Ligand flexibility and binding pocket malleability cooperate to allow selective PXR activation by derivatives of a promiscuous nuclear receptor ligand

SUMMARY

The human nuclear receptor (NR) family of transcription factors contains 48 proteins that bind lipophilic molecules. Approved NR therapies have had immense success treating various diseases, but lack of selectivity has hindered efforts to therapeutically target the majority of NRs due to unpredictable off-target effects. The synthetic ligand T0901317 was originally discovered as a potent agonist of liver X receptors (LXRα/β) but subsequently found to target additional NRs, with activation of pregnane X receptor (PXR) being as potent as that of LXRs. We previously showed that directed rigidity reduces PXR binding by T0901317 derivatives through unfavorable protein remodeling. Here, we use a similar approach to achieve selectivity for PXR over other T0901317-targeted NRs. One molecule, SJPYT-318, accomplishes selectivity by favorably utilizing PXR’s flexible binding pocket and surprisingly binding in a new mode distinct from the parental T0901317. Our work provides a structure-guided framework to achieve NR selectivity from promiscuous compounds.

Graphical Abstract

eTOC Blurb

Huber et al. use structure-guided design to develop a selective PXR agonist from a classically known promiscuous nuclear receptor ligand. The newly derived analog binds PXR in a different pose compared to the parental molecule, highlighting the use of PXR’s flexible binding pocket to recognize ligands.

INTRODUCTION

The nuclear receptor (NR) superfamily is a group of ligand-activated transcription factors with diverse roles in reproduction, development, physiology, and disease.¹ There are 48 NRs encoded in the human genome that regulate transcription of various target genes,² and alternative splicing further expands NR functionality.³ NRs share a common modular structure consisting of a variable N-terminal domain, a DNA-binding domain (DBD), and a ligand-binding domain (LBD). An activation function 2 (AF-2) helix (alpha helix 12, α12) on the LBD C-terminus regulates ligand-dependent cofactor interactions and receptor activity.⁴ The LBD is responsible for ligand recognition and resulting cofactor interactions through ligand-modulated α12 conformations.⁵ Agonists induce receptor association with coactivators, such as steroid receptor coactivator 1 (SRC-1), and inverse agonists stabilize receptor complexes with nuclear receptor corepressors like NCOR1 and NCOR2.⁶ Thus, the LBD dictates transcriptional transactivation activity of NRs while the DBD determines target gene specificity. The variety of DNA-binding sequences and the propensity of different NRs to either homodimerize (e.g., androgen receptor) or heterodimerize with retinoid X receptors results in a diverse set of transcriptional targets² including proteases (e.g., KLK3),⁷ regulators of inflammation (e.g., ANXA1),8 drug metabolism enzymes (e.g., CYP3A4),^9-12 and transcription factors (e.g., FOS).¹³

Most of the >1,600 transcription factors are typically difficult to develop drugs toward due to their disordered structures, but NRs are uniquely druggable because of their dedicated ligand-binding pockets.^14,15 These defined pockets combined with extensively characterized natural and synthetic ligands and broad biological roles have made NRs one of the most highly therapeutically targeted protein classes. However, drugs are only approved for approximately one-third of the 48 NRs.^16,17 A recurring challenge in drug development campaigns is multi-NR modulation by a single molecule. NRs are activated by an array of lipophilic molecules such as steroid hormones, vitamins, fatty acids, and xenobiotics,¹⁸ and there is considerable overlap of NR modulation by individual ligands. Both activation and inhibition of specific receptors can be advantageous in context-dependent manners, so ligand specificity is key to achieving a desired biological impact without adverse nonspecific effects.

The synthetic compound T0901317 is a representative promiscuous NR ligand that exhibits a range of agonistic and inverse agonistic activities on various NRs. First identified as a highly potent and efficacious agonist of the liver X receptors (LXRα and LXRβ),¹⁹ T0901317 was later found to modulate activities of multiple NRs, with activation of pregnane X receptor (PXR) being as robust as that of the LXRs.^20,21 PXR agonists have been proposed as potential treatments for inflammatory bowel disease (IBD),²² but lack of specificity of PXR agonists has prevented more meaningful studies and additional development. For example, LXR activation causes liver lipogenesis as well as additional systemic effects due to the more widespread expression of LXR compared to PXR, whose expression is generally limited to liver and intestine.^23,24 The therapeutic mechanism of the IBD drug rifaximin has at least partly been attributed to PXR activation, but the extent of PXR’s role is unknown because rifaximin is also a potent antibiotic that alters gut microbiota makeup.^25,26 Therefore, enhanced specificity of PXR agonists may be desirable for more effective IBD treatments. On the other hand, small molecule-mediated PXR activation might lead to drug-drug interactions and decreased drug efficacy, indicating that, like with other receptors, both activation and inhibition of PXR can be desirable in different contexts.

We previously used T0901317 as a case study chemical to reduce PXR binding by targeted expansion of the ligand binding pocket with N-position extensions of T0901317.²⁷ PXR has a large pocket volume (~1,200 ų) that can expand to >1,500 ų to incorporate large ligands,²⁸ although use of the expansion mechanism is unfavorable to binding affinity and subsequent activation.²⁷ Comparatively, LXRs have a small binding pocket volume of ~700-850 ų,^29,30 and we reasoned that N-position extensions of T0901317 would lose LXR activation disproportionately to PXR activation. Therefore, in this study, we expanded our T0901317 analog series and identified two molecules, SJPYT-305 and SJPYT-318, that are highly selective PXR agonists. Interestingly, while SJPYT-318 has a larger extension than SJPYT-319 (previously published as T0-BP),²⁷ SJPYT-318 more potently activated PXR than SJPYT-319. To investigate this unexpected observation, we solved the crystal structure of PXR LBD bound to SJPYT-318 and found that the N-position substituent was rotated ~180° compared to SJPYT-319 and T0901317, forming a network of new interactions that included stabilizing hydrophobic contacts with α12. These results highlight PXR’s ability to mold its ligand binding pocket to diverse ligand shapes and demonstrate that PXR’s flexibility can be utilized to develop PXR-selective ligands from a promiscuous initial scaffold. In contrast, ligand flexibility should be avoided when taking the approach of increasing ligand bulkiness to reduce PXR binding affinity.

RESULTS

Identification of PXR-selective T0901317 derivatives

Using a limited set of compounds, we previously showed that N-position extensions of T0901317 reduce PXR binding potency and activation potential in a stepwise manner.²⁷ Because T0901317 was originally reported as a potent LXR agonist¹⁹ and is known to be a promiscuous NR ligand,²⁰ we expanded our T0901317 analog series (Figure 1) and tested compound activities on PXR, LXRα, and LXRβ. We used a time-resolved fluorescence resonance energy transfer (TR-FRET) assay to measure recruitment of a 23-amino acid coactivator peptide (SRC-1) to purified NR LBDs (Figure 2A-C). To accurately reflect both potency (half maximal effective concentration, EC₅₀) and efficacy (maximal effect), we calculated area under the curve (AUC) for each compound (Table 1). We then analyzed the values along a linear gradient to detect PXR-selective molecules (Figure 2D). The large substitution in SJPYT-320 resulted in a complete loss of activity for all receptors, and we identified two analogs, SJPYT-305 and SJPYT-318, that exhibited selectivity for PXR LBD over LXR LBDs for recruiting SRC-1. Interestingly, the selectivity occurred in different ways for the two compounds. SJPYT-305 had more efficient SRC-1 recruitment than T0901317 for PXR LBD but became less efficient than T0901317 for LXR LBDs, thus resulting in PXR selectivity. SJPYT-318, however, was less efficient than T0901317 for all proteins but had a more pronounced activity loss for LXR LBDs than for PXR LBD.

Figure 1. Chemical structures of the studied compounds.

Figure 2. Identification of PXR-selective T0901317 derivatives.

(A) Schematic of TR-FRET assay used to measure recruitment of a 23-amino acid coactivator peptide (SRC-1) to purified NR LBDs. (B-C) The indicated compounds were assayed for SRC-1 recruitment to the LBDs of PXR (left panels), LXRα (middle panels), or LXRβ (right panels). Results are expressed as the mean ± standard deviation from four replicates. (D) The area under the curve (AUC) was calculated for each of the curves in (B-C), where larger AUC indicates more SRC-1 recruitment. See also Figure S3.

Table 1. SRC-1 recruitment to NR LBDs.

TR-FRET assays were performed with purified GST-tagged NR LBDs and a fluorescently labeled 23-amino acid SRC-1 peptide. ¹AUC, area under curve. ²SD, standard deviation. ³FC, fold change compared to T0901317.

Compound	PXR		LXRα		LXRβ
	AUC1 ± SD2	FC3	AUC ± SD	FC	AUC ± SD	FC
T0901317	124.0 ± 4.6	1	122.4 ± 2.6	1	201.5 ± 3.3	1
SJPYT-299	101.8 ± 15.7	0.82	67.1 ± 2.9	0.55	133.1 ± 1.4	0.66
SJPYT-300	102.2 ± 5.8	0.82	80.2 ± 1.9	0.66	136.4 ± 3.4	0.68
SJPYT-301	128.8 ± 12.9	1.04	95.1 ± 1.8	0.78	154.1 ± 2.3	0.76
SJPYT-302	101.9 ± 8.1	0.82	108.1 ± 4.4	0.88	148.9 ± 5.4	0.74
SJPYT-303	134.1 ± 3.1	1.08	63.6 ± 1.9	0.52	119.6 ± 2.4	0.59
SJPYT-304	112.6 ± 8.2	0.91	59.5 ± 1.8	0.49	124.9 ± 3.6	0.62
SJPYT-305	161.4 ± 6.4	1.30	33.4 ± 0.5	0.27	76.1 ± 2.3	0.38
SJPYT-315	57.4 ± 2.7	0.46	62.3 ± 0.4	0.51	117.1 ± 2.4	0.58
SJPYT-316	38.7 ± 2.2	0.31	60.8 ± 0.7	0.50	143.8 ± 3.6	0.71
SJPYT-317	39.5 ± 1.5	0.32	27.2 ± 1.4	0.22	71.3 ± 2.8	0.35
SJPYT-318	47.1 ± 2.4	0.38	5.0 ± 0.3	0.04	21.8 ± 1.1	0.11
SJPYT-319	23.6 ± 0.2	0.19	7.3 ± 0.2	0.06	21.5 ± 0.9	0.11
SJPYT-320	5.0 ± 1.2	0.04	2.5 ± 0.4	0.02	6.5 ± 1.1	0.03

PXR selectivity of molecules is accentuated in a cellular environment

Positive results in the SRC-1 recruitment assay indicated that most analogs may be agonists of PXR and LXRα/β to varying degrees. To test if SJPYT-305 and SJPYT-318 activate PXR and maintain their PXR selectivity in a cellular context, we performed activity assays in 293T cells using NR LBDs fused to the GAL4 DNA binding domain and a reporter under the control of a minimal promoter containing GAL4 binding sites (Figure 3A). Rifampicin and GW3965 were used as PXR-specific and LXR-specific agonists, respectively (Figure S1). To find a concentration range suitable for cellular experiments, we treated 293T cells with compounds and found that concentrations above 1 μM consistently reduced cell viability (Figure S2); therefore, 1 μM was chosen as the upper cutoff for T0901317 analog treatments in the cellular reporter assays.

Figure 3. PXR selectivity of molecules is accentuated in a cellular environment.

(A) Schematic of NR reporter activity assays. An NR LBD is fused to the yeast GAL4 DNA binding domain, which binds to a reporter plasmid. Expression of the reporter (firefly luciferase for PXR or beta-lactamase for LXRs) is increased in the presence of an agonist. (B-C) The indicated compounds were assayed for activity of PXR (left panels), LXRα (middle panels), or LXRβ (right panels). Results are expressed as the mean ± standard deviation from three replicates. (D) AUC was calculated for each of the curves in (B-C), where larger AUC indicates more NR activity. See also Figures S1-S3.

Reporter activity was measured after 24 h compound treatment, and activities were analyzed in the same manner as the SRC-1 recruitment assays (Figure 3B-D, Table 2, and Figure S3). The compound efficiencies for reporter activation were largely equal to those for SRC-1 recruitment, with a slight preference for cellular reporter activity over SRC-1 recruitment for PXR and LXRα and a marginal preference toward SRC-1 recruitment for LXRβ (Table 1, Table 2, and Figure S3). As expected based on our previous observations,²⁷ activation efficiency was roughly proportional to ligand size. Remarkably, however, despite the increased size of SJPYT-318 compared to SJPYT-319, SJPYT-318 had substantially enhanced PXR activation over SJPYT-319 (Figure 3C), and this observation was consistent with SRC-1 recruitment (Figure 2C). Furthermore, SJPYT-318 was strongly selective for PXR in cellular reporter activation, showing only marginal activation of both LXRα and LXRβ (Figure 3D). Because T0901317 is also a farnesoid X receptor (FXR) agonist and RAR-related orphan receptor gamma (RORγ) inverse agonist, we tested SJPYT-305 and SJPYT-318 and found that neither compound modulated these receptors (Figure 4A-B). The compounds also did not modulate additional receptors that have large ligand binding pockets, such as peroxisome proliferator-activated receptors delta and gamma (PPARδ and PPARγ) (Figure 4C-D). Thus, these molecules, particularly SJPYT-318, are highly selective PXR agonists.

Table 2. Cellular NR activation by T0901317 analogs.

NR LBDs are fused to the yeast GAL4 DNA binding domain, which binds to a reporter plasmid. Expression of the reporter (firefly luciferase for PXR or beta-lactamase for LXRs) is increased in the presence of an agonist.

Compound	PXR		LXRα		LXRβ
	AUC±SD	FC	AUC ± SD	FC	AUC ± SD	FC
T0901317	175.9 ± 6.3	1	186.8 ± 11.4	1	192.6 ± 28	1
SJPYT-299	111.4 ± 24.3	0.63	73.5 ± 3.5	0.39	78.4 ± 9.2	0.41
SJPYT-300	150.9 ± 4.8	0.86	131.4 ± 8.2	0.70	134.7 ± 10.9	0.70
SJPYT-301	164.8 ± 24.4	0.94	124.9 ± 10.1	0.67	137.4 ± 26.6	0.71
SJPYT-302	107.3 ± 17.3	0.61	120.8 ± 7.6	0.65	130.7 ± 19	0.68
SJPYT-303	154 ± 29.8	0.88	103.9 ± 2.9	0.56	107.9 ± 16.9	0.56
SJPYT-304	145.2 ± 25.7	0.83	103.5 ± 4.1	0.55	86.1 ± 18.7	0.45
SJPYT-305	119.9 ± 19.9	0.68	35.1 ± 4.1	0.19	70.4 ± 23.4	0.37
SJPYT-315	101.6 ± 19.8	0.58	98.0 ± 3.4	0.52	104.2 ± 18.9	0.54
SJPYT-316	67.8 ± 16.2	0.39	88.8 ± 5.9	0.48	108.7 ± 13	0.56
SJPYT-317	70.6 ± 6.8	0.40	43.1 ± 4.7	0.23	56.5 ± 10.1	0.29
SJPYT-318	102.5 ± 20.9	0.58	11.9 ± 4.1	0.06	36.4 ± 9.8	0.19
SJPYT-319	34.9 ± 1.3	0.20	12.1 ± 1.3	0.06	38.7 ± 14	0.20
SJPYT-320	9.1 ± 4.9	0.05	4.1 ± 3.3	0.02	21.1 ± 7.4	0.11

Figure 4. SJPYT-305 and SJPYT-318 do not modulate additional NRs.

NR reporter activity assays were performed for (A) FXR, (B) RORγ, (C) PPARδ, or (D) PPARγ. Cells were treated with DMSO or 1 μM of the indicated compounds for 24 h. T0901317 is a known FXR agonist and RORγ inverse agonist and served as the assay control for (A) and (B). T0901317 is not reported to modulate PPARδ or PPARγ, so GW0742 or rosiglitazone served as the respective control for (C) and (D). Results are plotted as fold change (FC) relative to the average DMSO control signal and are expressed as the mean ± standard deviation from at least three replicates.

SJPYT-305 and SJPYT-318 selectively activate endogenous PXR

Next, we aimed to investigate the impact of T0901317 analogs on PXR- and LXR-dependent transcription in a more biologically relevant context by examining endogenous CYP3A4 and ABCA1 RNA induction by endogenous PXR and LXRα/β, respectively. Because PXR activation in cell lines typically results in a low-magnitude transcription response, we first examined a panel of cell lines for the greatest CYP3A4 induction upon PXR activation. Guided by the Dependency Map portal (depmap.org),³¹ we selected eight cell lines with high PXR RNA expression, including SNU-C4 cells, which we have previously published for PXR studies.^32,33 We confirmed PXR RNA levels and assessed basal and rifampicin-induced CYP3A4 RNA levels by RT-qPCR (Figure S4). While SNU-C4 indeed had rifampicin-inducible CYP3A4 RNA expression, SNU719 inducibility was significantly greater (Figure S4C). Furthermore, the LXR target gene ABCA1 was highly inducible by LXR agonists, and rifampicin and GW3965 maintained their receptor specificity in SNU719 (Figure S5). Therefore, we chose SNU719 as the cell model for further study.

Importantly, SJPYT-305 and SJPYT-318 both retained high PXR selectivity in the SNU719 model (Figure 5). CYP3A4 induction by SJPYT-305 was similar to that by T0901317, but SJPYT-305 also slightly induced ABCA1 expression (2.3-fold for SJPYT-305 vs. 11.1-fold for T0901317). SJPYT-318, on the other hand, robustly induced CYP3A4 RNA without increasing ABCA1 RNA. None of the compounds notably altered expression of the control gene GAPDH. These results validate the selectivity of SJPYT-318 and SJPYT-305 for PXR and highlight the ability of PXR to incorporate bulky ligands that other NRs cannot.

Figure 5. SJPYT-305 and SJPYT-318 selectively induce endogenous PXR.

SNU719 cells were treated with DMSO or the indicated concentrations of (A) T0901317, (B) SJPYT-305, or (C) SJPYT-318 for 24 h. RNA was isolated and subjected to RT-qPCR to measure expression of CYP3A4, ABCA1, or GAPDH. 18S RNA was used for normalization of all samples. FC, fold change relative to DMSO controls. Results are expressed as the mean ± standard deviation from three replicates. See also Figures S4-S5.

Crystal structure of PXR LBD bound to SJPYT-318 demonstrates cooperativity between PXR flexibility and ligand conformational freedom

We achieved PXR selectivity of T0901317 derivatives by adding inflexible, bulky groups at the N-position. We previously showed that for PXR, analogs with bulky N-position substituents had decreased PXR activity due to unfavorable ligand binding pocket rearrangements. In this prior study, we found that SJPYT-319 (containing a biphenyl extension) substantially altered the position of alpha helix 2 (α2) of PXR LBD.²⁷ Oddly, SJPYT-318 has a larger group than SJPYT-319 at the N-position (phenoxyphenyl for SJPYT-318 vs. biphenyl for SJPYT-319), and yet, SJPYT-318 more potently activated PXR than SJPYT-319 (Figure 3C and Table 2). This unexpected finding combined with the selectivity of SJPYT-318 prompted us to investigate the binding mode of SJPYT-318 to PXR LBD, and we solved the co-crystal structure to 2.2 Å (Table 3).

Table 3. Data collection and model refinement statistics for the SJPYT-318-bound PXR LBD structure (PDB code 8SZV).

Values from the highest resolution shell are shown in parentheses. ¹Generated with MolProbity.

Data Collection
Resolution range (Å)	45.74-2.20 (2.27-2.20)
Space group	P 43 21 2
Unit cell dimensions
a, b, c (Å)	91.47, 91.47, 85.07
α, β, γ (°)	90.0, 90.0, 90.0
Wavelength (Å)	1.00003
Unique reflections	18851
Redundancy	14.6 (15.3)
Completeness (%)	99.9 (99.1)
I/σI	13.8 (2.4)
RMerge	0.100 (1.339)
CC1/2	0.998 (0.829)
Model Refinement
Rwork/Rfree	0.211/0.223
Number of atoms	2220
Protein	2137
Ligand	40
Water	43
RMSD
Bond length (Å)	0.005
Bond angles (°)	0.769
Ramachandran plot (%)
Preferred	98.52
Outliers	0.00
Clashscore1	0.94
MolProbity score1	0.78
Average B-factor (Å2)	56.29
Protein	55.76
Ligand	83.31
Water	57.57

As with all reported PXR LBD structures obtained in the absence of SRC-1 peptide, the asymmetric unit contained a single PXR LBD protomer (Figure 6A). The orientation of SJPYT-318 was clear, although the benzenesulfonamide moiety appeared to have at least two orientations – one toward the F288-W299-Y306 aromatic cage and one toward α2. The aromatic cage-oriented position seemed to be low occupancy, so we modeled only the α2 orientation (Figure 6B). The hexafluoro hydroxypropyl moiety was positioned similarly to T0901317 and previous derivatives, with the hydroxyl group forming a hydrogen bond with H407 (Figure 6C). Surprisingly, however, the benzenesulfonamide and N-position substituent of SJPYT-318 were rotated ~180° compared to the parental T0901317 and related SJPYT-319 (Figure 6D). Rather than being oriented over α2 like the biphenyl moiety of SJPYT-319, the phenoxyphenyl of SJPYT-318 was rotated toward α3 and α5, with the terminal benzene ring near α12 (also known as the AF-2 helix). This position required a rotamer flip of F251 and subsequent movement of L428 and F429 to accommodate the altered rotamer (Figure 6E). Interestingly, the position of the terminal benzene ring and resulting α12 interactions somewhat resembled that of our previously reported agonist SJB7, with M425, L428, and F429 forming a hydrophobic surface that interacts with the ligand (Figure 6F).³² The α3 residue F251 further stabilizes the complex by interacting with both the SJPYT-318 terminal benzene ring and the α12hydrophobic surface, pinning α12 to the ligand binding pocket (Figure 6F). Thus, SJPYT-318 is a more efficient PXR agonist than SJPYT-319 because it is reoriented to avoid the α2 clash, and other regions of the flexible ligand binding pocket move to not only accommodate the new ligand position, but to favorably incorporate the ligand in its new binding mode. The new ligand binding mode is made possible by 1) the flexibility of the PXR ligand binding pocket and 2) the flexibility imparted to the ligand by the addition of an oxygen between the two benzene rings.

Figure 6. SJPYT-318 adopts new binding mode compared to parental T0901317.

PXR LBD was co-crystal I ized with SJPYT-318 to a resolution of 2.2 Å. (A) One chain of PXR LBD was present in the asymmetric unit. The protein is shown as light blue cartoon, and SJPYT-318 is represented as deep blue spheres. (B) The 2Fo─Fc map for SJPYT-318 is contoured in mesh at 1.0 rmsd and carved around SJPYT-318 at 2.0 Å. (C) Residues of the ligand binding pocket within 5 Å of SJPYT-318 are shown as sticks. (D) Comparison of SJPYT-318 binding mode with binding modes of SJPYT-319 (PDB ID 8FPE) and T0901317 (PDB ID 2O9I, chain A). (E) SJPYT-318-bound PXR LBD (light blue protein with deep blue ligand) is overlaid with T0901317-bound PXR LBD (gray). α12 residues (L428 and F429) and the adjacent F251 move to incorporate the phenoxyphenyl group of SJPYT-318. (F) SJPYT-318-bound PXR LBD (light blue protein with deep blue ligand) is overlaid with SJB7-bound PXR LBD (PDB ID 5X0R, light teal). The α12 residues M425, L428, and F429 and the adjacent F251 are similarly arranged to incorporate the two ligands.

DISCUSSION

Both activation and inhibition of NRs can be beneficial or detrimental in context-specific manners. For example, oral contraceptives contain estrogen receptor agonists while estrogen receptor antagonists are staples of breast cancer treatments.^34,35 Similarly, PXR antagonists have potential value for enhancing drug bioavailability and treating conditions such as drug- or hemorrhagic shock-induced liver injury,^32,36,37 while agonists may be used for treatment of IBD.²² These dichotomies combined with multi-NR modulation of individual promiscuous chemicals reinforce the need for specificity in NR-targeting molecules. Achieving selectivity could open new avenues to treat pathologies in which NRs are known to play key roles.

Among the 48 human nuclear receptors, PXR is unique in its large ligand binding pocket volume (1200-1600 ų)²⁸ that allows binding of a vast array of molecules. Because of its propensity to be activated by diverse molecules and its role in regulating drug metabolism genes, PXR is often defined as a xenobiotic receptor. However, PXR also plays roles in homeostatic processes such as glucose metabolism, diabetes, intestinal barrier function, and general liver function.²⁴ PXR has been shown to mitigate intestinal inflammation through interaction with gut microbial metabolites and reduction of toll-like receptor 4 (TLR4) signaling.³⁸ Exogenous PXR agonists have also been shown to protect against IBD through crosstalk with nuclear factor kappa B (NF-κB) signaling.³⁹ Furthermore, PXR activation by the IBD drug rifaximin has been proposed as rifaximin’s therapeutic mechanism, but the extent of PXR’s role is unknown because rifaximin is also a potent antibiotic that alters gut microbiome composition.^25,26 PXR may therefore be a viable target for IBD treatment, but studies are currently limited by nonspecific agonists.

Though T0901317 has previously been used as a platform to generate modulators of NRs such as LXRs and retinoic acid-related orphan receptors,^40,42 and attempts at deriving PXR antagonists have even been made,⁴³ efforts aimed at enhancing PXR selectivity are lacking, and there is little information available specifically on the N-position. While the N-trifluoroethyl is directed at a flexible region of PXR LBD, it is in a relatively constrained area of the LXRβ LBD,⁴⁴ and it was reasonable to predict that N-position extensions would disproportionately diminish LXR activity over PXR. However, we were surprised to find that the phenoxyphenyl substitution in SJPYT-318 not only imparted PXR selectivity, but also completely flipped the binding mode. This was made possible both by the rotatable bonds in the molecule and the permissive nature of the PXR ligand binding pocket. The rotation brought the phenoxyphenyl group in close proximity to α12, similar to our previously reported agonist SJB7, resulting in productive stabilizing interactions. Interestingly, as we have previously studied SJB7 analogs as PXR antagonists, similar approaches might be used to generate antagonists from SJPYT-318. Thus, SJPYT-318 represents a PXR-selective agonist that can serve as a starting point for further development of both agonists and antagonists.

STAR METHODS

RESOURCE AVAILABILITY

Lead contact

Further information and requests for resources and reagents should be directed to and will be fulfilled by the lead contact, Taosheng Chen (taosheng.chen@stjude.org).

Materials availability

All unique/stable reagents generated in this study are available from the lead contact with a completed materials transfer agreement. There are restrictions to the availability of new compounds due to the lack of an external centralized repository for their distribution and our need to produce and maintain the stocks. We are glad to share new compounds with reasonable compensation by the requestor for its production, processing, and shipping.

Data and code availability

• The structure of SJPYT-318-bound PXR LBD was deposited in the Protein Data Bank (PDB) under accession code 8SZV.

• This paper does not report original code.

• Any additional information required to reanalyze the data reported in this paper is available from the lead contact upon request.

EXPERIMENTAL MODEL AND STUDY PARTICIPANT DETAILS

293T/17 cells (referred to as 293T) were obtained from the American Type Culture Collection (ATCC, cat. # CRL-11268) and maintained in Dulbecco's Modified Eagle Medium (DMEM, ATCC, cat. # 30-2002) with 10% fetal bovine serum (FBS, HyClone, cat. # SH30396.03). LS1034 cells were obtained from ATCC (cat. # CRL-2158) and maintained in Roswell Park Memorial Institute 1640 (RPMI) medium (ATCC, cat. # 30-2001) with 10% FBS. SW948 (cat. # CCL-237), SW1417 (cat. # CCL-238), and SW1463 (cat. # CCL-234) cells were obtained from ATCC and maintained in Leibovitz's L-15 Medium (ATCC, cat. # 30-2008) with 10% FBS. SNU-C4 (cat. # 0000C4), SNU81 (cat. # 00081), SNU719 (cat. # 00719), and SNU620 (cat. # 00620) cells were obtained from the Korean Cell Line Bank (KCLB) and maintained in RPMI with 10% FBS. GeneBLAzer LXRα-UAS-bla HEK 293T Cells (cat. # K1692) and GeneBLAzer LXRβ-UAS-bla HEK 293T Cells (cat. # K1699) were obtained from Thermo Fisher Scientific and maintained in DMEM with 10% FBS, 100 μg/mL Hygromycin B (Thermo Fisher Scientific, cat. # 10687010), and 100 μg/mL Zeocin (Thermo Fisher Scientific, cat. # R25001). GeneBLAzer PPARδ-UAS-bla HEK 293T Cells (Thermo Fisher Scientific, cat. # K1690) were maintained in DMEM with 10% FBS, 80 μg/mL Hygromycin B, and 100 μg/mL Zeocin. GeneBLAzer PPARγ-UAS-bla HEK 293H Cells (Thermo Fisher Scientific, cat. # K1701) were maintained in DMEM with 10% FBS, 100 μg/mL Hygromycin B, and 500 μg/mL geneticin (Thermo Fisher Scientific, cat. # 10131027). The GeneBLAzer cells were cultured on plates coated with 0.25% Matrigel Growth Factor Reduced Basement Membrane Matrix (Corning, cat. # 356230), as per the manufacturer’s recommendations. All cells were routinely verified to be mycoplasma free by using the MycoProbe Mycoplasma Detection Kit (R&D Systems), and, apart from the SW lines, were incubated in a humidified atmosphere at 37°C with 5% CO₂. SW948, SW1417, and SW1463 cells were incubated in a humidified atmosphere at 37°C without CO₂, according to ATCC recommendations. Cell counts were obtained with a Countess II Automated Cell Counter using trypan blue staining. The “Assay Medium” referenced in the following subsections was DMEM, high glucose, HEPES, no phenol red (Thermo Fisher Scientific, cat. # 21063029) supplemented with 2% charcoal/dextran treated FBS (HyClone, cat. # SH30068.03), 1 mM sodium pyruvate (Thermo Fisher Scientific, cat. # 11360070), and 0.1 mM non-essential amino acids (Thermo Fisher Scientific, cat. # 11140050). For protein expression, TurboCells Competent E. coli BL21(DE3) (Genlantis, cat. # C302020) were grown in terrific broth at 37°C.

METHOD DETAILS

General chemistry methods and synthesis

Dimethyl sulfoxide (DMSO) was purchased from Fisher Scientific (cat. # BP231-100). T0901317 was purchased from Cayman Chemical (cat. # 71810). Rosiglitazone (cat. # HY-17386) and GW0742 (cat. # HY-13928) were purchased from MedChemExpress. Organic reagents were purchased from commercial suppliers unless otherwise noted and were used without further purification. All solvents were analytical or reagent grade and the solvents were dried using the Glass Contour Solvent Systems by SG Water USA. All reactions with water- and/or air-sensitive starting materials were carried out in pre-dried glassware under argon atmosphere with standard procedure. Flash column chromatography was performed by using Biotage Isolera Flash Systems and Biotage SNAP Ultra or Biotage^® SNAP Ultra C18 columns (Biotage, Charlotte, NC). All reactions as well as compound purity were monitored by UPLC-MS by using a Waters Acquity UPLC MS system with a C18 column in a 2-min gradient (H₂O + 0.1% formic acid → acetonitrile + 0.1% formic acid) and detectors of PDA (215–400 nm), ELSD, and Acquity SQD ESI-positive MS (Waters Corporation, Milford, MA). High-resolution mass spectra were determined by using a Waters Acquity UPLC system with a C18 column (H₂O + 0.1% formic acid → acetonitrile + 0.1% formic acid gradient over 2.5 min) and Xevo G2Q-TOF ESI-positive MS in resolution mode. Compounds were internally normalized to leucine-enkephalin lock solution, with a calculated error of <5 ppm. All final compounds used for SAR studies have purity at 95% or greater (purity was calculated based on the average of the ELSD detector purity and UV detector purity). All NMR spectra were recorded on a Bruker 500 MHz spectrometer (Bruker Corporation) in the solvents indicated and spectra were processed using MestReNova (14.1.0) (Mestrelab Research). The chemical shift values are expressed in parts per million (ppm) relative to tetramethylsilane as the internal standard. Coupling constants (J) are reported in hertz (Hz). ¹H NMR, ¹³C NMR, and HRMS spectra and HPLC chromatograms are shown in Data S1.

Syntheses of SJPYT-302, 315-317, and 319 have been described previously.²⁷ Syntheses of SJPYT-300, 301, 303, 304, 318, and 320 were prepared by a similar procedure. Sulfonamidation of 2-(4-Aminophenyl)-1,1,1,3,3,3-hexafluoropropan-2-ol (1a) with benzenesulfonyl chloride in the presence of 2,6-Lutidine gave N-[4-(1,1,1,3,3,3-Hexafluoro-2-hydroxypropan-2-yl)phenyl]benzenesulfonamide (1b). The resulting sulfonamide (1b) was then alkylated with respective alkyl-halides in the presence of potassium carbonate and heating to produce various N-alkyl analogs (SJPYT-300 to SJPYT-304 and SJPYT-315 to SJPYT-320). For SJPYT-299 and SJPYT-305, the aniline 1a was alkylated under reductive amination conditions with corresponding aldehydes in the presence of sodium cyanoborohydride to generate the secondary aniline intermediates (2a and 2b) which were then converted to SJPYT-299 and SJPYT-305 via sulfonamidation with benzenesulfonyl chloride.⁴¹

Synthesis of SJPYT-299 to 305 and SJPYT-315 to 320

Reagents and conditions: (a) Benzenesulfonyl chloride, 2,6-lutidine, acetone, reflux; (b) RBr or RI, K₂CO₃, CH₃CN, 70°C; (c) 3,3,3-trifluoropropanal or 4-hydroxybenzaldehyde, NaBH₃CN, MeOH, room temperature; (d) PhSO₂Cl, pyridine, DMAP (cat.), toluene, reflux.

N-ethyl-N-(4-(1,1,1,3,3,3-hexafluoro-2-hydroxypropan-2-yl)phenyl)benzenesulfonamide (SJPYT-300)

Benzenesulfonyl chloride (2.59 ml, 20.26 mmol) and 2,6-dimethylpyridine (4.13 g, 38.6 mmol) were added to a solution of 2-(4-aminophenyl)-1,1,1,3,3,3-hexafluoropropan-2-ol (1a, 5g, 19.29 mmol) in acetone (30 mL) at room temperature. The mixture was stirred at 60°C overnight. After mass spectrometry (MS) demonstrated the complete consumption of the starting material 1a, the reaction mixture was concentrated, and water was added (100 mL). The aqueous layer was extracted with EtOAc (100 mL × 2). The combined organic layer was washed with brine. The organic layer was then dried over MgSO₄, filtered, and concentrated under reduced pressure. The residue was purified by silica gel chromatography (0% to 100% acetonitrile in water) to give N-(4-(1,1,1,3,3,3-hexafluoro-2-hydroxypropan-2-yl)phenyl)benzenesulfonamide (1b) as an off-white solid (6.78 g, 88% yield). ¹H NMR (500 MHz, DMSO-d₆) δ 11.21 (s, 1H), 9.10 (s, 1H), 8.45 – 8.24 (m, 2H), 8.18 – 8.12 (m, 1H), 8.11 – 8.00 (m, 4H), 7.81 – 7.67 (m, 2H).

K₂CO₃ (138 mg, 1.002 mmol) and iodoethane (44.3 μL, 0.551 mmol) were added to a solution of N-(4-(1,1,1,3,3,3-hexafluoro-2-hydroxypropan-2-yl)phenyl)benzenesulfonamide (1b, 200 mg, 0.501 mmol) in 5 mL CH₃CN at room temperature. The suspension was stirred at 70°C overnight. Water (50 mL was added, and the reaction mixture was extracted with EtOAc (50 mL × 2). The EtOAc layer was washed with water, dried with anhydrous Na₂SO₄, and concentrated. The residue was purified by silica gel chromatography (0% to 100% EA in hexane) to give product SJPYT-300 as a white solid, 161.3 mg, 75% yield, 96.86% purity. ¹H NMR (500 MHz, DMSO-d₆) δ 8.81 (s, 1H), 7.71 (tt, J = 5.9, 2.6 Hz, 1H), 7.67 (d, J = 8.4 Hz, 2H), 7.62 – 7.55 (m, 4H), 7.28 – 7.22 (m, 2H), 3.62 (q, J = 7.1 Hz, 2H), 0.97 (t, J = 7.1 Hz, 3H). ¹³C NMR (126 MHz, DMSO-d₆) δ 140.64, 138.23, 133.74, 130.29, 129.86, 128.70, 128.09, 127.57, δ 123.29 (q, J = 288.7 Hz), 77.14 (p, J = 29.4 Hz), 45.31, 14.34. ESI-TOF HRMS: m/z 428.0760 (C₁₇H₁₅F₆NO₃S + H⁺ requires 428.0750).

N-(4-(1,1,1,3,3,3-hexafluoro-2-hydroxypropan-2-yl)phenyl)-N-propylbenzenesulfonamide (SJPYT-301)

The title compound (SJPYT-301) was synthesized using a procedure similar to that described for SJPYT-300, with the utilization of 1b and 1-iodopropane to give a white solid, 173.1 mg, 70% yield, 97.92% purity. ¹H NMR (500 MHz, DMSO-d₆) δ 8.80 (s, 1H), 7.70 (ddt, J = 8.6, 6.9, 1.7 Hz, 1H), 7.66 (d, J = 8.4 Hz, 2H), 7.62 – 7.52 (m, 4H), 7.32 – 7.23 (m, 2H), 3.54 (t, J = 7.0 Hz, 2H), 1.31 (h, J = 7.2 Hz, 2H), 0.82 (t, J = 7.3 Hz, 3H). ¹³C NMR (126 MHz, DMSO-d₆) δ 140.82, 138.13, 133.72, 130.30, 129.82, 128.68, 128.07, 127.58, 126.73, 124.44, 122.14, 119.84, 77.61, 77.38, 77.14, 76.91, 76.68, 51.68, 21.47, 11.21. ESI-TOF HRMS: m/z 442.0918 (C₁₈H₁₇F₆NO₃S + H⁺ requires 442.0906).

N-benzyl-N-(4-(1,1,1,3,3,3-hexafluoro-2-hydroxypropan-2-yl)phenyl)benzenesulfonamide (SJPYT-303)

The title compound (SJPYT-303) was synthesized using a procedure similar to that described for SJPYT-300, with the utilization of 1b and benzyl bromide to give a white solid, 179.9 mg, 73% yield, 98.34% purity. ¹H NMR (500 MHz, DMSO-d₆) δ 8.72 (s, 1H), 7.79 – 7.72 (m, 1H), 7.68 – 7.59 (m, 4H), 7.56 (d, J = 8.5 Hz, 2H), 7.32 – 7.24 (m, 6H), 7.19 (ddd, J = 8.7, 4.9, 3.9 Hz, 1H), 4.85 (s, 2H). ¹³C NMR (126 MHz, DMSO-d₆) δ 140.81, 138.16, 136.49, 133.93, 130.06, 129.95, 128.86, 128.46, 128.34, 127.98, 127.87, 127.68, 126.66, 119.78, 77.54, 77.31, 77.07, 76.84, 76.61, 53.50. ESI-TOF HRMS: m/z 490.0905 (C₂₂H₁₇F₆NO₃S + H⁺requires 490.0906).

N-(4-(1,1,1,3,3,3-hexafluoro-2-hydroxypropan-2-yl)phenyl)-N-(4-methoxybenzyl)benzenesulfonamide (SJPYT-304)

The title compound (SJPYT-304) was synthesized using a procedure similar to that described for SJPYT-300 with the utilization of 1b and 1-(bromomethyl)-4-methoxybenzene to give a white solid, 445.2 mg, 73% yield, 98.38% purity. ¹H NMR (500 MHz, DMSO-d₆) δ 8.73 (s, 1H), 7.73 (tt, J = 6.7, 1.9 Hz, 1H), 7.67 – 7.59 (m, 4H), 7.56 (d, J = 8.5 Hz, 2H), 7.27 – 7.22 (m, 2H), 7.18 – 7.13 (m, 2H), 6.83 – 6.77 (m, 2H), 4.76 (s, 2H), 3.67 (s, 3H). ¹³C NMR (126 MHz, DMSO-d₆) δ 159.06, 140.73, 138.23, 133.87, 130.03, 129.93, 129.77, 128.55, 128.15, 127.83, 127.65, 126.67, 124.38, 122.08, 119.79, 114.25, 77.54, 77.31, 77.08, 76.85, 76.62, 55.44, 52.97. ESI-TOF HRMS: m/z 518.0881 (C₂₃H₁₉F₆NO₄S - H⁺ requires 518.0861).

N-(4-(1,1,1,3,3,3-hexafluoro-2-hydroxypropan-2-yl)phenyl)-N-(4-phenoxybenzyl)benzenesulfonamide (SJPYT-318)

The title compound (SJPYT-318) was synthesized using a procedure similar to that described for SJPYT-300 with the utilization of 1b and 1-(bromomethyl)-4-phenoxybenzene to give a white solid, 205.9 mg, 71% yield, 95.38% purity. ¹H NMR (500 MHz, DMSO-d₆) δ 8.75 (s, 1H), 7.78 – 7.69 (m, 1H), 7.71 – 7.59 (m, 4H), 7.57 (d, J = 8.5 Hz, 2H), 7.39 – 7.31 (m, 2H), 7.30 – 7.23 (m, 4H), 7.16 – 7.06 (m, 1H), 6.94 – 6.85 (m, 4H), 4.82 (s, 2H). ¹³C NMR (126 MHz, DMSO-d₆) δ 155.89, 155.31, 139.70, 137.08, 132.85, 130.36, 129.39, 129.13, 129.02, 128.88, 127.54, 126.79, 126.60, 125.60, 123.30, 122.89, 121.00, 118.72, 117.90, 76.23, 76.00, 75.77, 75.54, 51.96. ESI-TOF HRMS: m/z 580.1010 (C₂₈H₂₁F₆NO₄S - H⁺ requires 580.1017).

N-(4-(1,1,1,3,3,3-hexafluoro-2-hydroxypropan-2-yl)phenyl)-N-(4-(1,2,2-triphenylvinyl)benzyl)benzenesulfonamide (SJPYT-320)

The title compound (SJPYT-320) was synthesized using a procedure similar to that described for SJPYT-300 with the utilization of 1b and (2-(4-(bromomethyl)phenyl)ethene-1,1,2-triyl)tribenzene to give a white solid, 304.8 mg, 82% yield, 99.39% purity.¹H NMR (500 MHz, DMSO-d₆) δ 8.77 (s, 1H), 7.75 – 7.69 (m, 1H), 7.67 – 7.54 (m, 6H), 7.18 – 7.14 (m, 2H), 7.12 – 6.98 (m, 11H), 6.95 – 6.91 (m, 2H), 6.90 – 6.77 (m, 6H), 4.75 (s, 2H). ¹³C NMR (126 MHz, DMSO-d₆) δ 142.44, 142.37, 142.35, 141.77, 140.06, 139.66, 139.47, 137.24, 133.54, 132.81, 130.01, 129.95, 129.04, 128.86, 127.59, 127.20, 127.15, 127.05, 126.86, 126.71, 126.59, 125.94, 125.84, 125.64, 123.34, 121.05, 118.75, 76.49, 76.25, 76.02, 75.79, 75.55, 52.43. ESI-TOF HRMS: m/z 742.1843 (C₄₂H₃₁F₆NO₃S - H⁺ requires 742.1851).

N-(4-(1,1,1,3,3,3-hexafluoro-2-hydroxypropan-2-yl)phenyl)-N-(3,3,3-trifluoropropyl)benzenesulfonamide (SJPYT-299)

NaCNBH₃ (485 mg, 7.72 mmol) and 3,3,3-trifluoropropanal (173 mg, 1.543 mmol) were added to a solution of 2-(4-aminophenyl)- 1,1,1,3,3,3-hexafluoropropan-2-ol (1a, 400 mg, 1.543 mmol) in MeOH (5 mL) at room temperature. The suspension was stirred at room temperature overnight. The reaction mixture was diluted with water (50 mL) and extracted with EtOAc (50 mL × 2). The EtOAc layer was washed with water (50 mL × 2), dried with anhydrous Na₂SO₄, and concentrated. The residue (crude 2a) was mixed with pyridine (5 mL), toluene (10 mL), DMAP (cat.), and benzenesulfonyl chloride (296 μL, 2.32 mmol). The reaction mixture was under reflux overnight. The reaction mixture was poured into water (50 mL) and extracted with EtOAc (50 mL × 2). The EtOAc layer was washed with water (50 mL × 2), dried with anhydrous Na₂SO₄, and concentrated. The residue was purified by silica gel chromatography (0% to 100% EA in hexane) to give product SJPYT-299 as a white solid, 212.4 mg, 28% yield (two steps), 98.14% purity. ¹H NMR (500 MHz, DMSO-d₆) δ 8.83 (d, J = 0.9 Hz, 1H), 7.76 – 7.70 (m, 1H), 7.67 (d, J = 8.4 Hz, 2H), 7.62 – 7.53 (m, 4H), 7.30 – 7.24 (m, 2H), 3.87 (t, J = 6.7 Hz, 2H), 2.43 (qt, J = 11.1, 6.7 Hz, 2H). ¹³C NMR (126 MHz, DMSO-d₆) δ 139.29, 136.31, 133.00, 129.71, 128.97, 128.85, 127.71, 127.15, 126.77, 126.62, 125.63, 124.56, 123.34, 122.36, 121.05, 118.76, 76.53, 76.30, 76.07, 75.83, 75.60, 43.15, 43.11, 43.09, 43.06, 31.88, 31.67, 31.45, 31.23. ESI-TOF HRMS: m/z 496.0630 (C₁₈H₁₄F₉NO₃S+ H⁺ requires 496.0623).

N-(4-(1,1,1,3,3,3-hexafluoro-2-hydroxypropan-2-yl)phenyl)-N-(4-hydroxybenzyl)benzenesulfonamide (SJPYT-305)

The title compound (SJPYT-305) was synthesized using a procedure similar to that described for SJPYT-299 with the utilization of 1a and 4-hydroxybenzaldehyde to give a white solid, 103.2 mg, 13% yield (two steps), 99.26% purity. ¹H NMR (500 MHz, DMSO-d₆) δ 9.34 (s, 1H), 8.72 (s, 1H), 7.73 (td, J = 6.7, 2.1 Hz, 1H), 7.66 – 7.59 (m, 4H), 7.55 (d, J = 8.5 Hz, 2H), 7.25 – 7.18 (m, 2H), 7.06 – 6.96 (m, 2H), 6.63 – 6.55 (m, 2H), 4.70 (s, 2H).¹H NMR (500 MHz, DMSO-d₆) δ 9.34 (s, 1H), 8.72 (s, 1H), 7.73 (td, J = 6.7, 2.1 Hz, 1H), 7.66 – 7.59 (m, 4H), 7.55 (d, J = 8.5 Hz, 2H), 7.25 – 7.18 (m, 2H), 7.06 – 6.96 (m, 2H), 6.63 – 6.55 (m, 2H), 4.70 (s, 2H). ESI-TOF HRMS: m/z 550.0728 (C₂₂Hi₇F₆NO₄S + HCOO⁻ requires 550.5765).

Plasmids

The pG5luc vector containing five GAL4 binding sites upstream of a minimal TATA box and the firefly luciferase gene was obtained from Promega as a component of the CheckMate Mammalian Two-Hybrid System (cat. # E2440). The pBIND vector containing the yeast GAL4 DNA-binding domain upstream of a multiple cloning region was also a component of the CheckMate Mammalian Two-Hybrid System (Promega, cat. # E2440). The PXR expression vector pcDNA3-FLAG-PXR has been described previously.⁴⁵ To generate pBIND-PXR LBD, which contains the GAL4 DNA-binding domain fused to PXR LBD (residues 130-434), the PXR LBD was amplified from pcDNA3-FLAG-PXR using primers 5’-TCTAGAGCGGCCGCAGGTACCAGTGAACGGACAGGGACTC-3’ and 5’-GGCCTTAGTTATTCAGGTACCTCAGCTACCTGTGATGCC-3’. The amplicon was then inserted into KpnI-digested pBIND using NEBuilder HiFi DNA Assembly Master Mix (New England Biolabs, cat. # E2621S). To generate pBIND-FXR LBD and pBIND-RORγ LBD, KpnI-digested pBIND was ligated to gBlocks (Integrated DNA Technologies) containing either FXR LBD (residues 214-472) or RORγ LBD (residues 251-518) using NEBuilder HiFi DNA Assembly Master Mix. The pETDuet-1 vector (Sigma-Aldrich, cat. # 71146) containing His-tagged PXR LBD (residues 130-434) and untagged mouse SRC-1 (residues 623-710) has been described previously.^27,32

Time-resolved fluorescence resonance energy transfer (TR-FRET) PXR LBD coactivator recruitment assay

The TR-FRET PXR LBD coactivator recruitment assay measures PXR LBD interaction with a fluorescently labeled SRC-1 peptide (FAM-SRC1-B) and was performed as previously described with minor modifications.⁴⁶ The fluorescently labeled peptide FAM-SRC1-B with the sequence of FAM-(PEG)₅-CPSSHSSLTERHKILHRLLQEGSPS-NH₂ was synthesized by the Macromolecular Synthesis Section of the Hartwell Center for Bioinformatics and Biotechnology at St. Jude Children's Research Hospital. The assay buffer composition was 50 mM Tris (pH 7.5), 20 mM MgCl₂, 0.1 mg/mL bovine serum albumin (BSA), and 0.05 mM dithiothreitol (DTT). FAM-SRC1-B peptide solution (10 μL/well, 150 nM in assay buffer) was dispensed into ProxiPlate-384 Plus F black 384-shallow well microplates (PerkinElmer, cat. # 6008260). An Echo 650 Acoustic Liquid Handler then dispensed 15 nL/well of compound stocks or DMSO. Lastly, 5 μL/well of 15 nM LanthaScreen Tb-anti-GST antibody (Thermo Fisher Scientific, cat. # PV3550) and 15 nM GST-PXR LBD protein (Thermo Fisher Scientific, cat. # PV4841) in assay buffer was added. The final assay volume per well was 15 μL, and the final concentrations for the assay components were: 100 nM FAM-SRC1-B peptide, 5 nM Tb-anti-GST, 5 nM GST-PXR LBD, and 0.2% DMSO (0.1% from the compound addition and 0.1% from the peptide solution). 0.2% DMSO and 1 μM T0901317 were included in each plate as negative (0% signal) and positive (100% signal) controls, respectively. The plates were shaken at 900 rpm (80 × g) on an IKA MTS 2/4 digital microtiter shaker for 1 min then centrifuged at 200 × g for 30 s in an Eppendorf 5810 centrifuge equipped with an A-4-62 swing-bucket rotor. The plates were protected from light exposure and incubated for 90 min at room temperature. The TR-FRET signal from each well was collected with a PHERAstar FS microplate reader (BMG LABTECH) using a 340 nm excitation filter, 520 nm and 490 nm emission filters, a 100 μs delay, and a 200 μs integration time. The measured RFU were normalized for each well using equation 5,

$$Signal=\frac{RFUat520nm}{RFUat490nm}$$ Equation 5

and equation 6,

$$TR-FRETSignal(\%)=100\times \frac{(Signa{l}_{Compound}-Signa{l}_{DMSO})}{(Signa{l}_{T0901317}-Signa{l}_{DMSO})}$$ Equation 6

TR-FRET LXRα LBD and LXRβ LBD coactivator recruitment assays

The assays were performed similarly to the PXR LBD coactivator recruitment assay, with modifications. GST-LXRα LBD (cat. # PV4657) and GST-LXRβ LBD (cat. # PV4660) were purchased from Thermo Fisher Scientific. The assay buffer composition was 50 mM Tris (pH 7.5), 0.002% Pluronic F-127, 0.01% BSA, and 5 mM DTT. The final FAM-SRC1-B peptide concentrations were 50 nM for LXRα assay and 125 nM for the LXRβ assay.

Cytotoxicity assay

293T cells (5 × 10³/well in 25 μL Assay Medium) were plated and treated as in the reporter assays. After 24 h compound treatment, the CellTiter-Glo Luminescent Cell Viability Assay (Promega, cat. # G7572) and an EnVision microplate reader were used to assess cell viability. Cells treated with 0.1% DMSO served as positive controls (100% cell viability), and wells without cells (medium only) served as negative controls (0% cell viability). The percent cell viability for each well was calculated using equation 4,

$$CellViability(\%)=100\times \frac{(Signa{l}_{Compound}-Signa{l}_{Medium})}{(Signa{l}_{DMSO}-Signa{l}_{Medium})}$$ Equation 4

PXR transactivation assay

293T cells (1 × 10⁶/well in 2 mL/well DMEM with 10% FBS) were reverse transfected with pG5luc and pBIND-PXR LBD (1 μg/well each) using FuGENE HD Transfection Reagent (Promega, cat. # E2311). The following day, cells were trypsinized and resuspended in Assay Medium, and 5 × 10³ cells/well in 25 μL were added to white CulturPlate-384 plates (PerkinElmer, cat. # 6007680). An Echo 650 Acoustic Liquid Handler (Beckman Coulter) was used to dispense 25 nL/well of DMSO or stock compounds, resulting in 0.1% DMSO and the indicated concentrations of compounds. Cells treated with 0.1% DMSO served as negative controls (0% activity), cells treated with 1 μM T0901317 served as positive controls (100% activity), and wells without cells (medium only) were used for background correction. After 24 h, the steadylite plus Reporter Gene Assay System (PerkinElmer, cat. # 6066751) and an EnVision microplate reader (PerkinElmer) were used to measure luminescence. Data normalization occurred in two stages. First, the average measured relative light units (RLU) of the medium-only wells were subtracted from the RLU for all wells to give background-corrected values for each reading. Second, the percent activity was calculated for each well using equation 1,

$$Activity(\%)=100\times \frac{(CorrectedRL{U}_{Chemical}-CorrectedRL{U}_{DMSO})}{(CorrectedRL{U}_{T0901317}-CorrectedRL{U}_{DMSO})}$$ Equation 1

LXRα and LXRβ GeneBLAzer assay

GeneBLAzer cells (1 × 10⁴/well in 25 μL Assay Medium) were plated in black 384-well clear-bottom plates (Corning, cat. # 3764). An Echo 650 Acoustic Liquid Handler was used to dispense 25 nL/well of DMSO or stock compounds, resulting in 0.1% DMSO and the indicated concentrations of compounds. Cells treated with 0.1% DMSO served as negative controls (0% activity), cells treated with 1 μM T0901317 served as positive controls (100% activity), and wells without cells (medium only) were used for background correction. After 24 h, the LiveBLAzer FRET-B/G Loading Kit with CCF4-AM (Thermo Fisher Scientific, cat. # K1096) and a CLARIOstar Plus microplate reader (BMG LABTECH) were used to detect reporter activity. Readings were taken with excitation/emission wavelengths (nm) of 409/460 and 409/530. Data normalization occurred in three stages. First, the average relative fluorescence units (RFU) of the medium-only wells at 460 nm and 530 nm were subtracted from the RFU of all wells to give background-corrected values for each reading. Second, the “signal” for each well was calculated using equation 2,

$$Signal=\frac{CorrectedRFUat460nm}{CorrectedRFUat530nm}$$ Equation 2

Third, the percent activity was calculated for each well using equation 3,

$$Activity(\%)=100\times \frac{(Signa{l}_{Compound}-Signa{l}_{DMSO})}{(Signa{l}_{T0901317}-Signa{l}_{DMSO})}$$ Equation 3

FXR, RORγ, PPARδ, and PPARγ transactivation assays

Cellular FXR and RORγ assays were conducted by the same protocol as used for the PXR transactivation assay. However, rather than calculating percent activity, the background-corrected RLU values were normalized to DMSO control wells to yield fold change (FC). PPARδ and PPARγ GeneBLAzer assays were performed by the same protocol as used for LXRα and LXRβ. However, rather than calculating percent activity, the well signal values from equation 2 were normalized to DMSO control wells to yield FC.

Nuclear receptor target gene quantitative real-time polymerase chain reaction (RT-qPCR)

SNU-C4 cells (2 × 10⁵/well in 1 mL/well RPMI with 10% FBS) were plated in tissue culture-treated 12-well plates (Corning, cat. # 3512). Cells were grown for six days with fresh medium added every two days. Cells were washed with PBS, and 800 μL Assay Medium containing 0.5% DMSO and the indicated concentrations of compounds was added. After 24 h, total RNA was isolated from cells with Maxwell 16 LEV SimplyRNA Tissue Kits (Promega, cat. # AS1280), and cDNA was generated from 1 μg of RNA with the Superscript VILO cDNA Synthesis Kit (Thermo Fisher Scientific, cat. # 11754050). RT-qPCR was conducted with 2 μL of cDNA using TaqMan Fast Advanced Master Mix (Thermo Fisher Scientific, cat. # 4444557) in an Applied Biosystems 7500 Fast Real-Time PCR System. TaqMan gene expression assays specific for CYP3A4 (Hs00604506_m1), ABCA1 (Hs01059118_m1), GAPDH (Hs03929097_g1), and RNA18S (Hs03928990_g1) were purchased from Thermo Fisher Scientific. Fold induction values were calculated according to the 2^−ΔΔCt method, where ΔCt represents the differences in cycle threshold numbers between the target gene and reference gene and ΔΔCt represents the relative change in these differences between the control and treatment groups.⁴⁷ RNA18S was used as the reference gene for relative quantification of all other genes.

Protein purification, crystallization, and structure determination

PXR LBD was expressed and purified as previously described,^27,32 with minor modifications. Codon-optimized sequences for His-tagged PXR LBD (residues 130-434) and untagged mouse SRC-1 (residues 623-710) were cloned into pETDuet-1 (Novagen), which allows co-expression of two genes from separate inducible T7 promoters. The plasmid was transformed into TurboCells Competent E. coli BL21(DE3) (Genlantis, cat. # C302020), grown in terrific broth at 37°C to an OD₆₀₀ of 3-4, and induced overnight at 16°C with 500 μM isopropyl β-D-thiogalactoside (IPTG). Cells were pelleted by centrifugation at 4,000 × g and resuspended in lysis buffer [20 mM Tris (pH 7.5), 250 mM NaCl, 5% (v/v) glycerol, 10 mM imidazole] supplemented with EDTA-free SIGMAFAST protease inhibitor cocktail Tablets (MilliporeSigma, cat. # S8830-20TAB) and 1 mg/mL lysozyme (Gold Biotechnology, cat. # 12650-88-3). The suspension was sonicated and then centrifuged at 20,000 × g for one hour, and the supernatant was applied to a 5 mL HisTrap FF column (Cytiva). The column was washed with 50 mL lysis buffer, 25 mL lysis buffer with 50 mM imidazole, and 25 mL lysis buffer with 100 mM imidazole, and bound proteins were eluted with lysis buffer containing 500 mM imidazole. Elution fractions were collected and analyzed by SDS-PAGE for protein amount and purity. Selected fractions were pooled, and a 2:1 molar ratio of SRC-1 peptide was added to stabilize PXR LBD (N-CPSSHSSLTERHKILHRLLQEGSPS-C, prepared by the Macromolecular Synthesis Section at St. Jude Children’s Research Hospital). The PXR LBD/peptide mixture was concentrated to ≤10 mL in a Amicon Ultra-15 centrifugal filter unit with 10 kDa cutoff (MilliporeSigma), filtered through a 0.22 μm syringe filter, and loaded onto a HiLoad 26/600 Superdex 200 pg size exclusion column (Cytiva) equilibrated with storage buffer [20 mM Tris (pH 7.8), 200 mM NaCl, 5% (v/v) glycerol, 5 mM DTT, 2.5 mM EDTA]. Elution fractions were collected and analyzed by SDS-PAGE, pure fractions were pooled, and a 2:1 molar ratio of SRC-1 peptide was again added. The PXR LBD/peptide mixture was concentrated to 3 mg/mL, aliquoted, flash frozen in liquid nitrogen, and stored at −80°C.

PXR LBD (3 mg/mL, 83 μM) was mixed with 2 mM SJPYT-318 and incubated for 1 h at room temperature. This mixture contained 2% DMSO from the compound dilution. Hanging drop trays were set with 1 μL protein-ligand complex and 1 μL reservoir solutions containing 50 mM imidazole (pH 6.8-7.8) and 8-14% isopropanol. Crystals grew in various conditions within 1-3 days and were cryoprotected in respective mother liquors containing 1 mM SJPYT-318, 1% DMSO, and 40% ethylene glycol. The data presented were collected from a single crystal grown in 50 mM imidazole (pH 6.8) with 10% isopropanol.

X-ray diffraction data were collected to a resolution of 2.2 Å at Southeast Regional Collaborative Access Team (SER-CAT) 22-ID beamline at the Advanced Photon Source, Argonne National Laboratory. Frames were processed with XDS.⁴⁸ The crystals belonged to space group P4₃2₁2 with one molecule in the asymmetric unit. The structure was solved by molecular replacement in Phaser⁴⁹ using PDB ID 1M13 as the search model.⁵⁰ The search model was stripped of solvent and ligand prior to molecular replacement. Iterative cycles of model building and refinement were performed in Coot⁵¹ and Phenix.⁵² The data collection and refinement statistics are shown in Table 3. All crystallographic Figures were made in PyMOL (Schrodinger). The structure is deposited as PDB ID 8SZV.

QUANTIFICATION AND STATISTICAL ANALYSIS

All plots were made in GraphPad Prism 9. Results are expressed as the mean ± standard deviation from at least three independent experiments, and basic dose response curves were fitted as needed. To appropriately reflect both potency and efficacy, area under the curve (AUC) values were calculated in GraphPad Prism.

REAGENT or RESOURCE	SOURCE	IDENTIFIER
Antibodies
LanthaScreen Tb-anti-GST Antibody	Thermo Fisher Scientific	Cat#PV3550
Bacterial and virus strains
TurboCells BL21 (DE3) Competent E. coli	Genlantis	Cat#C302020
Chemicals, peptides, and recombinant proteins
Dimethyl sulfoxide	Thermo Fisher Scientific	Cat#BP231-100
T0901317	Cayman Chemical	Cat#71810
Rosiglitazone	MedChemExpress	Cat#HY-17386
GW0742	MedChemExpress	Cat#HY-13928
SJPYT-299	This paper	N/A
SJPYT-300	This paper	N/A
SJPYT-301	This paper	N/A
SJPYT-302	Lin et al.27	N/A
SJPYT-303	This paper	N/A
SJPYT-304	This paper	N/A
SJPYT-305	This paper	N/A
SJPYT-315	Lin et al.27	N/A
SJPYT-316	Lin et al.27	N/A
SJPYT-317	Lin et al.27	N/A
SJPYT-318	This paper	N/A
SJPYT-319	Lin et al.27	N/A
SJPYT-320	This paper	N/A
Dulbecco's Modified Eagle's Medium (DMEM)	ATCC	Cat#30-2002
RPMI-1640 Medium	ATCC	Cat#30-2001
Leibovitz's L-15 Medium	ATCC	Cat#30-2008
DMEM, high glucose, HEPES, no phenol red	Thermo Fisher Scientific	Cat#21063029
Fetal bovine serum (FBS), Canadian Origin	Cytiva	Cat#SH30396.03
HyClone Charcoal/Dextran treated FBS, US Origin	Cytiva	Cat#SH30068.03
Hygromycin B (50 mg/mL)	Thermo Fisher Scientific	Cat#10687010
Zeocin Selection Reagent	Thermo Fisher Scientific	Cat#R25001
Geneticin Selective Antibiotic (G418 Sulfate) (50 mg/mL)	Thermo Fisher Scientific	Cat#10131027
Matrigel Growth Factor Reduced Basement Membrane Matrix	Corning	Cat#356230
Sodium Pyruvate (100 mM)	Thermo Fisher Scientific	Cat#11360070
MEM Non-Essential Amino Acids Solution (100X)	Thermo Fisher Scientific	Cat#11140050
NEBuilder HiFi DNA Assembly Master Mix	New England Biolabs	Cat#E2621S
KpnI-HF	New England Biolabs	Cat#R3142S
FAM-SRC1-B Peptide: FAM-(PEG)5-CPSSHSSLTERHKILHRLLQEGSPS-NH2	Lin et al.46	N/A
Magnesium chloride, 1M aq. soln., sterile-filtered	Thermo Fisher Scientific	Cat#J62411AD
Bovine Serum Albumin	Sigma-Aldrich	Cat#A3294
Ethylenediamine Tetraacetic Acid (EDTA)	Thermo Fisher Scientific	E478-500
Isopropanol	Thermo Fisher Scientific	Cat#A416-4
Ethylene glycol	Sigma-Aldrich	Cat#102466-500ML
Pluronic F-127, 0.2 μm filtered (10% Solution in Water)	Thermo Fisher Scientific	Cat#P6866
Lysozyme, Egg White	Gold Biotechnology	Cat#12650-88-3
Terrific Broth, Modified, Granulated, 5 Kilograms	Research Products International	Cat#T15100-5000.0
D-(+)-Glucose	Sigma-Aldrich	Cat#G8270
SIGMAFAST Protease Inhibitor Cocktail Tablets, EDTA-Free	Sigma-Aldrich	Cat#S8830
IPTG	Gold Biotechnology	Cat#I2481C100
Ampicillin (Sodium), USP Grade	Gold Biotechnology	Cat#A-301-100
DTT (Dithiothreitol) (> 99% pure) Protease free	Gold Biotechnology	Cat#DTT100
Tris Base Ultra Pure [Tris (Hydroxymethyl) Aminomethane]	Research Products International	Cat#T60040-1000.0
Sodium Chloride	Research Products International	Cat#S23020-500.0
Glycerol, 99+%, extra pure	Thermo Fisher Scientific	Cat#158920025
Imidazole, 99%	Thermo Fisher Scientific	Cat#122025000
HisTrap FF	Cytiva	Cat#17525501
HiLoad 26/600 Superdex 200 pg	Cytiva	Cat#28989336
FuGENE HD Transfection Reagent	Promega	Cat#E2311
ProxiPlate-384 Plus F, Black 384-shallow well Microplate	PerkinElmer	Cat#6008260
CulturPlate-384, White Opaque 384-well Microplate, Sterile and Tissue Culture Treated	PerkinElmer	Cat#6007680
PXR (SXR) Ligand Binding Domain Protein	Thermo Fisher Scientific	Cat#PV4841
LXR-a Recombinant Human Protein, Ligand Binding Domain, GST-Tagged	Thermo Fisher Scientific	Cat#PV4657
LXR-b Recombinant Human Protein, Ligand Binding Domain, GST-Tagged	Thermo Fisher Scientific	Cat#PV4660
His-tagged PXR LBD	This paper	N/A
Unlabeled SRC-1 Peptide: N-CPSSHSSLTERHKILHRLLQEGSPS-C	This paper	N/A
Critical commercial assays
CellTiter-Glo Luminescent Cell Viability Assay	Promega	Cat#G7572
steadylite plus Reporter Gene Assay System	PerkinElmer	Cat#6066751
LiveBLAzer FRET-B/G Loading Kit with CCF4-AM	Thermo Fisher Scientific	Cat#K1096
Maxwell 16 LEV simplyRNA Tissue Kit	Promega	Cat#AS1280
SuperScript VILO cDNA Synthesis Kit	Thermo Fisher Scientific	Cat#11754050
TaqMan Fast Advanced Master Mix	Thermo Fisher Scientific	Cat#4444557
TaqMan gene expression assay for CYP3A4	Thermo Fisher Scientific	Cat#Hs00604506_m1
TaqMan gene expression assay for ABCA1	Thermo Fisher Scientific	Cat#Hs01059118_m1
TaqMan gene expression assay for GAPDH	Thermo Fisher Scientific	Cat#Hs03929097_g1
TaqMan gene expression assay for RNA18S	Thermo Fisher Scientific	Cat#Hs03928990_g1
Deposited data
Crystal structure of PXR LBD bound to SJPYT-318	This paper	PDB ID 8SZV
Experimental models: Cell lines
293T/17 Cells	ATCC	Cat#CRL-11268
GeneBLAzer LXRα-UAS-bla HEK 293T Cells	Thermo Fisher Scientific	Cat#K1692
GeneBLAzer LXRβ-UAS-bla HEK 293T Cells	Thermo Fisher Scientific	Cat#K1699
GeneBLAzer PPARδ-UAS-bla HEK 293T Cells	Thermo Fisher Scientific	Cat#K1690
GeneBLAzer PPARγ-UAS-bla HEK 293H Cells	Thermo Fisher Scientific	Cat#K1701
SNU-C4 Cells	Korean Cell Line Bank	Cat#0000C4
SNU81 Cells	Korean Cell Line Bank	Cat#00081
SNU719 Cells	Korean Cell Line Bank	Cat#00719
LS1034 Cells	ATCC	Cat#CRL-2158
SNU620 Cells	Korean Cell Line Bank	Cat#00620
SW948 Cells	ATCC	Cat#CCL-237
SW1417 Cells	ATCC	Cat#CCL-238
SW1463 Cells	ATCC	Cat#CCL-234
Oligonucleotides
5’-TCTAGAGCGGCCGCAGGTACCAGTGAACGGACAGGGACTC-3’	Integrated DNA Technologies	N/A
5’-GGCCTTAGTTATTCAGGTACCTCAGCTACCTGTGATGCC-3’	Integrated DNA Technologies	N/A
Recombinant DNA
pG5luc	Promega	Cat#E2440
pBIND	Promega	Cat#E2440
pETDuet-1-PXR LBD-mSRC-1	Lin et al.32	N/A
pcDNA3-FLAG-PXR	Lin et al.45	N/A
pBIND-PXR LBD	This paper	N/A
pBIND-FXR LBD	This paper	N/A
pBIND-RORγ LBD	This paper	N/A
Software and algorithms
XDS	Kabsch48	https://xds.mr.mpg.de/
Phenix	Liebschner et al.52	https://phenix-online.org/
Coot	Emsley et al.51	http://www2.mrc-lmb.cam.ac.uk/personal/pemsley/coot/
PyMOL	Schrödinger	https://pymol.org/2/
Prism 9	Graphpad Software	https://www.graphpad.com/features

Highlights.

1. Analogs of a promiscuous nuclear receptor ligand are selective PXR agonists

2. Selectivity is achieved by utilizing PXR’s large spherical ligand binding pocket

3. Selective ligand binds PXR in new mode compared to parental compound