Modeling, Synthesis, and Biological Evaluation of Potential Retinoid-X-Receptor (RXR) Selective Agonists: Analogs of 4-[1-(3,5,5,8,8-Pentamethyl-5,6,7,8-tetrahyro-2-naphthyl)ethynyl]benzoic Acid (Bexarotene) and 6-(Ethyl(4-isobutoxy-3-isopropylphenyl)amino)nicotinic Acid (NEt-4IB)

Abstract

Five novel analogs of 6-(ethyl)(4-isobutoxy-3-isopropylphenyl)amino)nicotinic acid-or NEt-4IB-in addition to seven novel analogs of 4-[1-(3,5,5,8,8-pentamethyl-5,6,7,8-tetrahydro-2-naphthyl)ethynyl]benzoic acid (bexarotene) were prepared and evaluated for selective retinoid-X-receptor (RXR) agonism alongside bexarotene (1), a FDA-approved drug for cutaneous T-cell lymphoma (CTCL). Bexarotene treatment elicits side-effects by provoking or disrupting other RXR-dependent pathways. Analogs were assessed by the modeling of binding to RXR and then evaluated in a human cell-based RXR-RXR mammalian-2-hybrid (M2H) system as well as a RXRE-controlled transcriptional system. The analogs were also tested in KMT2A-MLLT3 leukemia cells and the EC₅₀ and IC₅₀ values were determined for these compounds. Moreover, the analogs were assessed for activation of LXR in an LXRE system as drivers of ApoE expression and subsequent use as potential therapeutics in neurodegenerative disorders, and the results revealed that these compounds exerted a range of differential LXR-RXR activation and selectivity. Furthermore, several of the novel analogs in this study exhibited reduced RARE cross-signaling, implying RXR selectivity. These results demonstrate that modification of partial agonists such as NEt-4IB and potent rexinoids such as bexarotene can lead to compounds with improved RXR selectivity, decreased cross-signaling of other RXR-dependent nuclear receptors, increased LXRE-heterodimer selectivity, and enhanced anti-proliferative potential in leukemia cell lines compared to therapeutics such as 1.

Keywords: leukemia; retinoid; retinoid-X-receptor; rexinoid; small molecule therapeutic; structure–activity relationship.

Figure 1

Structures of bexarotene (1), disilabexarotene (2), 9-cis retinoic acid, and all trans retinoic acid.

Figure 2

Structures of reported rexinoids 3–27.

Figure 3

Structures of novel target rexinoids 28–36, 37a, and 37b.

Figure 4

Illustration of AutoDock Vina simulation for RXR binding with bexarotene. (A) Cartoon representation of the human RXR alpha ligand binding domain (PDB:1FBY) in green and the compound bexarotene in orange. N and C terminals are labeled. (B) 2-dimentional depiction of the interactions between protein residue sidechains with bexarotene using PoseView (BioSolvIT). Hydrogen bonds are presented as dashed lines between interaction partners, and hydrophobic interactions are depicted as smooth contour lines.

Scheme 1

Synthesis of 45 and 46 from 38.

Scheme 2

Synthesis of 50, 51, and 52 from 42.

Scheme 3

Synthesis of 53, 54, 55, and 56 from 45, 46, 50, and 51, respectively.

Scheme 4

Synthesis of 58 from 52.

Scheme 5

Synthesis of 25, 28, 29, and 30 from 53, 54, 55, and 56, respectively.

Scheme 6

Synthesis of 31 from 58.

Figure 5

X-ray crystal structure of 31.

Scheme 7

Synthesis of 60 from 59.

Scheme 8

Synthesis of 63 and 64 from 61 and 62, respectively.

Scheme 9

Synthesis of 65 and 66 from 63 and 64, respectively.

Scheme 10

Synthesis of 26 from 66.

Scheme 11

Synthesis of 32 and 33 from 65.

Figure 6

X-ray crystal structure of 32.

Scheme 12

Synthesis of 73, 74, and 75 from 63, 64, and 72 respectively.

Scheme 13

Synthesis of 27 from 74.

Scheme 14

Synthesis of 34 from 73.

Scheme 15

Synthesis of 35 from 77.

Scheme 16

Synthesis of 36 from 75.

Scheme 17

Synthesis of 84 from 80.

Scheme 18

Synthesis of 37a from 84.

Scheme 19

Synthesis of 37b from 84.

Figure 7

Effect of pharmacological targeting of RXRA on KMT2A-MLLT3 proliferation in vitro. (A) EC₅₀ (nM) values for each compound were calculated based on the ratio of GFP+ mCherry+ cells to total mCherry+ cells in UAS-GFP × KMT2A-MLLT3 cells transduced with Gal4-RXRA retrovirus and treated as indicated. ** p < 0.01 compared with 1 (bexarotene) result. (B) GFP activation used to calculate EC₅₀ in (A), each dose evaluated in duplicate. (C) EC₅₀ (nM) values for each compound were calculated based on the luciferase relative intensity of 293T cells transduced with pBABE-RXRA and ApoA1-Luc and treated as indicated. * p < 0.05, ** p < 0.01 compared with 1 (bexarotene) result. (D) Luciferase luminescence results used to calculate EC₅₀ in (C), each dose evaluated in duplicate. (E) UAS-GFP x KMT2A-MLLT3 cells were treated as indicated, replated after 48 h, and total viable cells in 50 µL assessed in duplicate after 96 total hours of treatment. ** p < 0.01 comparing results with and without ATRA (all trans retinoic acid) treatment. Pairwise T-test.

Figure 8

Evaluation of RXR agonists to potentiate LXRE-mediated transactivation in the absence and presence of LXR ligand T0901317. (A) HEK-293 human embryonic cells were transfected with pCMX-hLXRα, an expression vector for human LXRα, an LXRE-luciferase reporter gene with three tandem copies of the LXRE from the human ApoE gene, and a renilla control plasmid. Cells were transfected for 24 h utilizing a liposome-mediated transfection protocol and then treated with the ethanol vehicle, or 100 nM of the indicated compound alone or in combination with 100 nM TO901317 (TO). LXRE-directed activity was compared to compound 1 (bexarotene), set to 100%. (B) The “Heterodimer Specificity Score” (LHS) was determined by calculating the LXRE:RXRE activity ratio of each analog, with compound 1 set to 1.0. Values are means ± SD with indicated analogs exhibiting greater LHS vs. compound 1 (* p < 0.05).

Figure 9

Evaluation of additional RXR agonists to potentiate LXRE-mediated transactivation in the absence and presence of LXR ligand T0901317. (A) Human HEK-293 cells were transfected and treated as described in Figure 8. LXRE-directed activity was compared to compound 1 (bexarotene), set to 100%. Values are means ± SD with indicated analog+TO exhibiting greater activity vs. compound 1+TO (* p, 0.05). (B) The “Heterodimer Specificity Score” (LHS) was determined by the LXRE:RXRE ratio with compound 1 set to 1.0. Values are means ± SD with indicated analogs exhibiting greater LHS vs. compound 1 (* p < 0.05).

Figure 10

Assessment of RXR agonists via a RARE-luciferase reporter based assay in human cells. (A–E) Human embryonic cells (HEK293) were co-transfected with expression vectors for hRXRα, a RARE-luciferase reporter gene, and a renilla control plasmid for 24 h utilizing a liposome-mediated transfection protocol. Cells were treated with bexarotene, analog, or all-trans-retinoic acid (RA) at 10 nM for 24 h. The RARE activity for RA was set to 100%. Values are means ± SD with all analogs tested displaying lowered RARE activity vs. compound 1 (p < 0.05).