Bioorganic & Medicinal Chemistry Letters 13 (2003) 2625–2628 Hit-to-Lead Studies: The Discovery of Potent, Orally Bioavailable Triazolethiol CXCR2 Receptor Antagonists Andrew Baxter,* Colin Bennion, Janice Bent, Kerry Boden, Steve Brough, Anne Cooper, Elizabeth Kinchin, Nicholas Kindon, Tom McInally, Mike Mortimore, Bryan Roberts and John Unitt AstraZeneca R&D Charnwood, Bakewell Road, Loughborough LE11 5RH, UK Received 11 April 2003; revised 4 June 2003; accepted 5 June 2003 Abstract—A Hit-to-Lead optimisation programme was carried out on the high throughput screening hit, the triazolethiol 1, resulting in the discovery of the potent, orally bioavailable triazolethiol CXCR2 receptor antagonist 45. # 2003 Elsevier Ltd. All rights reserved. The use of High Throughput Screening (HTS) is now widespread in the pharmaceutical industry. There was an expectation that, once a screen was established for a particular target, then potent lead compounds or candidate drugs would be found. The reality is often far from this. Bridging the gap between the end of a HTS and the start of a full Lead Optimisation (LO) project has been described as Hit-to-Lead (HtL).1 Hits from HTS are profiled and compared to a generic target lead criteria. The lead target profile used is shown in Figure 1. Lead series then have a balance of properties — potency and SAR an encouraging metabolic and selectivity profile—such that rapid (less than 2 years) further optimisation should provide Candidate Drugs (CDs). Chemokines play an important role in immune and inflammatory responses in various diseases and dis- Figure 1. Hit-to-Lead generic lead target profile. *Corresponding author. Tel.: +44-1509-644772; fax: +44-1509645513; e-mail: andrew.jg.baxter@astrazeneca.com 0960-894X/03/$ - see front matter # 2003 Elsevier Ltd. All rights reserved. doi:10.1016/S0960-894X(03)00561-4 orders, including asthma and allergic disease, as well as autoimmune pathologies such as rheumatoid arthritis and atherosclerosis. These small secreted molecules are a growing superfamily of 8–14 kDa proteins characterised by a conserved four-cysteine motif. The chemokine superfamily can be divided into two main groups exhibiting characteristic structural motifs, the Cys-X-Cys (CXC) and Cys-Cys (CC) families.2 The CXC chemokines include several potent chemoattractants and activators of neutrophils such as interleukin-8 (IL8), GROa and neutrophil activating peptide 2 (NAP2). Studies have demonstrated that the actions of the chemokines are mediated by subfamilies of G proteincoupled receptors, among which are the receptors designated CCR1 to CCR10 and CXCR1 to CXCR5. These receptors represent good targets for drug development since agents which modulate these receptors would be useful in the treatment of disorders and diseases such as those mentioned above. The CXCR2 receptor is one of the receptors that IL8 activates and antagonists of this receptor should find particular use in the treatment of a number of diseases. A HTS was undertaken to identify compounds that blocked the binding of [125I]-IL8 to human recombinant CXCR2 (hrCXCR2) expressed in HEK 293 membranes using a Scintillation Proximity Assay (SPA). Subsequently at the hit evaluation stage compounds with binding inhibition seen in the HTS SPA assay were initially validated in a more conventional filter wash 2626 A. Baxter et al. / Bioorg. Med. Chem. Lett. 13 (2003) 2625–2628 hrCXCR2 [125I]-IL8 binding assay. Confirmation of functional antagonism was shown by blockade of GROa stimulated intracellular calcium mobilisation in isolated human neutrophils using a Fluorescence Imaging Plate Reader (FLIPR).3 and its size and steric environment are essential for receptor binding as smaller or sterically demanding acidic groups (OH, NHAc and NHSO2Ph) are inactive. All subsequent analogues prepared in HtL preserved the triazolethiol structure. The triazole 1 emerged inter alia from the CXCR2 HTS and, while having only modest potency in the binding assay (IC50 4.6 mM), had demonstrable and comparable functional activity in the FLIPR assay (IC50 2.4mM). The profile of this CXCR2 hit is in Table 1 compared with the generic lead target profile. The DMPK profile was satisfactory at this stage and HtL focused on exploring SAR to increase potency and assessing the importance of the thiol-thione functionality. Table 2. CXCR2 antagonist binding potencies Table 1. Hit profile of triazolethiol 1 2 3 4 5 6 7 8 a Generic lead criteriaa Binding IC50 <0.1 mM Ca Flux IC50 <0.1 mM Rat hepatocyte Cl <14 Human microsome Cl <23 Rat iv Cl <35 Rat iv Vss>0.5 L/kg Rat iv T1/2 >0.5 h Molecular weight< 450 ClogP<3.0 Triazolethiol 1 4.6 mM 2.4 mM 19 13 12 1.3 3.4 268 3.4 a Units as Figure 1 where not stated. Replacement of the pyridinyl ring by phenyl (compound 2)4 gave a slightly more potent analogue which was the starting point for variation of the thiol (Table 2). Analogues with hydroxy (3), amino (4), and without the thiol (5) were prepared using literature methods;5
7 all were inactive as CXCR2 antagonists. Methylation of the thiol (6) and acetylation (7) and benzenesulphonylation (8) of the amino also gave inactive compounds. It is assumed that the acidic nature of the thiol (pKa  6.5) R X CXCR2 IC50 (mM)a PhCH2 PhCH2 PhCH2 PhCH2 PhCH2 PhCH2 PhCH2 SH OH NH2 H SMe NHAc NHSO2Ph 2.4 NA NA NA NA NA NA NA,<50% inhibition at 10 mM. Next variation of the 2-benzyl substituent was investigated. Published routes to these compounds utilise the reaction of 2-substituted thiosemicarbazides with acid chlorides,4 or acylisocyanates with hydrazines7 followed by cyclisation (Scheme 1a and b). The compounds 9–14 in Table 3 were prepared from commercially available starting materials and offered no improvements in potency over 2-benzyl. Phenyl, methyl and hydrogen were all inactive and only the 3-hydroxybenzyl analogue 14 had any activity. The substitution on the 5-phenyl was investigated next, keeping the 2-benzyl constant, using the route described above. Table 3 lists the compounds prepared (15–37) and their activity. Various substitutions were seen as potency enhancing, in particular 2-chloro (34), 2,4-dichloro (35), 2-bromophenyl (36) and 2,3-dichloro (37) gave 10-fold increases in potency compared to the unsubstituted phenyl (2). In order to look for further variations of substituents on the 2-benzyl group, it was necessary to find an alternative route that would allow simpler starting materials Scheme 1. (i) Pyridine, 16 h, rt; (ii) 1 M NaHCO3, 16 h, 100  C; (iii) Et3N, toluene, 1 h, 80  C; (iv) aldehyde, MeOH/HCl; (v) Et3SiH, TFA, 0  C; (vi) EtOH/HCl, NH4SCN, 18 h, 75  C. A. Baxter et al. / Bioorg. Med. Chem. Lett. 13 (2003) 2625–2628 to be used. Benzylation of the 3-H compound (11) proceeds first on sulphur and a di-benzylation followed by S-debenzylation route did not work in our hands. The route followed6 is shown in Scheme 1c. Substituted aldehydes were condensed with benzoylhydrazides to give the corresponding hydrazones. These hydrazones were then reduced using triethylsilane in trifluoroacetic acid to give the substituted benzoylhydrazines. Reaction with ammonium thiocyanate in hot ethanol containing hydrogen chloride gave the intermediate semicarbazides, which were cyclised in hot sodium bicarbonate solution to give the required triazolethiols (38–45). A number of analogues were prepared keeping the 5substituent constant as 2,4-dichlorophenyl; most potent of these was the 3-chlorobenzyl having IC50 0.092 mM. Checking back to other potent 5-substituents led to the preparation of the lead compound, the 2-(3-chlorobenzyl)-5-(2-chlorophenyl) analogue (45) (Table 3). Some preliminary SAR conclusions can be drawn on the basis of the compounds prepared in this paper. The substitution patterns, orientation in space and presence Table 3. CXCR2 antagonist binding potencies 2 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 a R R0 CXCR2 IC50 (mM)a PhCH2 Ph Me H 4-CF3PhCH2 2-ClPhCH2 3-OHPhCH2 PhCH2 PhCH2 PhCH2 PhCH2 PhCH2 PhCH2 PhCH2 PhCH2 PhCH2 PhCH2 PhCH2 PhCH2 PhCH2 PhCH2 PhCH2 PhCH2 PhCH2 PhCH2 PhCH2 PhCH2 PhCH2 PhCH2 PhCH2 4-MeOPhCH2 3-MeOPhCH2 3-MePhCH2 PhCH2CH2 4-ClPhCH2 3-PhOPhCH2 3-ClPhCH2 3-ClPhCH2 Ph Ph Ph Ph Ph Ph Ph Cyclohexyl Me 4-MePh 2-OHPh 4-Pyridinyl 2-Furanyl 4-CNPh 3-CF3Ph 4-CF3Ph 4-MeOPh 3,5-DiClPh 2-Thienyl 2-MePh 2-MeOPh 3-ClPh 2-FPh 4-ClPh 3,4-DiClPh 2,5-DiClPh 2-ClPh 2,4-DiClPh 2-BrPh 2,3-DiClPh 2,4-DiClPh 2,4-DiClPh 2,4-DiClPh 2,4-DiClPh 2,4-DiClPh 2,4-DiClPh 2,4-DiClPh 2-ClPh 2.4 NA NA NA NA NA 4.4 NA NA NA NA 7.7 4.2 3.5 3.5 2.8 2.3 2.0 2.0 1.4 1.4 1.0 0.89 0.83 0.80 0.67 0.45 0.41 0.35 0.35 10 4.2 0.73 0.45 0.30 0.17 0.092 0.028 NA, <50% inhibition at 10 mM. 2627 of the two phenyl rings is seen as important. In the 5position of the triazole (Table 3), cyclohexyl (15) and methyl (16) were inactive whilst almost all aromatic groups tried had some antagonist activity. Phenyl and thiophene (2 and 26) were more potent than pyridyl and furan (19 and 20). In general it was found that analogues with a 2-substituted phenyl were the most potent and that the preferred substituent was chlorine or bromine. 2-Chlorophenyl became the 5-substituent of choice. 2-Chloro (34) was better than 2-fluoro (30), 2methoxy (28) or 2-methyl (27) and 2-hydroxy (18) was inactive possibly due to an internal hydrogen bond with the triazole. 2-Chloro was preferred to 2-bromo because of reduced lipophilicity and better pharmaceutical acceptability. Some of the potency enhancement could be lipophilicity driven but the 2-chloro substituent would be expected to cause a twist out of plane between the phenyl and triazole. With a single substituent this could be up to 30  without loss of resonance between the rings. In general, addition of 3- or 4-substituents did not cause appreciable potency changes; for example. compare 2-chloro (34) with 2,3-, 2,4-, 2,5-dichloro (37, 35, 33). Optimisation at the 2-position was undertaken in two separate series (5-phenyl analogues 2, 9–14) and 5-(2,4-dichlorophenyl) analogues 35, 38–44). Initially only benzyl (2) gave any activity, phenyl, methyl and hydrogen being inactive (9–11). Activity was removed by the addition of 2-chloro (13) and 4-trifluoromethyl (12), only 3-hydroxy (14) maintained activity. The preference for a 3-substituent was confirmed and extended in the more potent 5-(2,4-dichlorophenyl) series. 3Chloro (44) gave an increase in potency compared to the unsubstituted compound (35), other 3- and 4-substitutuents were less active (38–43). One interesting observation was that the phenethyl analogue (41) had similar potency to benzyl (35). Investigation of the role of the central triazole ring was also undertaken. A number of heterocycles substituted with phenyl and benzyl were obtained from the corporate compound collection and from external suppliers but none showed any activity (data not shown). Limited variation of the 3-position on the triazole was carried out but only 3-thiol had any activity (2 vs 3–8). Initially the acidic nature of the thiol (pKa 6.20) was thought to be important. This trend has been noted by others; SB2250028 and other urea based CXCR2 antagonists9 are all acidic in nature (Fig. 2). The role of the anionic group has been investigated further by this group,10 finding that alteration of the phenolic pKa to give uncharged molecules gave greatly reduced potency. In the case of the triazoles though more subtle SAR exists. The hydroxy compound 3 has appreciable ionised form (pKa 7.80) and the sulphonamide 8 (pKa 6.71) is even more acidic but both lack activity. With the sulphonamide, the presence of a large extra substituent might be Figure 2. Structure of acidic urea CXCR2 antagonists. 2628 A. Baxter et al. / Bioorg. Med. Chem. Lett. 13 (2003) 2625–2628 expected to interfere with binding to the receptor but at present there is no explanation of the lack of activity of the hydroxy compound 3. Table 4. Lead profile of triazolethiol 45 Generic lead criteriaa Binding IC50 <0.1 mM Ca flux IC50 <0.1 mM Rat hepatocyte Cl <14 Human microsome Cl <23 Rat iv Cl <35 Rat iv Vol>0.5 L/kg Rat iv T1/2 >0.5 h Rat po bioavail.>10% Plasma prot. bind.<99.5% Molecular weight< 450 Solubility>10 mg/mL clogP<3.0 Log D<3.0 Triazolethiol 45 0.028 mM 0.048 mM 26 14 12 8 9 61% 99.0% 336 20 4.3 3.2 a Units as Figure 1 where not stated. The profile of the lead compound (45) is shown in Table 4 compared with the lead target profile. Compound (45) has good lead-like potency, both binding and functional in the calcium flux assay, and the in vitro rat hepatocyte and human microsome data was acceptable. It is a weak acid (pKa 6.4) and has borderline lipophilicity and this is reflected in the high plasma protein binding. However the compound has maintained the satisfactory in vivo DMPK profile having a 9 h intra venous half-life and over 60% oral bioavailability. The SAR presented clearly shows scope for further improvement in potency by looking for additional interactions via further aromatic substituents and alteration of linker group between the triazole and phenyl group in the 2-position. Selectivity data was acceptable and analogues with substituents on both aromatic rings are novel. The triazolethiols exemplified by compound 45 formed the basis of a new LO project. References and Notes 1. Michne, W. F. Pharm. News 1996, 3, 19. 2. For reviews, see: (a) Rossi, D.; Zlotnik, A. Annu. Rev. Immunol. 2000, 18, 217. (b) Horuk, R.; Ng, H. P. Med. Res. Rev. 2000, 20, 155. (c) Trivedi, B. K.; Low, J. E.; Carson, K.; LaRosa, G. J. Ann. Rep. Med. Chem. 2000, 35, 191. (d) Onuffer, J. J.; Horuk, R. Trends Pharmacol. Sci. 2002, 23, 459. 3. For full details of the biological assays used, see: Baxter, A.; Bennion, C.; Cage, P.; Kindon, N.; Mortimore, M.; Roberts, B. Int. Patent Appl. WO 00/12489. 4. Durant, G. J.; Kline, S. J. Chem. Soc. C 1967, 92. 5. Tsuge, O.; Kanemasa, S.; Tashiro, M. Tetrahedron 1968, 24, 5205. 6. Davidson, J. S.; Dhami, S. S. Chem. Ind. (London) 1978, 92. 7. Behringer, H.; Ramert, R. Liebigs Ann. Chim. 1975, 1264. 8. Podolin, P. L.; Bolognese, B. J.; Foley, J. J.; Schmidt, D. B.; Buckley, P. T.; Widdowson, K. L.; Jin, Q.; White, J. R.; Lee, J. M.; Goodman, R. B.; Hagen, T. R.; Kajikawa, O.; Marshall, L. A.; Hay, D. W. P.; Sarau, H. M. J. Immunol. 2002, 169, 6435. 9. White, J. R.; Lee, J. M.; Young, P. R.; Hertzberg, R. P.; Jurewicz, A. J.; Chaikin, M. A.; Widdowson, K.; Foley, J. J.; Martin, L. D.; Griswold, D. E.; Sarau, H. M. J. Biol. Chem. 1998, 273, 10095. 10. Widdowson, K.; Nie, H.; Jurewicz, A. J.; Hertzberg, R. P.; Sarau, H. M.; Foley, J. J.; Lee, J.; White, J. R.; Veber, D. F. Lett. Peptide Sci. 1998, 5, 235.