Initial SAR studies on apamin-displacing 2-aminothiazole blockers of calcium-activated small conductance potassium channels
Abstract
An initial SAR study on a series of apamin-displacing 2-aminothiazole KCa2 channel blockers is described. Potent inhibitors such as N-(4-methylpyridin-2-yl)-4-(pyridin-2-yl)thiazol-2-amine (13) are disclosed, and for select members of the series, the relationship between the observed activity in a thallium flux, a binding and a whole-cell electrophysiology assay is presented.
Small conductance calcium-activated potassium (KCa) channels are expressed predominately in neuronal cells.1 They are voltage insensitive and open in response to elevated concentrations of intracellular calcium. Calcium ions bind to calmodulin, a protein that is constitutively associated with the C-terminal domain of the KCa channel. This protein modulates the state of the channel in response to its calcium occupancy, whereby higher occupancy is associated with opening of the channel.2–6 KCa channel opening results in hyperpolarization of the plasma membrane7 with a con- comitant reduction in neuronal excitability. The hyperpolarization phenomenon (termed an afterhyperpolarization or AHP) can per- sist for several hundreds of milliseconds,8 and significantly affects the rate and pattern of neuronal firing.9,10
Three isoforms of KCa channels have been identified and cloned: KCa2.1, KCa2.2 and KCa2.3. In situ hybridization, Northern blot, and RT-PCR studies have demonstrated differential regional expression levels of the isoforms in both rat and human brain.
Correspondingly, there exists the possibility for site selective modulation of neuronal excitability if suitable pan-KCa or sub-type selective KCa channel modulators can be identified. Such agents may have utility in the treatment of a number of pathological con- ditions13 including epilepsy, depression, and Parkinson’s disease,14
A limited number of compounds have been reported to be active at KCa channels. The cyclic octapeptide apamin is a highly selective and potent blocker with reported IC50’s between 30 pM and 20 nM.1,18 The larger peptide Scyllatoxin is exceptionally potent (Ki = 75 pM)19,20 and related derivatives display moderate selectivity among the KCa channel subtypes. As shown in Figure 1 above, dequalinium (IC50 = 1.0 lM) was the first non-peptidic KCa selective blocker identified.21 Subsequent development of this chemotype led to the discovery of the cyclophane derivative, UCL 1684, a compound that displayed similar potency to apamin (IC50 = 3.0 nM).22 More recently, quinoline appended diazepines23 (Ki = 140 nM) and isoquinoline analogs related to bicuculline and N-methyl laudanosine24–28 have been reported, as well as the non-apamin displacing 2-aminobenzimidazoles, such as NS8593 (IC50 = ~500 nM).29 In this Letter, we describe a preliminary SAR study on a series of aminothiazoles that display significant KCa channel activity. In addition, we confirm the previously reported activities of some of these analogs30 in alternative assay systems, and significantly expand on reported selectivity and mechanism of inhibition studies.
The thiazole chemotype discussed here was identified from a high-throughput screen that employed a thallium flux assay,33 in which compounds were tested against a HEK 293 cell line recom- binantly expressing specific KCa channel isoforms. Several chemo- types were identified with one of the more interesting being the 2-aminothiazole derivative 1, as shown in Figure 2.
Figure 1. Examples of reported, non-peptidic blockers of KCa channels.
Figure 2. Aminothiazoles evaluated for activity at KCa channels.
This compound displayed significant potency (89% inhibition at 30 lM) in the original screen. It was resynthesized, and its IC50 in the thallium assay was determined to be 0.5 lM. In a subsequent whole-cell electrophysiology experiment, the IC50 of 1 was found to be in close agreement with the value obtained in the thallium assay, as shown in Table 1. In addition to the compound’s impres- sive activity, we anticipated that its physicochemical properties [MW = 254, CLog P = 2.71, HBD = 1, HBA = 3, ACD pKa (conjugate acid) = 3.8] held excellent promise for its advancement and utility in investigating CNS effects of KCa channel block.
Subsequently, we investigated the activities of the series of re- lated aminothiazoles shown in Figure 2. These compounds are either commercially available, or can be synthesized from readily accessible starting materials using the methodology depicted in progressive replacement, as shown in analogs 7 and 8. Both com- pounds displayed very limited activity in the thallium assay. In additional studies, the importance of the pendant amino function- ality in 1 was explored in analogs 6 and 9, in which the amino group was either replaced by a methylene group or simply elimi- nated. Again, both substitutions were associated with a significant loss of activity at the KCa channel.
The above observations raised the possibility that the activity of 1 was due to the formation of a chelate. 2-Aminothiazoles are known ligands in a number of chelate complexes,30,34 and the idea that the chelate may be the active species is partly supported by reports30 showing that thiazoles and the related iron or zinc chelates have similar activity in a K 2.2 rubidium flux assay. In relation to the assay systems reported here, the most likely ion that would participate in chelate formation is Mg2+, and several com- plexes of this type have been reported.35 Based on the SAR pre- sented above, the most probable structure of such a complex would be I, as shown in Figure 4, although II and III cannot be excluded.
Figure 3. The majority of the analogs were prepared using the mod- ified Hantzsch31 procedure shown in reaction A. The 2,4-di-(pyri- din-2-yl)thiazole analog 9 was synthesized using the same methodology employing a thioamide, as shown in B. Lastly, the 4-(pyridin-2-yl)-2-(pyridin-2-ylmethyl)thiazole derivative 6 was synthesized using the two-step procedure shown in reactions C32 and D.
Initially, we sought to investigate the importance of the relative disposition of the pyridine nitrogens in 1. Thus, retaining the con- nectivity of the 4-(pyridin-2-yl)-N-2-aminothiazole moiety, the nitrogen atom of the N-pyridinyl group was transposed as shown in structures 2 and 3 above.
As can be seen from Table 1, both analogs lost significant activ- ity at the KCa2.3 channel. A similar exercise was conducted, in which the heteroatom of the pyridinyl group at the 4-position of the thiazole ring was varied in the context of the N-(pyridin-2-yl) moiety, as shown in compounds 4 and 5. Again, a significant loss of activity was observed. The importance of the nitrogens in the two terminal pyridinyl groups was further demonstrated by their
With the key pharmacophoric elements now established, we next attempted to introduce additional functionality into the N- (pyridin-2-yl) group while retaining the pyridin-2-yl group at the 4-position of the aminothiazole. The introduction of a nitrogen atom as shown in the pyrimidin-2-yl analog 10 was again associ- ated with a loss of activity, and we tentatively attributed this to a reduction in the basicity of the heterocycle (ACD pKa’s: 3.8 vs 2.4). Additional attempts to introduce steric probes at either C3 or C3 in combination with C5 as in analogs 11 and 12, resulted in a significant loss of activity.
In a related study focused on the effects of substitution at the C4 vector of the N-pyridinyl moiety, we found that the N-(4-methyl- pyridin-2-yl) derivative 13 displayed significantly improved po- tency relative to 1. Rather than modulating the basicity of the heterocycle, we envisage that the methyl group interacts directly with the channel. The activity of this compound was also deter- mined in EP experiments, and good agreement between the assays was observed (54 nM vs 56 nM). This level of activity at the KCa2.3 channel is unprecedented in a neutral small molecule, and 13 should prove useful for KCa2 channel functional studies.
To determine the site of action of 1 and 13, we employed a Scin- tillation Proximity Assay (SPA) to assess their ability to displace compounds competed off the peptide with IC50’s as shown in Table 2. These results suggest that both may function by blocking the pore of the channel, as is observed with apamin (see Table 3). To assess the KCa channel selectivity of the thiazole chemotype, selected analogs were assessed in the thallium flux assay against cell lines recombinantly expressing the KCa2.1, KCa2.2, and KCa3.1 channels. No significant selectivity was observed, although it was noted that both 1 and 13 displayed essentially no activity at the KCa3.1 channel.
Figure 3. Synthetic methods used for the synthesis of the 2-aminothiazoles listed in Table 1.
In conclusion, we present a series of N-(pyridine-2-yl)-4-(pyri- dine-2-yl)-2-aminothiazoles that display excellent potency as KCa2 blockers. In binding studies, these compounds appear to inter- act with the channel at the apamin binding site, and presumably exert their effect by mechanically blocking the pore of the channel. We speculate that the active species may be the thiazole itself, or a metal chelate in which the thiazole functions as a ligand.