Wide-ranging behavioral dysfunction in two mouse models of pathological human variants in the GRIK2 kainate receptor gene

Background

Glutamate is the brain’s main excitatory neurotransmitter, and its receptors come in two major classes: fast-acting ionotropic receptors and slower, modulatory metabotropic receptors. Among the ionotropic receptors, kainate receptors (KARs) are often overshadowed by their more “famous” AMPA and NMDA relatives, yet they also play important roles in shaping synaptic transmission, regulating release probability, and influencing circuit development.

Two pathogenic GRIK2 (Glutamate Ionotropic Receptor Kainate type) variants as, p.Ala657Thr (A657T) and p.Thr660Lys (T660K) are an example of that. Clinical reports show that A657T causes intellectual disability and unstable gait, while T660K is associated with profound impairment, early-onset epilepsy, and in some cases structural brain changes.

Structurally, KARs are formed from GluK1–5 subunits, withGluK2 being a key “principal” subunit. The channel’s M3 helix is critical for gating, and small changes here can dramatically alter how the receptor opens, closes, and responds to glutamate. Residues at the tip of the M3 helix are affected in both, A657T and T660K variants.

In this preprint, Webb and colleagues from Geoffrey Swanson lab at Northwestern University, Chicago, USA have generated the CRISPR knock-in mouse lines carrying each variant to understand how these subtle changes in channel structure produce such divergent phenotypes.

Key Findings

Both mutations i.e., A657T and T660K, turned out to be gain-of-function, making GluK2-containing receptors more sensitive to glutamate, slower to desensitize, and inappropriately active even at normal glutamate levels. Yet their consequences were far from identical. In A657T mice, receptor levels and hippocampal localization were normal. In T660K mice, however, hippocampal GluK2/3 was drastically reduced, suggesting defects in receptor trafficking or stabilization at synapses.

The differences were apparent early in life. Both variants reduced viability, but T660K was especially severe: pups lagged in growth by postnatal day 10, while A657T mice grew normally until later juvenile stages, when weight began to drop. Maternal deaths during parturition meant the lines had to be maintained via male breeders.

Seizure testing revealed just how excitable these receptors had become. At a high dose of kainic acid (25 mg/kg), wild-type mice took nearly 40 minutes to reach severe seizures, whereas both mutants reached that stage within about five minutes. At a dose harmless to wild-type animals, A657T mice seized and died, while half of the T660K mice survived but continued to convulse. T660K went further, with spontaneous seizures triggered by handling, and EEG recordings showing epileptiform activity even at rest.

Behaviourally, both variants showed altered gait patterns, reduced engagement in naturalistic behaviours like digging and nest building, and signs of proprioceptive dysfunction. A657T mice largely maintained coordination, while T660K animals had poorer balance, more paw slips, and weaker grip strength. Cognitive testing revealed working memory deficits in both lines, though A657T’s impairment was specific to longer delays, while T660K was broadly impaired and slower to train. In neonatal tests, A657T pups had persistent abnormalities in ultrasonic vocalizations, while T660K pups displayed delayed walking and reflex development along with transient vocal changes.

Why I highlight this preprint

This work provides a beautiful yet rare and detailed link from ion channel biophysics to complex neurodevelopmental phenotypes. It is striking that two mutations only three amino acids apart produce such different developmental trajectories: A657T allows early survival and normal synaptic localization before later dysfunction emerges, while T660K disrupts receptor trafficking from the start, leading to early growth deficits, seizures, and broad behavioural impairment. What stands out most to me is how the authors combined precise genome editing, detailed electrophysiology, and an extensive behavioural battery to tell a complete mechanistic story. This study not only models two rare neurological disorders but also offers a template for dissecting the effects of other ion channel mutations, showing how a single amino acid change can ripple through the nervous system from molecular function to whole-animal behaviour.

Questions for the authors

1. Given that heterozygous incross progeny were reported as non-viable, were embryonic stages examined to identify the exact point of lethality? Furthermore, since heterozygous outcrosses yielded sub-Mendelian ratios, were early embryonic progeny genotyped to confirm the presence or absence of heterozygous individuals?
2. Maternal death was noted during parturition. Were any analyses performed to determine whether this was due to obstructed labor, seizure activity, or systemic collapse?
3. Did any pharmacological blockade of kainate receptors was used during early development to assess whether growth deficits or postnatal lethality phenotypes can be rescued?
4. As the A657T variant appeared more deadly to KA-induced seizures than T660K, have hippocampal deep electrophysiological recordings been performed to elucidate the underlying mechanisms?
5. In relation to motor dysfunction, have cerebellar Purkinje cells, deep cerebellar nuclei, or basal ganglia synapses been examined? Additionally, in early communication behavior assays, were any sex-specific differences observed?
6. Given that hindlimb suspension and righting reflex performance may be influenced by both motor ability and motivational state, have neurotransmitter analyses been conducted to differentiate these contributions?

References:

1. Hollmann, Michael, and Stephen Heinemann. “Cloned glutamate receptors.” Annual review of neuroscience(1994).
2. Contractor, Anis, Christophe Mulle, and Geoffrey T. Swanson. “Kainate receptors coming of age: milestones of two decades of research.” Trends in neurosciences3 (2011): 154-163.
3. Hansen, Kasper B., et al. “Structure, function, and pharmacology of glutamate receptor ion channels.” Pharmacological reviews4 (2021): 1469-1658.
4. Mayer, Mark L. “Structural biology of glutamate receptor ion channel complexes.” Current opinion in structural biology41 (2016): 119-127.
5. Guzmán, Yomayra F., et al. “A gain-of-function mutation in the GRIK2 gene causes neurodevelopmental deficits.” Neurology: Genetics1 (2017): e129.
6. Stolz, Jacob R., et al. “Clustered mutations in the GRIK2 kainate receptor subunit gene underlie diverse neurodevelopmental disorders.” The American Journal of Human Genetics9 (2021): 1692-1709.

 

doi: https://doi.org/10.1242/prelights.41228

Read preprint (No Ratings Yet)