Rationale:
A host of early-life epilepsies (ELEs) are thought to arise as a consequence of altered excitation/inhibition balance resulting in circuit hyperexcitability and vulnerability to seizures. Reduced inhibitory tone from GABAergic interneurons represents a common mechanism across many ELEs and other neurodevelopmental disorders as well. In hippocampus and cortex, the most common inhibitory neurons are called fast-spiking, parvalbumin-expressing (PV) interneurons. In principle, selective manipulation of PV interneuron activity in early life could restore network balance in ELE, thereby preventing harmful circuit hyperactivity and sudden unexpected death (SUDEP). However, outside of Cre-expressing models, it is unclear whether PV interneurons can be specifically targeted at a neonatal timepoint for subsequent functional or genomic manipulation.
Methods:
Whether enhancement of PV interneuron activity during early life is sufficient to protect against seizures and SUDEP is unclear. To address this question, we developed a systemic, PV interneuron-specific targeting method using enhancer-AAVs, abrogating the need for Cre-based genetic crosses. This was achieved by performing retro-orbital injections in young wild-type or Scn1a(+/-) (Dravet) mice to virally express fluorescent or chemogenetic activator (hM3Dq) constructs in PV cells throughout cortex and hippocampus. Subsequent application of the chemogenetic ligand CNO was then performed two to three weeks later in acute slices or in vivo. Changes in PV excitability or seizure susceptibility were examined using patch clamp electrophysiology or febrile seizure induction, respectively.
Results:
PV-specific targeting was first ascertained using fluorescent (GFP)-guided patch clamp in wild-type or
Scn1a(+/-) mice at P21. We found that 100% of GFP-expressing neurons displayed a fast-spiking, non-accommodating phenotype characteristic of PV cells. GFP+ PV interneurons in
Scn1a(+/-) mice showed AP firing deficits consistent with previous work. We also immunostained for parvalbumin and found consistent somatic overlap with GFP. Next, we evaluated the effect of CNO 2-3 weeks after retro-orbital AAV injections expressing a PV-specific chemogenetic activator. CNO application increased AP firing in PV interneurons due to an apparent increase in input resistance. Lastly, CNO injections
in vivo were found to reduce the susceptibility of
Scn1a(+/-) mice to febrile seizures.
Conclusions:
Our findings show that PV-INTs can be systematically targeted in the young mouse brain for selective and long-lasting functional manipulation throughout development. Chemogenetic enhancement of PV interneurons using this method shows promise in reducing seizure activity in Dravet model mice. Overall, this approach may increase our understanding of the developmental contribution of PV dysfunction across different disease models, including Dravet. This approach may also have translational relevance, as the cell-type-specificity of several enhancer-viral approaches has been shown to scale well from rodent models to human brain tissue.
Funding: MJMR: R56AG072473, RF1AG079269, Emory ADRC grant 00100569