TXA has been shown to be a

TXA has been shown to be a competitive antagonist on the GABAA receptor and potentially also the glycine [7], [12], [15] receptor. Furtmuller et al. showed that TXA right-shifted the GABA CRC by one order of magnitude [7], while in our study TXA similarly right-shifted the GABA EC50 value from 0.33 μM to 2.5 μM. Furthermore, TXA at a concentration of 1 mM right-shifted the EC50 for glycine from 94 μM to 0.21 mM in rat PU-WS13 [14], compared to our results where TXA at 3 mM shifted the potencies from 11 μM to 0.14 mM. Together these results show that the DMR assay can be applied for probing of PBI selectivity towards human inhibitory neuron receptors. TXA was 10 times more potent as an antagonist at extrasynaptic glycine receptors compared to synaptic glycine receptors [14], [15]. This can be an important chemical property to identify during lead optimization. Picrotoxin at low concentrations can be used to discriminate between α-homomeric and αβ-heteromeric glycine receptors [2]. Addition of taurine at a concentration which only activates the glycine receptor showed that picrotoxin could partially block the response. This suggests that hiPSC neurons express both homomeric and heteromeric glycine receptors and that glycine composition selectivity studies can be performed with the DMR assay. GABAA and glycine receptors are widespread in the CNS and are of interest for selectivity studies in drug discovery since blocked inhibitory transmission via these receptors can induce seizures [7], [12], [15]. Previous PBI development programs have not been able to sufficiently address unwanted receptor activities [3]. Hence the usefulness of our assay was tested on three potent PBIs. AZD6465 did not show GABAA receptor binding, while TXA effectively bound to the GABAA receptor, as previously reported using a low-throughput rat membrane binding assay [3], [4]. Additionally, we confirm that the discontinued AZD6564 analog AZ13267257 is also a weak agonist of the GABAA receptor. The hiPSC-derived EC50 value of 15.4 ± 5.0 μM (n = 3) is in good agreement with screening data from the rat brain membrane binding assay at 12 μM (CRC with 10 data points, AZ, data on file). The incomplete concentration-response curve indicates that it is a partial agonist. Interestingly, 4-PIOL, the origin for the development of AZD6564, is also a low-efficacy partial GABAA agonist [13]. We conclude that the hiPSC-derived neuron assay is more informative and convenient than the rat membrane binding assay As previously reported the DMR assay is useful for pharmacological characterization of receptors or ligand gated ion channels. Our results also show the possibility to use hiPSC-derived neurons for high-throughput drug discovery efforts, including selectivity studies in a physiologically relevant cellular background with a likely high translatability to humans. In this proof of concept study, we could show that the competitive antagonistic effect towards the GABAA and glycine receptor in rat neurons, can be identified in human neurons using DMR. This further supports the mechanistic explanation of seizures following high dose administration of TXA in humans. Additionally, in the development of novel candidate drugs this method can be effectively used to identify unwanted CNS receptor effects.
Author contribution
Acknowledgements This work was supported by The Swedish Research Council (grant number: 2013-5731). K.L., L.A., and R.E are employees of AstraZeneca. G.D., F.J., and T.F. are employees of Emeriti Bio. Thanks to Dr. Thomas Lundbäck for proofreading the article.
Introduction Glycine is of highest importance in the adult spinal cord and brainstem mediating fast inhibitory neurotransmission accounting for coordinated movement and respiratory rhythm. Mutations in genes encoding for glycine receptor (GlyR) subunits α1 (GLRA1) and β (GLRB) are the major underlying causes of the human neurological motor disorder hyperekplexia (OMIM # 149400) [1], [2].