporter ATP-binding cassette subfamily G member 1. The HFHC-fed LXRα S196A mice also showed a unique induction in the expression of the cholesterol transporter adenosine triphosphate (ATP)-binding cassette subfamily G member 5. Thus, reduced cholesterol accumulation in S196A mice is likely due to increased hepatobiliary secretion of cholesterol.

Mechanism

RNA sequencing (RNA-seq) was performed on livers from LXRα S196A and WT mice fed an HFHC diet to gain insight into the mechanism of steatosis managed by LXRα phosphorylation. This revealed increased expression of a subset of the hepatic enzymes involved in fatty-acid elongation and fatty-acid oxidation. These changes likely contribute to the distinct hepatic fatty-acid profile of S196A mice.

LXRα S196A mice also showed a decrease in the levels of wound healing and fibrotic mediators, including collagen genes and enzymes responsible for collagen stabilization. Moreover, expression of a subset of genes involved in extracellular matrix remodeling and tissue regeneration that distinguish between low-risk and high-risk NAFLD among presymptomatic patients (38) were remarkably different between WT and S196A mice. Thus, LXRα phosphorylation could alter preclinical NAFLD progression by regulating tissue regenerative and remodeling pathways.

Variations in the hepatic transcriptome observed in the liver of LXRα S196A mice compared to WT coincided with changes in chromatin modifications: differences in H3K27 acetylation, which mark active genes, between LXRα WT and S196A mice strongly overlapped with changes in gene expression. For example, expression of the carboxylase 1F (Ces1f) gene was strongly induced in livers from LXRα S196A on an HFHC diet while showing a significant increase in H3K27Ac compared to WT. In silico analysis of the Ces1f gene for putative LXR binding sites revealed a degenerate direct repeat 4 sequence resembling the LXRE consensus binding site (7). In livers from the LXRα S196A HFHC-fed mice, this site was preferentially bound by LXR compared to WT, and was associated with increased RNA polymerase II (Pol II) and phosho-Ser2 Pol II occupancy, indicative of enhanced transcriptional initiation and elongation. This suggests that impaired LXRα pS196 allows for the transcriptional activation of new target genes containing noncanonical direct repeat 4 sequences.

We have previously reported using a macrophage cell line that on LXR ligand activation, phosphorylation affected the transcriptional activity of LXRα by modulating the binding of the NCoR corepressor to genes controlled by LXRα phosphorylation (22). However, differences in NCoR occupancy in the livers of mice exposed to the HFHC diet were not detected, suggesting responses to cholesterol in vivo involve other transcriptional regulatory factors whose interaction with LXRα is sensitive to phosphorylation. One such factor is the transducin β-like 1 X-linked receptor 1 (TBLR1), which participates in nuclear receptor cofactor exchange (39) and modulates LXR target gene expression in hepatic cells (40). TBLR1 was found to preferentially bind to LXRα S196A and occupy the Ces1f LXRE in livers from LXRα S196A as compared to WT HFHC-fed mice. Ces1f is a carboxylesterase that regulates very low-density lipoprotein (LDL) lipid packaging and assembly in the hepatocytes (41), with enhanced expression expected to result in increased hepatic lipid clearance. This suggests changes in LXRα phosphorylation at Ser196 in the context of a cholesterol-rich diet affect its interactions with regulatory factors such as TBLR1 to drive new sets of target genes that influence diet-induced responses in liver.

Liver X Receptor α Phosphorylation in Atherosclerosis

Atherosclerosis is a metabolic and inflammatory disease that begins when cholesterol-rich LDL and other apolipoprotein B–containing lipoproteins in the circulation are deposited in the subendothelial space of large arteries, where they and their modified forms promote inflammatory responses (42). In macrophages, LXRα signaling is critical for the homeostatic response to cellular lipid loading (43). Macrophage uptake of normal and oxidized LDL leads to increased cellular concentrations of cholesterol and oxysterols. Activation of LXRs by oxysterols induces the expression of genes involved in cellular cholesterol trafficking and efflux. However, in the face of persistent high cholesterol (hyperlipidemia), the LXR-mediated cholesterol homeostatic mechanism in macrophages is overwhelmed. This results in their becoming foam cells retained in the subendothelial space and the growth of an atherosclerotic plaque (42). In addition, plaque macrophage content reflects additional parameters including monocyte recruitment, macrophage proliferation and apoptosis, and efferocytosis, the process by which dead cells are removed by phagocytosis. We initially reported an increase in LXRα pS196 in plaque macrophages under inflammatory conditions associated with atherosclerosis progression, and a decrease in LXRα pS196 during the resolution of inflammation in regressing plaques. We also demonstrated in vitro that LXRα pS196 selectively regulates a subset of LXRα target genes in macrophages, including genes with anti-inflammatory and atheroprotective functions (22, 27). However, the specific contribution of LXRα phosphorylation in macrophages and immune cells to atherosclerosis development remained unknown.

Altering liver X receptor α phosphorylation in macrophages during atherosclerosis progression: impact on forkhead box protein M1 expression and macrophage proliferation

Phenotype.

To address the impact of LXRα phosphorylation in macrophages in atherosclerosis, Gage et al (28) generated a mouse model expressing LXRα S196A specifically in myeloid cells on the LDL receptor (LDLR)-deficient (Ldlr-KO) atherosclerotic-prone background (M-S196A^Ldlr-KO). This was accomplished using a conditional LXRα S196A mouse whereby LysM-Cre–mediated recombination of lox sites results in the replacement of the WT LXRα with the nonphosphorylatable S196A allele. The LysM-Cre promotes significant recombination in the monocyte/macrophage lineage, as well as promoting recombination in neutrophils (44).

M-S196A^Ldlr-KO mice fed a Western diet for 12 weeks developed accelerated plaque progression compared to WT^Ldlr-KO controls, shown through increased lipid content measured by en face examination of aortas and plaque coverage in aortic roots. Despite no change in the metabolic characteristics such as systemic glucose or lipid homeostasis, atherosclerotic plaques were larger in M-S196A^Ldlr-KO compared to control mice. This reflected increased proliferation of cells within the plaque (likely macrophages). Smaller necrotic cores and reduced fibrous caps of the M-S196A^Ldlr-KO plaques were also observed.

Mechanism.

To determine the mechanism of increased atherosclerosis in M-S196A^Ldlr-KOmice, the transcriptomic profiles of bone marrow–derived macrophages (BMDMs) from the Western diet-fed mice were compared. This revealed significant genome-wide transcriptional changes and an enrichment in G2/M checkpoint and E2F targets, indicating cell cycle and cell proliferation pathways were induced in macrophages expressing LXRα S196A. This included a 3-fold increase in the proto-oncogene forkhead box protein M1 (FoxM1) and several of its target genes from Western diet-fed LXRα S196A compared to WT LXRα BMDMs. Macrophages expressing the LXRα S196A also showed increased proliferation. Interrogation of chromatin immunoprecipitation sequencing data and in silico analysis identified specific LXRα occupancy at the FoxM1 gene in macrophages. Furthermore, pharmacological inhibition of FoxM1 using the specific inhibitor FDI-6 (45) reduced the LXRα S196A macrophage proliferation and FoxM1 target gene expression, confirming that FoxM1 upregulation mediates the increased macrophage proliferation observed in M-S196A^Ldlr-KO mice. Interrogation of the LXRα S196A-regulated transcriptome also revealed selective expression of several prophagocytic and antiphagocytic molecules predicted to increase efferocytosis, which may contribute to the decreased necrotic core size observed.

To further understand the magnitude of the transcriptional changes imposed by the LXRα phosphorylation disruption, response to an LXR ligand GW3965 was explored through RNA-seq analysis on Western diet–exposed M-196A^Ldlr-KO and WT^Ldlr-KO BMDM. GW ligand activation promoted substantial changes in macrophage gene expression that was different and of a different magnitude in cells expressing the LXRα S196A mutant compared with WT macrophages, highlighting the significance of LXRα pS196 in rewiring the LXRα transcriptome.

Inhibiting liver X receptor α phosphorylation in bone marrow cells during atherosclerosis progression promotes an antiatherogenic phenotype

In addition to macrophages, other immune cells accumulate in the arterial wall and contribute to atherosclerosis (46-48). In a complementary approach to the macrophage specific knockin described earlier, Voisin and colleagues (30) studied the requirement of LXRα phosphorylation on the entire cadre of immune cells expressing LXRα S196A compared to WT using a bone marrow transplantation model into the atherosclerotic-prone Ldlr-KO fed a Western diet to promote plaque formation. This strategy can provide evidence for communication between immune cells types in atherosclerosis that respond to LXRα phosphorylation.

Phenotype.

In contrast to myeloid-specific LXR S196A, plaques from LXRα S196A bone marrow–transplant mice showed reduced lipid accumulation in the plaques and reduced macrophage proliferation. LXRα S196A decreased the recruitment of inflammation-prone Ly6C^high monocytes, which was opposed by a decrease in macrophage apoptosis. Consequently, we observed a small reduction in plaque size and macrophage (CD68+ cells) accumulation. In this model, unlike the myeloid-specific knockin, lipoprotein profiles were affected with an increase in atheroprotective high-density lipoprotein cholesterol (HDL-C) and a decrease in atherogenic LDL-C in LXRα S196A compared to WT. This increase in HDL-C may occur through recruitment of LXRα S196A expressing resident macrophages (Kupffer cells) to the liver in the context of bone marrow transplantation after whole-body irradiation to influence liver cells to either increase HDL-C particle secretion or decrease HDL-C uptake (49, 50).

Mechanism.

To characterize the mechanisms governing macrophage phenotype in atherosclerotic plaques when LXRα phosphorylation at S196 is inhibited in all bone marrow–derived cells, we examined by RNA-seq the gene expression profiles of CD68+ plaque macrophages collected by laser-capture microdissection from LXRα S196A and LXRα WT mice (51). Genes repressed by LXRα S196A aligned to proinflammatory pathways, whereas genes induced by LXRα S196A were associated with mitochondrial function and oxidative phosphorylation pathways but not fatty-acid oxidation. This indicates that mitochondrial activity and energy metabolism are induced in LXRα S196A CD68+ cells relative to WT. These are characteristics of metabolic pathways upregulated in anti-inflammatory macrophages. In parallel, we found that LXRα S196A M1 BMDMs had increased basal respiration, ATP production, fatty-acid metabolism, and mitochondrial abundance. Because the metabolic dependence of macrophages on oxidative phosphorylation is a hallmark of the anti-inflammatory response by M2 macrophages (52), our findings indicate that inhibiting LXRα phosphorylation at S196 in CD68+ plaque macrophages promotes the acquisition of an anti-inflammatory metabolic phenotype.

The difference in phenotype between the myeloid-specific expression of S196A described earlier vs the bone marrow–transplant model, in which the entire complement of hematopoietic cells express LXRα S196A, suggests that LXRα S196A expression within additional bone marrow–derived cells restrains macrophage proliferation in the plaque to deter atherosclerosis. Recent advances in single-cell RNA-seq revealed different leukocyte clusters, including T cells, B cells, dendritic cells, macrophages, monocytes, and natural killer cells, are present in the plaques of Ldlr-KO mice fed a Western diet (53). Elucidating which cell subset influences macrophage proliferation within the plaques of LXRα S196A mice remains an interesting and open question.

Liver X Receptor α Phosphorylation in Cardiometabolic Diseases

Disrupting liver X receptor α phosphorylation in bone marrow during atherosclerosis progression protects against diet-induced obesity

Obesity shares pathological features with atherosclerosis, including chronic inflammation mediated both by innate and adaptive immune cells that accumulate in the visceral adipose tissue and contribute to adipocyte dysfunction (54). Given the pivotal role of LXRα in cholesterol metabolism and inflammation and the influence of LXRα pS196 on gene expression, strategies that reduce LXRα pS196 could help diminish obesity. To determine the effect of LXRα pS196 on diet-induced obesity, we took advantage of the bone marrow–transplant model of LXRα S196A in Ldlr-KO background (30).

Phenotype.

Ldlr-KO mice reconstituted with LXRα WT or S196A bone marrow fed a Western diet become obese. Interestingly, LXRα S196A mice gained less weight (15% less than WT mice) after 16 weeks of the Western diet. Despite equivalent food intake, bone mineral density, and lean body mass, LXRα S196A mice showed reduced fat mass, in particular perigonadal white adipose tissue mass (pWAT), compared to LXRα WT mice. Consistent with the reduced lipid accumulation in perigonadal adipose tissue, we also observed a decrease in adipocyte size when LXRα S196 phosphorylation was inhibited.

Mechanism.

Associated with the reduction of adipose tissue, flow cytometry of the perigonadal adipose tissue revealed that LXRα S196A altered a unique type of immune cell population. Total adipose tissue macrophages (ATMs) and recruited proinflammatory triple-positive F4/80+CD11b+CD11c+ (FBC) macrophages (M1-like macrophages) were reduced, although the number of resident anti-inflammatory double-positive F4/80+CD11b+ CD11c– (FB) macrophages (M2-like macrophages) did not change. This resulted in a higher ratio of inflammation-resolving FB to inflammation-promoting FBC macrophages in LXRα S196A adipose tissue compared to LXRα WT (55). Dendritic cell and T-cell numbers were not affected by LXRα S196A. RNA-seq analysis of ATMs revealed a large change in the LXRα S196A transcriptomes between ATM subtypes consistent with a less-inflammatory phenotype, suggestive of reduced inflammatory-mediated signaling to adipocytes.

Thus, expression of LXRα S196A in the bone marrow significantly decreased mouse weight, total fat, pWAT, and adipocyte size in pWAT compared to WT. This phenotype was not reported in myeloid-specific S196A expressing mice in the Ldlr-KO background, suggesting that other bone marrow–derived cells in combination with macrophages are promoting this phenotype (30).

Disrupting liver X receptor α phosphorylation alters atherosclerosis regression associated with diabetes

Diabetic patients have earlier onset and more extensive atherosclerosis than nondiabetic patients (56, 57). Clinical data show that diabetes significantly hampers the regression of human atherosclerosis, even when lipid levels are lowered (58). This effect has been recapitulated in a mouse model of atherosclerosis in which the favorable effects of lowering lipid levels on plaque development are impaired in diabetic mice (59). Plaque regression—the ability to resolve inflammation on lipid lowering—is also significantly impaired in diabetic mice compared to nondiabetic mice, even when LDL-C is similarly reduced in both (59). We have demonstrated that a chronically high level of glucose alters LXR-dependent gene activation in BMDM, suggesting that LXR function is dysregulated by hyperglycemia in diabetes (60). Given that phosphorylation of LXRα at S196 was increased by diabetes-relevant high glucose levels in a macrophage cell line and also modulated LXRα-gene expression, we examined whether LXRα function in atherosclerosis regression is impaired in diabetic mice as a function of pS196. To determine whether LXRα phosphorylation affected macrophage function during atherosclerosis regression in diabetes, we used an inducible LXRα pS196A-deficient knockin mouse so plaques could develop normally under LXRα WT expression and then be switched to LXRα S196A to isolate the effect of LXR phosphorylation on regression. Mice homozygous for LXRα S196A floxed allele (S196A^FL/FL) with or without a tamoxifen-inducible CRE recombinase (RosaCreER^T2;Cre+ or Cre–) were used as donors for bone marrow transplantation into recipient Reversa mice (Ldlr^−/−; ApoB^100/100; microsomal triglyceride transfer protein [Mttp^fl/fl]; Mx1-Cre^+/+) (61). This is an Ldlr-KO model in which the hyperlipidemia can be reversed after conditional inactivation of microsomal triglyceride transfer protein. The mice were fed a Western diet for 16 weeks to allow plaques to develop. Tamoxifen was administered at week 14.5 to all groups. This induced recombination of LoxP sites in the Cre-expressing mice, resulting in the replacement of LXRα WT with the LXRα S196A allele after the development of atherosclerosis. The regression/diabetic mice were administered an intraperitoneal injection of streptozotocin for 5 days to induce hyperglycemia (29).

Phenotype.

Expression of LXRα S196A in hematopoietic cells had differential effects on atherosclerosis regression in normoglycemic vs hyperglycemic conditions. Under normoglycemia, LXRα S196A expression increased macrophage accumulation, macrophage apoptosis, and plaque necrotic area. In diabetes, LXRα S196A reduced macrophage retention in the plaque, which would be predicted to be antiatherogenic and enhance plaque regression. However, this favorable effect on regression was negated by increased monocyte infiltration in the plaque, attributed to increased leukocytosis in LXRα S196A mice.

Mechanism.

The increase in macrophage plaque accumulation under normoglycemia observed in LXRα S196A mice is due to an increase in the number of circulating white blood cells, especially monocytes. In fact, elevated white blood cell counts have been linked to an increased incidence of coronary heart disease in humans (62). Monocytosis occurs in hypercholesterolemic animal models as a result of Western diet feeding, which promotes accumulation of macrophages in the plaque (63). Plaques from LXRα S196A mice also showed an increased necrotic area and higher numbers of apoptotic macrophages compared to WT. This is likely the result of a defect in efferocytosis because plaque macrophages expressing LXRα S196A showed reduced expression of the efferocytosis promoting factor Mertk, a known LXR target gene (64).

Although LXRα S196A reduced macrophage retention in plaques in diabetes, this favorable effect on regression is masked by increased monocyte infiltration in the plaque attributed to increased leukocyte production in LXRα S196A mice. Thus, in the absence of leukocytosis, the phosphorylation-deficient LXRα S196 has atheroprotective effects in diabetes by reducing the retention and accumulation of macrophages in atherosclerotic plaque (29).

Liver X Receptor α Serine 196 Phosphorylation as a Potential Drug Target

LXRα represents a promising target for the treatment of diseases involving dysregulation of metabolism and inflammation given that its activity can be controlled both by small molecule ligands and phosphorylation. This includes atherosclerosis (15), diabetes (65), obesity (36), and fatty liver disease (37, 66), as well as cancer (67, 68), and Alzheimer disease (35). However, undesirable side effects of targeting LXR using ligands—such as elevated liver triglycerides via LXR-dependent activation of sterol regulatory-element binding proteins—have hindered the clinical utility of compounds targeting LXRα (69-71). Therefore, finding ways to selectively activate (or repress) LXRα targets genes via changes in LXRα phosphorylation could represent a new therapeutic strategy because alterations in LXRα phosphorylation elicit different patterns of gene expression. This is consistent with the finding that phosphorylation of peroxisome proliferator-activated receptor γ at S273 in the hinge region by cyclin-dependent kinase 5 changes the expression of specific genes involved in insulin sensitivity (72). In fact, peroxisome proliferator-activated receptor γ ligands that block cyclin-dependent kinase 5–mediated phosphorylation of the receptor have potent anti-diabetic activity through selective changes in gene expression, suggesting the possibility of novel pharmacology achieved by modulating phosphorylation of nuclear receptors (72, 73). We have reported that we can stimulate the nonphosphorylated form of LXRα pharmacologically by blocking the upstream kinase (eg, CK2) or using a combination of LXRα ligands, such as T0901317 and 9cRA (22). To circumvent systemic side effects of LXR activation, site-specific delivery of LXRα ligands and/or kinase inhibitors that block phosphorylation could be accomplished using nanoparticles. Nanomedicine using LXRα ligands has demonstrated success in treating atherosclerosis in a mouse model while minimizing systemic effects in the liver (74). However, a better understanding of the complex activities of LXRα pS196 in pathological contexts, especially in models that recapitulate human pathophysiology, is necessary to develop alternative targeted therapies.

Conclusions and Future Directions

We have shown that LXRα phosphorylation at S196 plays a key role in the control of LXRα transcriptional activity in cardiometabolic disease. One overarching theme from these studies is that pS196 reprograms the LXR-modulated transcriptome, and this appears distinct from the LXR transcriptome in response to ligand (28, 31). For example, Gage et al found that the majority of the genes in the macrophage transcriptional response that were sensitive to the LXRα phosphomutant were observed in the absence of LXR ligand stimulation (28). This suggest that environmental cues such as diet are perceived by LXR phosphorylation. These findings are also supported by work from Ramón-Vázquez and colleagues in immortalized BMDMs from LXR-deficient mice reconstituted with either LXRα or LXRβ that found most of the gene expression differences were seen when receptors were expressed rather than with pharmacological activation of LXR (75). In addition, comparison of gene expression profiles between macrophages from the plaque and macrophages from the visceral adipose tissue from the same mouse were nonoverlapping (30), highlighting the role of the local tissue microenvironment in determining gene expression managed by LXRα.

Our studies have established the physiological relevance of LXRα phosphorylation in cardiometabolic diseases, but several questions remain. For example, which kinase(s), CK2 or protein kinase A, phosphorylate LXRα in vivo, and is this dependent on the diet or cell type? To address these questions, we could use genetics and pharmacological manipulation via kinase conditional knockout mice, and kinase inhibitors. Another remaining question is how does LXRα S196A induce such a dramatic change in transcription? LXRα S196A can selectively occupy certain genes via preferential interaction with coregulators (22, 31). This reinforces work on LXRβ phosphorylation at S426 by calcium/calmodulin-dependent protein kinase II γ, leading to the derepression of inflammatory genes via the release of coronin 2, a component of the NCoR repression complex (76). But are these mechanisms the exception or the rule? To gain insight into this question, chromatin immunoprecipitation sequencing studies of LXRα WT and LXRα S196A on dietary changes, as well as characterization of the LXRα S196 phosphocistrome using LXRα phosphospecific antibody in macrophages and liver, could inform LXRα phosphorylation-dependent gene regulatory mechanisms by identifying potential collaborating transcription factors.

We are also interested in applying the LXRα S196A mouse model to other diseases linked to LXRα signaling, including neurodegenerative diseases such as Alzheimer (77). Additionally, an atlas of LXRα phosphorylation in normal and diseased human tissues would help inform homeostatic or pathological states in which LXRα phosphorylation could be relevant. Together, such studies would reveal mechanisms underlying LXRα phosphorylation-dependent gene expression, and inform physiological settings in which LXRα phosphorylation is important.

Glossary

Abbreviations

ATM

adipose tissue macrophage

ATP

adenosine triphosphate

BMDM

bone marrow–derived macrophage

Ces1f

carboxylase 1F

CK2

casein kinase 2

FoxM1

forkhead box protein M1

HDL-C

high-density lipoprotein cholesterol

HFHC

high-fat, high-cholesterol

LDL

low-density lipoprotein

LDLR

LDL receptor

LXR

liver X receptor

LXRE

LXR response element

M-S196A^Ldlr-KO

mouse model expressing LXRα S196A in myeloid cells on LDLR-deficient atherosclerotic-prone background

NAFLD

nonalcoholic fatty liver disease

NCoR

nuclear receptor co-repressor

Pol II

RNA polymerase II

pS196

S196 phosphorylation

pWAT

perigonadal white adipose tissue mass

RNA-seq

RNA sequencing

RXR

retinoid X receptor

S196

serine 196

S196A

serine-to-alanine mutation at Ser196

S198

serine198

TBLR1

transducin β-like 1 X-linked receptor 1

WT

wild-type

Additional Information

Disclosure Summary: The authors have nothing to declare.

Data Availability. Data sharing is not applicable to this article as no datasets were generated or analyzed during the current study.