Design and Evaluationof Selective Positive AllostericModulators Targeting Kainate Receptors

Abstract
Kainate-type ionotropic glutamate receptors (KARs) are emerging therapeutic targets for the modulation of excitatory neurotransmission in epilepsy, psychiatric disorders, and neurodegenerative diseases, yet selective positive allosteric modulators (PAMs) remain scarce. Starting from the benzothiadiazine AMPA receptor (AMPAR) PAM BPAM344 (1), we pursued a structure-based scaffold repurposing strategy to identify KAR-directed analogues. A focused library of benzothiadiazine 1,1-dioxides (7a–7z) was designed through systematic diversification at the N4 position and modulation of the C7 substituent (H, F, and Cl). Compounds were synthesized via a convergent route involving benzothiadiazine core assembly, N4 alkylation with aliphatic (including nitrile-terminated), trifluorinated, benzyl, or phenethyl substituents, followed by reduction to the corresponding 3,4-dihydro derivatives. All compounds were fully characterized and evaluated for modulatory activity at GluA2 and GluK1–3 receptors using Ca2+ fluorescence assays in HEK cells. Clear structure–activity relationships (SAR) were established. Short nitrile-terminated aliphatic substituents (7a–7g) retained GluA2 potentiation but lacked KAR activity, indicating that flexible polar chains favor AMPAR selectivity. In contrast, the N4-benzyl series revealed a more complex structure–activity relationship. The para-nitro benzyl analogue 7y was inactive across the receptor subtypes tested, whereas the 3-fluoro-4-nitrobenzyl derivative 7q showed preferential potentiation of GluK1 and GluK3, with minimal GluA2 modulation. Notably, the meta-fluorinated benzyl analogue 7u preferentially potentiated GluK3 while remaining inactive at GluK1, GluK2, and GluA2, whereas the other fluorinated derivatives were devoid of detectable activity across the receptor subtypes tested. These findings reveal a distinct subtype-dependent SAR and support the existence of a spatially constrained polar microenvironment within the GluK1/3 allosteric sites, which appears to be optimally engaged by rigid para-polar benzyl substituents. In line with our docking observations, acidic substituents may further exploit this polar region while the core scaffold remains favorably accommodated in the surrounding hydrophobic pocket. On the basis of the pharmacological testing, meta-fluorination appears to further refine subtype selectivity toward GluK3 and may represent a promising starting point for further exploratory work. Collectively, these results establish 1,2,4-benzothiadiazine 1,1-dioxides as a tractable chemotype for the development of subtype-selective KAR PAMs and provide a medicinal chemistry framework for second-generation optimization, while the docking-derived contacts should be regarded as plausible structural hypotheses rather than experimentally confirmed interactions, given that they are based on SwissDock-predicted poses rather than a crystallographic ligand–receptor complex.