Liver X Receptor α
Created by William Walsh
Cholesterol (5-Cholesten-3β-ol) is vital to the human body and is only synthesized by animals (1). As a precursor for steroid and cortisone-like hormones as well as bile acids, cholesterol serves a variety of functions in the human body (1). In fact, cholesterol is one of the main components of cell membranes and is a major component of the phospholipid bilayer (1).
While cholesterol is vital to animals, it is important for cells to tightly regulate cholesterol levels because excess cholesterol is toxic to cells and organ systems (2). High cholesterol levels (LDL) are associated with atherosclerosis, which increase the chance of heart attack or stroke (16). Since "heart disease is the number one killer of women and men in the United States" the mechanisms of cholesterol regulation have been well studied and methods to reduce cholesterol level are tirelessly pursued (16).
Cholesterol homeostasis within animal cells is maintained through transcriptional regulation of cholesterologenic enzymes and cholesterol removal proteins (2)(3). The nuclear receptor responsible for transcriptional regulation of these proteins is Liver X Receptor α (LXRα). LXRα activation by oxysterols (cholesterol metabolites) inhibits intestinal absorption of cholesterol, stimulates removal from cells, activates the conversion of cholesterol to bile acids, and activates de novo fatty acid synthesis (3)(Image 2 & 3). Because cholesterol and oxysterol levels are directly proportional, activation of LXRα by oxysterols leads to the repression of cholesterologenic enzyme gene expression and the activation of cholesterol disposal gene expression (2)(3). 
LXRα is a 447 amino acid protein weighing 50.4 kD (PDB= 1UHL) with an isoelectric point at pH 7.61 (5)(6). The protein’s secondary structure contains twelve α-helices and two β-sheets (7). The Ramachandran plot of LXRα shows the majority of the values in the (-60, -45) region of the graph which confirms the predominately α-helical structure (Image 4). Composed of four domains, LXRα contains “an N-terminal ligand-independent activation function domain which may stimulate transcription in the absence of a ligand, 2) a DNA-binding domain containing two zinc fingers, 3) a hydrophobic ligand-binding domain required for ligand binding and receptor dimerization, and 4) a C-terminal ligand-dependent transactivation sequence, which stimulates transcription in response to ligand binding” (3). The ligand-binding domain (LBD) is the region crystallized by researchers to study as a site for drug therapy.
The LBD of LXRα is a hydrophobic pocket made up of multiple residues essential for oxysterol and synthetic agonist binding (Image 5). The intermolecular forces of W-443 and H-421 with the 22-, 24-, 27- hydroxyl/epoxy group of oxysterols are pivotal to ligand induced conformational change (7). The hydrogen bonding between the 3-hydroxyl group of oxysterols to R-305 and E-267 is key for stabilization of the active conformation (7). As shown, R305 and E267 do not interact with the synthetic agonist but form a salt bridge instead(7).
Agonist binding in the LBD induces a conformational change in the C-terminal helix 12 (AF2), which forms a "lid over the ligand-binding pocket" in the active state (7). The movement of AF2 over the ligand binding pocket leads to formation of a hydrophobic groove by helix 3, helix 4, and helix 12(7). This hydrophobic groove is the site of coactivator binding which allows for stronger gene regulation (7). Composed of multiple LXXLL motifs, the coactivator-LXRα pocket interacts with the LXR-response element of genes (8). The LXR-response element is a tandem repeat of AG(G/T)TCA separated by four nucleotides (9).
Biologically, LXRα exists as a heterodimer with the retinoid x receptor (RXR) but comparative analysis determined that the coactivator-binding surface of LXR is critically important to interaction with the coactivor protein and confirmed that RXR is an allosteric activator of SRC-1-LXR interaction (10). Therefore, LXRα can be activated through three possible mechanisms: activation by dimerization, dual-ligand permissiveness (ligand binding of LXR or RXR), and ligand-dependent allosteric effect (ligand binding of LXR and RXR).
Although LXRα acts as the cholesterol sensor, RXRβ is a necessary component of the functional protein and should be analyzed. RXRβ is a 533 amino acid long, 57 kD nuclear receptor with isoelectric point of 8.52 and a secondary structure similar to LXRα (13). The ligand-binding pocket of RXR is mostly hydrophobic and made up of helices 3,7 and 11 with two key residues in the active site. The carboxylate of the methoprene acid (MPA) is coordinated through ion-ion interaction with R387 while N-377 regulates the size of the binding pocket (7).
The dimerization of LXRα/RXRβ takes place along the mostly hydrophobic interface of helix 9 and helix 10 of each protein. While the hydrophobic core is a substantial component of the interface the H383, E387 and H390 of LXRα form salt bridges with E465, A469, and E472 of RXRβ(7).
As activation of LXRα leads to increased fatty acid synthesis, fatty acids are natural inhibitors of LXRα, especially polyunsaturated fatty acids (7)(11). One of the strongest LXR antagonists, arachidonic acid, competes for the LBD of LXRα by interacting with R305, which is believed to coordinate the carboxylate group of fatty acids through hydrogen bonding (7). An E267A mutation in LXRα led to stronger inhibition by arachidonic acid, suggesting that the loss of glutamate allows for enhanced arginine-carboxylate interaction (7). This arginine residue is conserved in other nuclear receptors, specifically retinoic acid receptor (RXR) and liver x receptor β (LXRβ) to coordinate the carboxylate of retinoic acid (7).
Another inhibitor, 22(S)-hydroxy-cholesterol, contains a hydroxyl group, which fits the ligand-binding pocket as well as natural agonists but antagonizes LXRα through interaction with Q424 instead of H421 (7). This finding confirms the importance of H421 and W443 for inducing the correct conformational change for coactivator binding (7).
In addition to negative feedback inhibition, LXRα may also be regulated by covalent modification. Phosphorylationof the 195-196 serines or 290-291 threonine or serine results in inactivation of LXRα by conformational change (12). Activation of proteinkinase A (PKA) by cyclic AMP is responsible for phosphorylation of LXRα.
LXRα belongs to a larger family of nuclear receptors and has many conserved structural components. LXRβ (PDB=1PQC),an isoform of LXRα, shares nearly 80% of its amino acid sequence with LXRα (3). BLAST results for LXRβ showed aquery coverage of 92% and an E value of 2E-148. While LXRα and LXRβ share asimilar affinity for oxysterols, it is the location of LXRα expression that has made it the focus of research. LXRα is expressed in the liver, intestine, kidney, spleen and adipose tissue while LXRβ is expressed at lower levels throughout the body with less tissue specificity (3). Another protein that showed strong similarities with LXRα in primary structure was the bile acid receptor FXR (PDBID= 3L1B), which had an 85% query coverage and an E value of 8E-72 (14).
After comparison of primary structure, the DALI Structural Alignment server was used to compare the tertiary structure of LXRα with other proteins. The LXRα isoform LXRβ showed striking similarity as expected. The RXRβ (PDB=1UHL) showed even more similarity to LXRα than LXRβ with a Z score of 40.3 and an rmsd of 0 (15). The bile acid receptor FXR results were not as similar as expected (Z score 28.0 and rmsd 1.7) but superimposition showed very similar tertiary structure except for the lack of β-sheets in FXR compared to LXRα. The similarities between LXRα, LXRβ, FXR and RXRβ is logical because all four bind hydrophobic ligands with oxygen containing head groups and bind to DNA. The similarity in structure correlates to a similarity in function as nuclear receptors for gene regulation.