Campaign to identify novel modulators of the Retinoic acid receptor-related Orphan Receptors (ROR)

Relevant biology of the probes

RORα: The protein encoded by this gene is a member of the NR1 subfamily of nuclear hormone receptors. It can bind as a monomer or as a homodimer to hormone response elements upstream of several genes to enhance the expression of those genes. The specific functions of this protein are not known, but it has been shown to interact with NM23-2, a nucleoside diphosphate kinase involved in organogenesis and differentiation, as well as with NM23-1, the product of a tumor metastasis suppressor candidate gene. Four transcript variants encoding different isoforms have been described for this gene.

RORγ: The protein encoded by this gene is a DNA-binding transcription factor and is a member of the NR1 subfamily of nuclear hormone receptors. The specific functions of this protein are not known; however, studies of a similar gene in mice have shown that this gene may be essential for lymphoid organogenesis and may play an important regulatory role in thymopoiesis. In addition, studies in mice suggest that the protein encoded by this gene may inhibit the expression of Fas ligand and IL2. Two transcript variants encoding different isoforms have been found for this gene.

Probe SID-85257301 is a novel RORα/γ inverse agonist probe that binds to and decreases the transcriptional activity of both RORα and RORγ receptors. The second probe, SID-85257298 decreases RORα transcriptional activity and lacks activity for LXR and RORγ.

2.1. Assays

(Click on the hyperlinks to obtain itemized protocols directly from PubChem; see also Summary AID-2139)

2.2. Probe Chemical Characterization

Synthetic route. The RORα and RORα/γ inverse agonist probes were synthesized as summarized below.

Probe 1 (ML125): N-[4-(1,1,1,3,3,3-hexafluoro-2-hydroxypropan-2-yl)phenyl]-N-(2,2,2- trifluoroethyl)benzenesulfonamide.

Probe 2: (ML124): 4-tert-butyl-N-[4-(1,1,1,3,3,3-hexafluoro-2-hydroxypropan-2-yl)phenyl]benzenesulfonamide.

Structure verification with 1H NMR and LCMS results

Probe 1 (ML125) was obtained as a white powder (m.p. 105°C) with 99% purity (HPLC analysis): ¹H NMR (400 MHz, (CD₃)₂SO): 4.63 (q, J = 9.0 Hz, 2H), 7.29 (d, J = 8.8 Hz, 2H), 7.55–7.64 (m, 4H), 7.66 (d, J = 8.8 Hz, 2H), 7.73 (tt, J = 7.1, 1.6 Hz, 1H), 8.82 (s, 1H). ¹³C NMR (100 MHz, (CD₃)₂SO): 127.3 (2C), 127.7 (2C), 128.7 (2C), 129.4 (2C), 130.7, 133.8, 137.1, 140.4. Five carbon resonances are missing in the ¹³C spectrum of SID-85257301. These are the three carbons of the hexafluoropropanol moiety and the two carbons of the trifluoroethyl group. The fluorine coupling with these carbons gives multiplets which were difficult to detect in the ¹³C spectrum even with increased number of scans. FTIR: 3430, 1510, 1450, 1425, 1343, 1269, 1229, 1215, 1174, 1150, 1138, 1108, 1086, 976, 953, 928, 855, 769, 752, 735, 716, 705, 686, 673 cm⁻¹. MS (ES-) m/z = 480 (found for C₁₇H₁₂F₉NO₃S-H⁺).

Probe 2 (ML124) was obtained as a white powder (m.p. 161°C) with >98% purity (HPLC analysis): ¹H NMR (400 MHz, (CD₃)₂SO): 1.26 (s, 9H), 7.22 (d, J = 8.8 Hz, 2H), 7.53 (d, J = 8.8 Hz, 2H), 7.59 (d, J = 8.8 Hz, 2H), 7.76 (d, J = 8.8 Hz, 2H), 8.58 (s, 1H), 10.66 (s, 1H). ¹³C NMR (100 MHz, (CD₃)₂SO): 30.7 (3C), 34.9, 118.3 (2C), 125.2, 126.2 (2C), 126.5 (2C), 127.9 (2C), 136.8, 139.6, 156.10. Three carbon resonances are missing in the ¹³C spectrum of SID-85257298. These are the carbons of the hexafluoropropanol moiety. The fluorine coupling with these carbons gives multiplets which were difficult to detect in the ¹³C spectrum even with increased number of scans. FTIR: 3430, 3245, 1517, 1471, 1403, 1332, 1308, 1267, 1228, 1192, 1162, 1150, 1112, 1088, 963, 926, 827, 752, 735, 702, 664 cm⁻¹. MS (ES-) m/z = 454 (found for C₁₉H₁₉F₆NO₃S-H⁺).

Solubility. The solubility of the probes was measured in phosphate buffered saline (PBS: 137 mM NaCl, 2.7 mM KCl, 10 mM sodium phosphate dibasic, 2 mM potassium phosphate monobasic and a pH of 7.4) at room temperature (23°C). The solubility of probe ML124 and ML125 was found to be 16 μM and 3 μM, respectively.

Stability. The stability of the probes was measured at room temperature (23ºC) in PBS (no antioxidants or other protectants; DMSO concentration below 0.1%). The stability, represented by the half-life, was found to be > 48 hours for both probes. Below is a graph showing loss of compound with time over a 48 hour period with a minimum of 6 time points. The table indicates the percent of compound remaining at the end of the 48 hours.

Table

Table

Table

The probes were measured for their ability to form glutathione adducts. At concentrations of 100 μM reduced GSH, 10 μM of the probes do not appear to be Michael acceptors [21, 22].

CID, SID and ML# of the probe and five related analog samples submitted to the SMR collection

Table

2.3. Probe Preparation

Detailed experimental procedures for synthesis of Probe 1 (ML125) and Probe 2 (ML124) follow.

To a solution of 2-(4-aminophenyl)-1,1,1,3,3,3-hexafluoropropan-2-ol¹ (1.5M in THF, 1.29 mL, 1.93 mmol) in CH₂Cl₂ (2.7 mL) under argon were successively added at 0°C N, N-Diisopropylethylamine (370 μL, 2.12 mmol) and trifluoroacetic anhydride (268 μL, 1.93 mmol). The mixture was stirred for 10 min at 0°C, then 2 hours at room temperature. The solution was quenched by the addition of saturated NH₄Cl solution and the organic phase was washed with saturated NH₄Cl solution, brine, dried over Na₂SO₄, filtered and concentrated under reduce pressure to give 630 mg of 2,2,2-trifluoro-N-{4-[2,2,2-trifluoro-1-hydroxy-1-(trifluoromethyl)ethyl]phenyl}acetamide (92%) as purple oil. The crude residue was used without purification for the next step. ¹H NMR (400 MHz, (CD₃)₂SO): 7.72 (d, J = 8.8 Hz, 2H), 7.81 (d, J = 8.8 Hz, 2H), 8.75 (s, 1H), 11.46 (s, 1H).

To a solution of 2,2,2-trifluoro-N-{4-[2,2,2-trifluoro-1-hydroxy-1-(trifluoromethyl)ethyl]phenyl}acetamide (630 mg, 1.77 mmol) in Et₂O (2.7 mL) under argon was portion-wise added at 0°C lithium aluminum hydride (95%, 709 mg, 17.7 mmol). The mixture was stirred for 30 min at 0°C, then overnight at room temperature. The solution was cooled down at 0°C and quenched by H₂O (800 μL), followed by NaOH (1N, 1.6 mL), then H₂O (3*800 μL). The mixture was stirred 30 min at room temperature, filtered over Celite, washed by Et₂O and the filtrate was concentrated under reduce pressure to give 414 mg of 1,1,1,3,3,3-hexafluoro-2-{4-[(2,2,2-trifluoroethyl)amino]phenyl}propan-2-ol (69%) as purple solid. The crude residue was used without purification for the next step. ¹H NMR (400 MHz, (CD₃)₂SO): 3.52 (m, 2H), 6.60 (t, J = 6.8 Hz, 2H) 6.81 (d, J = 8.8, 2H), 7.38 (d, J = 8.8 Hz, 2H), 8.30 (s, 1H).

Synthesis of Probe 1 (ML125)

Probe 1, SID-85257301

To a solution of 1,1,1,3,3,3-hexafluoro-2-{4-[(2,2,2-trifluoroethyl)amino]phenyl}propan-2-ol (630 mg, 0.42 mmol) in pyridine (1.4 mL) under argon was added at room temperature benzenesulfonyl chloride (80 mg, 0.46 mmol). The mixture was heated overnight at 100°C, then diluted by toluene and concentrated under reduced pressure. The crude product was directly applied to a silica gel column and eluted with hexane-EtOAc (85/15) to obtain 121 mg of Probe 1 (SID-85257301) (60% yield, purity 99% by HPLC analysis) as a white powder (m.p. 105°C). Spectroscopic characterization data for Probe 1 (ML125) are provided in Section 2.2 above.

Synthesis of Probe 2 (ML124)

Probe 2, SID-85257298

To a solution of 2-(4-aminophenyl)-1,1,1,3,3,3-hexafluoropropan-2-ol (1.5M in THF, 128 μL, 0.193 mmol) in acetone (643 μL) were successively added at room temperature 2,6-Lutidine (29 μL, 0.251 mmol) and 4-tert-butylbenzenesulfonyl chloride (46 mg, 0.193 mmol). The mixture was heated overnight at 80°C, then diluted by AcOEt and quenched at room temperature by the addition of saturated NaHCO₃ solution. The aqueous phase was extracted two times by AcOEt and combined organic phases were dried over Na₂SO₄, filtrated and evaporated. The crude residue was purified by silica gel column and eluted with hexane-EtOAc (70/30) to obtain 64 mg of Probe 2 (SID-85257298) (73% yield, purity >98% by HPLC analysis) as a white powder (m.p. 161°C). Spectroscopic characterization data for Probe 2 (ML124) are provided in Section 2.2 above.

3.1. Summary of Screening Results

This project was initiated as an outreach effort. During this outreach effort 64,925 substances from the MLPCN collection were tested in primary screening in singlicate to identify compounds that inhibit RORα activity (AID-561). A total of 278 substances were identified as active in the primary screen. The 273 available hits from this assay were then screened in titration assays in triplicate to determine RORα potency (AID-610). No RORα specific inverse agonists were identified and the project was pursued as a Center-based component (CBC). The SRIMSC CBC focuses on the use of a genomic screening platform to support selectivity profiling of high value compounds that emerge from the MLPCN program. One area of focus for the CBC was the development of a nuclear receptor (NR) library containing expression vectors for all 48 human NRs in a GAL4 format. In an effort to validate the NR library (AID-2277), we screened it against a small chemical library containing mostly well characterized NR modulators. Analysis of the results from this validation screen revealed that the LXR agonist T0901317 (SID-85257301) was capable of repressing the activity of both GAL4-RORα and GAL4-RORγ, but not that of GAL4-RORβ. The activity of T0901317 was confirmed on wildtype receptor on native RORα/γ promoters and direct binding to these receptors was demonstrated. This is an excellent example of the utility of the tools being developed within the SRIMSC CBC.

Compounds derived from these initial candidates were purchased as powders or synthesized at the SRIMSC and were tested for their ability to inhibit RORα and RORγ in luciferase-based reporter assays performed at a single concentration of 10 μM or in dose response assays starting at a nominal concentration of 20 μM. Compounds were subsequently counterscreened at a single concentration and/or dose response assays against the liver X receptor (LXR), the farnesoid X receptor (FXR), and glucose-6-phosphatase to determine selectivity. Finally, compounds of interest were tested at a single concentration of 10 μM against VP16 to determine whether they were non-selective or cytotoxic.

The Center-based probe development efforts resulted in the identification of two probes. The benzenesulfonamide compound T0901317 (SID-85257301) previously identified as a selective agonist of LXR (4) was identified here as a novel RORα/γ inverse agonist probe that decreases the transcriptional activity of both ROR receptors (IC50 values = 2.0 and 1.73 micromolar, respectively). The second probe, SID-85257298 synthesized at the SRIMSC, was found to decrease RORα transcriptional activity (IC50 value = 2.47 micromolar). SID-85257298 represents an improvement over the prior art due to its lack of activity for LXR. Probe compound SID-85257298 does not have activity against RORγ (IC50 > 20 micromolar). These two probes are useful tools for examining ROR biology.

Following the screening and hit validation campaign, a probe compound (SID-85257301) belonging to the benzenesulfonamide scaffold was identified. We searched the MLSMR and commercial sources for structurally related compounds. Analogs were purchased in powder form or re-ordered from the MLSMR in liquid form and tested in dose response assays against both RORα and RORγ. Additional compounds were synthesized, by using the synthesis scheme summarized in the following section, and were also tested in dose response assays against both RORα and RORγ, as well as other nuclear receptors such as LXR and FXR. Results of these assays are summarized in the SAR table below. We found that deletion of the N-trifluoroethyl substituent and replacement of the original N-benzenesulfonyl group of the original screening hit (SID-85257301) with a Np- tert-butylbenzenesulfonyl group gave a RORα selective inverse agonist (SID-85257298) that was inactive against both anti-targets, LXR and FXR. The assays have been summarized in Summary AID-2139.

3.2. Dose Response Curves for Probes

Dose-response curves of RORα and RORα/γ inhibitor probes ML125 and ML124 in the X cell-based assay

Ten-point, 1:3 serial dilution of probe compound ML125 (SID-85257301; A and B) and probe compound ML124 (SID-85257298; C and D) were tested in six replicate in both the RORα (A and C) and RORγ (B and D) inhibition assays the protocols described in the technical section of this report (see AID-2139 and AID-2117, respectively). Error bars represent the standard deviation of six replicates.

3.3. Scaffold/Moiety Chemical Liabilities

There are no known chemical liabilities known for Probe 1 (ML125) and Probe 2 (ML124). While both compounds have 6 and 9 fluorine atoms, respectively, both have acceptable solubility in PBS (3 and 16 mM, respectively—see Section 2.2).

Analogs were synthesized or purchased in powder form or re-ordered from the MLSMR in liquid form and tested in dose response assays against RORα, RORγ, VP16, LXR, and FXR. Results for these compounds are summarized in the SAR table below and in AID-2117.

Analogs of T0901317 (SID-85257301) were purchased as powders or synthesized at the SRIMSC and were tested for their ability to inhibit RORα and RORγ in luciferase-based reporter assays performed at a single concentration of 10 μM or in dose response assays starting at a nominal concentration of 20 μM. Compounds were subsequently counterscreened at a single concentration and/or dose response assays against the liver X receptor (LXR), the farnesoid X receptor (FXR), and glucose-6-phosphatase to determine selectivity. Finally, compounds of interest were tested at a single concentration of 10 μM against VP16 to determine whether they were non-selective or cytotoxic. This led to the identification of CID-44237404 that was declared as a first generation selective RORα inverse agonist probe ML124 (SR-03000000995; SID-85257298: MLS-002554296).

Efforts to identify a more potent RORα-selective inverse agonist have continued. We have explored changes in the arylsulfonamide unit, the aniline unit, and the linker connecting the two. The SAR Table below provides structures of more than new 30 analogs synthesized or purchased to explore these points; data for CID- 44237404 and the analogs declared in the first –generation RORα probe report are also included. From these results it is apparent that the sulfonamide unit cannot be replaced by a carboxamide, and thus far all substituents introduced to replace the bis(trifluoromethyl)carbinol unit of the original probe have eliminated or substantially reduced activity. The most productive change has been the discovery that replacement of the arylsulfonamide unit of probe ML124 (CID-44237404) with a thiophenyl sulfonamide group, as in SR- 06000113335 (“SR-3335”) has led to a ca. 5-fold increase in potency as an RORα inverse agonist, while maintaining selectivity vs. RORγ (inactive). [Note that this SR-3335 compound is a next generation selective RORα probe discussed in a separate probe report]. However, introduction of additional substituents on the thiophene ring have led to decrease in activity as RORα inverse agonists, and has led to re-emergence of activity against RORγ. The SAR table summarizing these results is attached below. In future work, we will continue to explore structural modifications of this inverse agonist series, with the objective of further increasing potency while maintaining selectivity for RORα vs. RORγ.

Subsequent SAR efforts were performed after the present ML124 and ML125 probe report was submitted, Those efforts led to the identification of an improved, selective second-generation RORα inverse agonist probe, ML176 [also known as SR-06000113335 (“SR-3335”)]. ML176 is ca. 5-fold more potent than ML124 as a RORα inverse agonist, while maintaining electivity vs. RORγ (inactive). More than 30 analogs of ML124 and ML176 are reported in the second generation ML176 probe report.

3.4. SAR Table

Table

3.5. Cellular Activity

The probes were tested in a variety of cell-based assays performed by the SRIMSC and assay provider. These assays were performed to determine the probe’s selectivity and potency. The results of these studies demonstrated that probe ML125 is a selective RORα inverse agonist, while probe ML124 is a dual RORα/γ inverse agonist. Further, probe ML124 was were tested in a variety of cell-based assays performed by the SRIMSC and assay provider. These assays were performed to determine the probe’s selectivity and potency. The results of these studies demonstrated that probe ML125 is a selective RORα inverse agonist, while probe ML124 is a dual RORα/γ inverse agonist.

ML124 Suppresses RORα and RORγ mediated IL-17 transcription

HEK293 cells were transiently transfected with the IL-17-dependent reporter construct, renilla luciferase, and vectors containing full-length RORα (+RORα), full-length RORγ (+RORγ), or empty vector alone (endogenous, “no receptor”). Twenty-four hours later cells were treated with DMSO or increasing concentrations of ML124. Twenty-four hours post-treatment, IL-17 activity was determined by luciferase assay. The data are normalized to the vehicle (DMSO) treated cells.

Modulation of Glucose-6-Phosphatase (G6Pase) promoter activity by ML124 in HepG2 cells

Sequential chromatin immunoprecipitation (ChIP/reChIP) assay illustrating that probe ML124 (10 μM) treatment reduces the ability of RORα to recruit SRC-2 to a G6Pase gene promoter. HepG2 cells overexpressing Flag-tagged RORα were treated with vehicle (DMSO) or 10 μM ML124 for 24 h followed by sequential ChIP. The first immunoprecipitation was performed using α-Flag antibody and the second immunoprecipitation was performed using α-SRC-2 antibody. Mouse IgG was used as a negative control and anti-RNA pol II antibody was used as a positive control.

3.6. Profiling Assays

Human Nuclear Receptor Profiling Assay. In addition to the above cellular assays, we have also obtained profiling results for the probe at 2 μM using a 48 human nuclear receptor library. This assay has been published in PubChem as AID-2277. The nuclear receptor library was plated into 384-well plates. HEK293T cells were reverse transfected with the well-specific construct and the UAS luciferase reporter pGL4.31 using Fugene6 transfection reagent in a final volume of 40 microliters. Control wells containing constructs encoding for the GAL4 DBD alone (pBind) or GAL4 fused to VP16 were also analyzed. After 24 hours, optimized compounds (2 micromolar final concentration) or DMSO was added to the plates and allowed to incubate for 20 hours. Next, 40 microliters of BriteLite was added to all wells and luciferase activity was measured on the PerkinElmer Envision 2104. Compounds that attenuate the GAL4-VP16-dependent luciferase activity are considered promiscuous or cytotoxic. To correct for plate-to-plate variance, sample data was normalized to wells containing vector only (69 wells) to determine Normalized Data Values (NDV). For each nuclear receptor, the average of NDV treated with a particular test compound was divided by the average NDV of wells treated with DMSO. This resulted in a fold-activation or fold-inhibition value for each receptor-compound pair. Because the activity of VP16 should not be impacted by these compounds and a change in this value is indicative of cell toxicity or general disruptions in transcription/translation, this fold-change value was then normalized to the VP16 fold-change. Compounds that induced in any receptor an average fold change that was three standard deviations from the VP16 value were considered active for that particular receptor.

Gal4 Nuclear Receptor Profiling of Ml125. Gal4 NR clones were reverse transfected with a UAS reporter construct into HEK293T cells. After 24 hours, the LXRα agonist T0901317 (2 μM final concentration) or DMSO was added and incubated for 20 hours. The luciferase activity of each construct was measured and normalized to the mock (vector alone), then the fold change in signal compared to DMSO is calculated (n=6).

Table

4.1. Comparison to existing art and how the new probe is an improvement

Cholesterol, cholesterol sulfate, and 25-hydroxycholesterol have been shown to be ligands of the RORs. Although these molecules bind tightly to the receptor, none alters the transcriptional output of the RORs in cell-based assays suggesting that while some oxysterols bind to RORs, they do not induce a conformational change in the ligand binding domain that would be required to alter coactivator or corepressor interaction. Therefore, prior to this probe report there were no published functional modulators of the RORs. Regardless, sterols would represent a limited chemical starting point for optimization of potency and isoform selectivity.

Chemical optimization of ML125

A recent survey of the patent literature does not reveal any specific patent applications on synthetic ROR modulators. While there are patents filed on ML125, they are all specific to LXR activity. Additionally, the modifications made to obtain selectivity over LXR as demonstrated with ML124, render these compounds inactive on LXR thus they do not appear as examples in the ML125 patent material. We have filed a provisional application which is currently under conversion to non-provisional on the general scaffold covering ML124 as selective modulators of RORα and RORγ.

Table

4.2. Mechanism of Action Studies

Modulation of RORα mediated Glucose-6-Phosphatase (G6Pase) promoter activity. In addition to the cellular and profiling assays above, it was important to determine whether the probes ML125 and ML124 could block expression of relevant ROR target genes in cells. In these assays 293T cells were co-transfected with pS6 control plasmid or pS6 containing full length RORα along with G6Pase promoter. SRC-2 as a coactivator was also co-transfected with G6Pase promoter. Dose-response curve was determined by treating the transfected cells with varying concentrations of compound for 20 hours. Luciferase activity was measured and relative change was determined by normalizing to cells treated with vehicle only. Each data point was performed in eight replicates and represented as mean ±SEM, n=8. The maximum concentration tested was 10 μM.

4.3. Planned Future Studies

In future work, we will continue to explore structural modifications of this inverse agonist series, with the objective of further increasing potency while maintaining selectivity for RORα vs. RORγ. ML125 represents a chemical tool that has been studied extensively and is useful to scientists that are interested in pharmacological observations made that are not easily attributable to LXR and FXR. ML124 represents a probe that is significantly less well studied, but is devoid of LXR and FXR activity. We have ongoing studies showing efficacy of ML124 and close analogs in animal models of diabetes 20 and autoimmune disorders. Subsequent to this probe report, SAR studies have uncovered RORα selective analogs. While these compounds have modest potency in cellular and biochemical assays, they show efficacy in animal models. Regardless, future effects are focused on structural studies and additional SAR to drive potency. More recent, we have uncovered a unique scaffold that offers RORγ isoform selectivity. Thus, the two probes presented here have provided an excellent foundation for the development of chemical probes to explore the biology of the RORs.