Pharmacology and crystal structure of novel 2,3-quinoxalinediones at kainate receptors

Research output: Contribution to journalConference abstract in journalResearchpeer-review

Stine Møllerud, Jakob Staun Pallesen, Diletta Pasini, Laura Marconi, Liwei Han, Jan Bornholt, Tommy Nørskov Johansen, Jette Sandholm Kastrup, Darryl S Pickering, Karla Andrea Frydenvang

Ionotropic glutamate receptors (iGluRs) are the primary mediators of fast excitatory neurotransmission in the mammalian CNS where they are involved in learning and memory formation. iGluRs are important for normal brain function and thus disturbances in the iGluR system are associated with the patho-physiology of CNS diseases such as epilepsy, schizophrenia and depression. Selective tool compounds are therefore needed to address the functional roles of
different types of iGluRs. A few selective compounds that can discriminate between AMPA and kainate (KA) receptors are available. However, within the KA receptor family (GluK1-5) only compounds with selectivity towards GluK1 exist [1]. Thus, there is an unmet need for Tool compounds with selectivity towards the remaining KA receptor subunits.
Here we report the pharmacology of a series of novel N1-substituted 2,3-quinoxalinediones, as well as the crystal structure of one compound (JP-10-7A) in the GluK1 ligand binding domain (GluK1-LBD) at 1.85 Å resolution. Radioligand binding experiments indicated that most of the compounds had similar binding affinities at GluK1 and GluK3, but a few had higher affinity at GluK3 and were thus GluK3-preferring.
The GluK1 binding mode of the JP-10-7A 2,3-quinoxalinedione scaffold is similar to that of another published 2,3-quinoxalinedione ligand, (S)-2-amino-4-(2,3-dioxo-1,2,3,4-tetrahydroquinoxalin-6-yl)-butanoic acid (PDB-entry 4QF9), with the substituent in the N1-position pointing out of the binding pocket. Whereas agonists induce a closure of domain D2 towards D1,
antagonists stabilize an open conformation of the GluK1-LBD. Domain opening of GluK1-LBD with JP-10-7A bound (compared to glutamate bound GluK1-LBD PDB-entry 2F36, molA) is approximately 30°, which is consistent with an antagonist binding mode. Functional electrophysiological (TEVC) experiments indeed showed these compounds to be antagonists at cloned, homomeric KA receptors. The structure and pharmacology will be valuable for design of new and more GluK3-selective quinoxalinedione analogues.
Original languageEnglish
Article numberMTU07-13
JournalJournal of Neurochemistry
Volume142
Issue numberS1
Pages (from-to)124
Number of pages1
ISSN0022-3042
DOIs
Publication statusPublished - Aug 2017

ID: 199122670