Discovery and characterization of a novel series of N-phenylsulfonyl-1H-pyrrole picolinamides as positive allosteric modulators of the metabotropic glutamate receptor 4 (mGlu₄)

Abstract

Herein we report the synthesis and characterization of a novel series of N-phenylsulfonyl-1H-pyrrole picolinamides as novel positive allosteric modulators of mGlu₄. We detail our work towards finding phenyl replacements for the core scaffold of previously reported phenyl sulfonamides and phenyl sulfone compounds. Our efforts culminated in the identification of N-(1-((3,4-dimethylphenyl)sulfonyl)-1H-pyrrol-3-yl)picolinamide as a potent PAM of mGlu₄.

The metabotropic glutamate receptors are Class C G-protein coupled receptors (GPCRs) and are further separated into three family classes based on their receptor structure, G-protein coupling and ligand selectivity (Group I: mGlu₁ and mGlu₅; Group II: mGlu₂ and mGlu₃; Group III: mGlu₄, mGlu₆, mGlu₇ and mGlu₈).¹ This family of GPCRs has received significant attention over the past 10 years as potential therapeutic targets for a number of CNS disorders, such as schizophrenia,^2–4 Fragile X,^5,6 generalized anxiety disorder^7,8 and Parkinson’s disease.^9–14 Until recently, the Group III family had received less attention than the Group I and Group II family receptors; however, many tool compounds for the mGlu₄ subtype have been disclosed recently from our laboratories,^15–17 and others.^18–20 A common motif in many of the disclosed compounds is the presence of an N-phenyl picolinamide core scaffold.^15,16,21,22 There have been reports of efforts aimed at discovering amide replacements²³; however, there have not been disclosures around replacement of the central phenyl ring system. Herein, we report our efforts towards phenyl ring replacements culminating in the discovery of a unique series of N-pyrrolesulfonamides as novel PAMs of mGlu₄.

The starting point for our exploration centered around a series of phenylsulfonamides and phenylsulfones that were previously disclosed (Fig. 1).^24,25 Our first attempts at replacing the phenyl group centered on a thiazole ring and utilized the sulfone moiety as in 2 in order to limit the number of H-bond donors in the molecule.²⁶ The synthesis started with the commercially available 5-bromo-thiazol-2-amine (4-Me, Et, or H) which was acylated to yield the picolinamide 5 (picolinyl chloride, DiPEA, CH₂Cl₂) (Scheme 1). Next, palladium catalyzed cross-coupling of the bromothiazole, 5, with an appropriately substituted benzylthiol (Pd₂(dba)₃, XantPhos, DIEA, 1,4-dioxane, 100 °C) yielded the N-(5-(benzylthio)-thiazol-2-yl)picolinamide, 6.²⁷ Lastly, oxidation of the sulfide to the sulfone (m-CPBA, CH₂Cl₂) yielded the desired compounds, 7a–h.

Figure 1.

Previously disclosed phenylsulfonamide and phenylsulfone mGlu₄ PAMs.

Scheme 1.

Reagents and conditions. (a) Picolinyl chloride·HCl, diisopropylethylamine, CH₂Cl₂, 24–39%; (b) Pd₂(dba)₃, XantPhos, HSCH₂Ar, diisopropylethylamine, 1,4-dioxane, 100 °C, 51–87%; (c) m-CPBA, CH₂Cl₂, 38–87%.

The SAR analysis of the thiazole sulfone analogs is summarized in Table 1. Gratifyingly, the thiazole is a tolerated replacement for the internal phenyl ring as the thiazole derivative, 7a, is equipotent with the phenyl derivative, 2 (7a, EC₅₀ = 189 nM vs 2, EC₅₀ = 237 nM). In addition to R¹ = H, the methyl and ethyl substituents are also tolerated in the 4-position (7b, EC₅₀ = 448 nM; 7e, EC₅₀ = 217 nM). Overall, the replacement of the phenyl group with a thiazole was a productive change in terms of potency with 7c being one of the most potent compounds discovered to date (EC₅₀ = 83 nM). The sulfide intermediates, 6, were tested and were significantly less potent (EC₅₀’s = 5000–8000 nM); the sulfoxides were not tested as the oxidation procedure led directly to the sulfones. Unfortunately, these compounds suffered from significant pharmacokinetic liabilities (poor brain penetration, metabolic instability) limiting the utility of these compounds to in vitro tool compounds.

Table 1.

SAR of the sulfonylthiazoles, 7a–h

Compd	R1	R2	mGlu4 EC50 a (nM)	pEC50 ± SEMa	%GluMaxb
7a	H		189	6.72 ± 0.13	30.6 ± 0.5
7b	Me		448	6.35 ± 0.06	94.8 ± 3.6
7c	H		83	7.08 ± 0.14	35.1 ± 1.2
7d	Me		4137	5.38 ± 0.06	112.5 ± 1.6
7e	Et		217	6.66 ± 0.09	119.9 ± 1.3
7f	Me		547	6.26 ± 0.05	93.5 ± 1.8
7g	Me		295	6.53 ± 0.09	99.3 ± 1.7
7h	Et		150	6.82 ± 0.12	109.0 ± 1.4

^aCalcium mobilization human mGlu₄ assay; values are the average of n = 3.

^bAmplitude of response in the presence of 30 µM test compound, normalized to a standard compound, PHCCC, and represented as %GluMax.

Next, we turned our attention to other replacement groups, namely, pyrrole and pyrazole.^28,29 The synthesis of these groups is detailed in Scheme 2. The nitro pyrrole (or pyrazole) was converted to the sulfonamide (DBU, RSO₂Cl) and then the nitro was reduced to the amino compound using Raney Nickel, 10 (EtOH, Ra–Ni, 40 psi H₂). The final compounds were completed via acylation of the amino group with the acid chloride (picolinyl chloride, DIEA) yielding the desired compounds, 11a–w.^30,31

Scheme 2.

Reagents and conditions. (a) DBU, ArSO₂Cl or BnSO₂Cl, CH₂Cl₂, 50–96%; (b) EtOH, Ra–Ni, 40 psi H₂, 2 h; 93–97%; (c) diisopropylethylamine, picolinyl chloride, CH₂Cl₂, 39–43%.

The initial set of compounds evaluated were the benzyl sulfonamides, comparators to the thiazole benzylsulfones. Similar to the thiazole core compounds, the pyrrole compounds were well tolerated, with the 2-chlorobenzyl sulfonamide being equipotent to the thiazole (7a) and phenyl (2) derivatives (11g, EC₅₀ = 174 nM). As seen previously, most of the benzyl compounds that we evaluated were active as mGlu₄ PAMs with most EC₅₀’s < 500 nM. Although these compounds were active as PAMs, they suffered from the same PK liabilities as the thiazole compounds, namely, metabolic stability issues. It was determined that oxidation of the benzylic CH₂ group was the major metabolic liability and efforts were undertaken to block this site. Thus, compounds 11i–k were synthesized. Unfortunately, these compounds were significantly less potent as mGlu₄ PAMs, with the mono-fluoro compound being the most active (11j, EC₅₀ = 1024 nM). In addition, blocking the presumed site of metabolism did not inherently improve the metabolic stability of these compounds, presumably due to a shift of the site of instability to other areas of the molecule (Table 2).

Table 2.

SAR of the pyrrole benzylsulfonamides, 11a–k

Compd	R1	R2	R3	mGlu4 EC50 a (nM)	pEC50 ± SEMa	%GluMaxb
11a		H	H	572	6.24 ± 0.04	98.6 ± 2.6
11b				188	6.72 ± 0.04	88.7 ± 4.4
11c				4179	5.38 ± 0.16	112.3 ± 5.3
11d				84	7.07 ± 0.04	122.4 ± 3.7
11e				398	6.40 ± 0.06	140.3 ± 1.7
11f				251	6.60 ± 0.07	117.2 ± 7.1
11g				174	6.76 ± 0.02	119.0 ± 3.5
11h				190	6.72 ± 0.04	108.9 ± 5.2
11i		Me	H	2970	5.53 ± 0.26	59.6 ± 12.7
11j		F	H	1021	5.99 ± 0.02	71.0 ± 2.6
11k		F	F	>30,000	<4.5	33.5 ± 4.9

^aCalcium mobilization human mGlu₄ assay; values are the average of n = 3.

^bAmplitude of response in the presence of 30 µM test compound, normalized to a standard compound, PHCCC, and represented as %GluMax.

Next, we turned our efforts to eliminating the benzylic site altogether by evaluating a series of phenyl sulfonamides. Much like the benzyl derivatives, the phenylsulfonamides were well tolerated as a substituent. Alkyl substituted pyrrole sulfonamides were the most potent of the series (11l, EC₅₀ = 122 nM; 11m, EC₅₀ = 104 nM; 11r, EC₅₀ = 62 nM), with the halogen substituted analogs being less potent (e.g., 11n, EC₅₀ = 740 nM; 11o, EC₅₀ = 1238 nM). Some potency could be recaptured by the addition of an alkyl group as in 11t (EC₅₀ = 163 nM) and 11u (EC₅₀ = 268 nM). The pyrazole moiety was also tolerated as a phenyl replacement (Table 3).

Table 3.

SAR of the pyrrole phenylsulfonamides, 11l–w

Compd	X	R1	mGlu4 EC50 a (nM)	pEC50 ± SEMa	%GluMaxb
11l	CH		122	6.91 ± 0.03	88.8 ± 3.4
11m	CH		104	6.98 ± 0.04	84.8 ± 5.3
11n	CH		740	6.13 ± 0.03	78.8 ± 3.5
11o	CH		1238	5.91 ± 0.18	102.8 ± 6.2
11p	CH		440	6.36 ± 0.10	72.9 ± 2.5
11q	CH		344	6.46 ± 0.10	83.0 ± 0.3
11r	CH		62	7.21 ± 0.01	80.3 ± 4.3
11s	N		148	6.83 ± 0.04	92.9 ± 1.9
11t	CH		163	6.79 ± 0.05	72.2 ± 1.7
11u	CH		268	6.57 ± 0.06	98.9 ± 2.3
11v	N		242	6.62 ± 0.06	35.3 ± 3.0
11w	CH		739	6.13 ± 0.07	35.6 ± 3.7

^aCalcium mobilization human mGlu₄ assay; values are the average of n = 3.

^bAmplitude of response in the presence of 30 µM test compound, normalized to a standard compound, PHCCC, and represented as %GluMax.

Having established the pyrrole scaffold as a novel phenyl replacement as mGlu₄ PAMs, we next evaluated the compounds in our battery of Tier 1 in vitro PK assays (Table 4). The intrinsic clearance (CL_INT) was assessed in rat hepatic microsomes and the subsequent predicted hepatic clearance (CL_HEP) was calculated.^23,32 Many of the compounds displayed high intrinsic and predicted hepatic clearance, except for the 3,4-dimethylphenyl analog, 11r, which had moderate predicted hepatic clearance (CL_HEP = 38.3 mL/min/kg). Utilizing an equilibrium dialysis approach, the protein binding of the compounds was evaluated in rat plasma. The fraction unbound (F_u) of the analogs tested was very low, except, again, in the case of 11r which showed slightly better unbound fraction (F_u = 0.012).

Table 4.

In vitro pharmacokinetic data for selected N-phenylsulfonamide pyrroles

Compd	CL (mL/min/kg)		PPB (Fu)
	Rat CLINT	Rat CLHEP	Rat
11l	2704	68.2	0.009
11m	4035.1	68.8	0.001
11o	1126.4	65.9	0.004
11r	293	38.3	0.012
11s	3804	68.7	0.005
11t	3445	68.6	0.004

Lastly, we evaluated two analogs in an in vivo IV clearance experiment in order to assess whether the in vitro data would be predictive of in vivo PK (in vitro:in vivo correlation, IV/IVC). The two compounds chosen were 11r and 11t, one predicted to be moderately cleared and the other highly cleared. Both compounds were dosed IV (1 mg/kg) and the in vivo plasma clearance closely resembles the in vitro predicted hepatic clearance suggesting the major route of clearance to be hepatic oxidative metabolism. In order to more fully establish this route of metabolism, we co-dosed 11t with 1-aminobenzotriazole (ABT), a pan-irreversible inhibitor of the cytochrome P450s. After PO dosing, a significant amount of 11t was detected in the HPV (hepatic portal vein); however, only ~1% of that amount was detected in the plasma, indicating a significant amount of oxidative metabolism. The brain:plasma ratio was ~1:1, although the total amount in the brain was very low, which is in contrast to the phenylsulfones which displayed low brain:-plasma ratios. When co-dosed with ABT, the concentrations of 11t in the HPV were nearly identical with the non-ABT dosed animals; however, the concentrations detected in the plasma were significantly higher (198.5 ng/mL vs 7.4 ng/mL). Again, while the brain:plasma ratio was ~1:1; as with the total plasma concentrations, the total amount in the brain was higher when co-dosed with ABT (Table 5).

Table 5.

In vivo Rat PK IV clearance (1 mg/kg)

		11r	11t
CL (mL/min/kg)		37.2	86.8
T1/2 (min)		252	167
MRT (min)		149	84.7
Vss (L/kg)		5.5	7.4
AUC (h ng/mL)		447.5	191.7
po dosing of 11t with and without ABT, 1 h
Dose (mg/kg)	Concentration (ng/mg or g)
	Plasma systemic	Plasma HPV	Brain
10/Veh	7.4	731	8.1
10/ABT	198.5	764.5	175.5

In conclusion, we report the discovery and characterization of unique thiazolesulfone and pyrrolesulfonamide core scaffolds as mGlu₄ PAMs. These series were synthesized as a replacement for the previously disclosed phenylsulfonamide/phenylsulfone compounds. The thiazolesulfone analogs were potent mGlu₄ PAMs (EC₅₀’s < 500 nM), but suffered from metabolic instability. The pyrrolesulfonamide scaffold also was a well-tolerated change with many compounds exhibiting EC₅₀’s < 500 nM. Compounds from this scaffold displayed an IV/IVC with regards to clearance (CLp) following IV adminstration; however, the compounds were not orally bioavailable and displayed poor exposure following PO dosing (systemic:HPV ratio <0.01). Co-dosing in vivo with the pan-irreversible P450 inhibitor, ABT, blocked extensive P450-mediated metabolism and significantly improved the systemic plasma and brain concentrations (plasma systemic:HPV ratio 0.26).