In neonatal mouse CA1 (27 pups, P5-P12), pyramidal responses to self-generated movement fell linearly and changed sign near P9.7, starting before the perisomatic Syt2 surge

Dard et al. (eLife 2022) reported that mouse CA1 pyramidal cells switch from activation to suppression by self-generated movement at about postnatal day 9 (P9), within 2 days, and that a rapid surge of perisomatic GABAergic innervation causes the switch. We re-analysed their public two-photon data (DANDI 000219; 49 sessions, 27 pups, 4,597 movement onsets, about 29,000 cells) with surrogate nulls and pups as the unit. We compared the functional time course with the CA1 Syt2 perisomatic marker and with electron-microscopy synapse counts of inhibitory axons in developing barrel cortex (Gour et al. 2021). The sign change reproduced: after P9, 7 of 7 pups had net pyramidal suppression; before P9, 15 of 16 had activation. GABAergic cells stayed activated at all ages. But the change was not a step. The pyramidal response fell at an almost constant rate from P5 to P12 and crossed zero near P9.7. A linear model predicted left-out pups with 2.2 to 2.7 times less error than the best step model, and a step added nothing to the linear trend (p = 0.78 to 0.95). Twitches alone and a within-litter analysis gave the same decline. By P7 the response had completed 41% (95% CI 23-59%) of its change, while the CA1 Syt2 cover was still at its P5 level. Three cell-level tests found no negative relation between GABAergic and pyramidal activity: per-movement covariation and rest coupling stayed positive and only fell with age, and nearness to activated GABAergic cells did not increase suppression. In barrel-cortex EM, soma targeting by inhibitory axons rose from 34% of its P28 level at P7 to 92% at P9. Thus the reversal is the zero crossing of a gradual decline that starts before the measured CA1 perisomatic surge. A gradual maturation of inhibition, or of the balance of drive to the two cell classes, agrees with the data; an abrupt switch caused by an abrupt perisomatic surge does not. Main limitations: only 7 pups older than P9, whole-day ages, mostly classifier-inferred GABAergic labels, Syt2 values available only as medians, and EM from a different region.

hippocampus development · CA1 · neonatal mouse · self-generated movement · twitches · perisomatic inhibition · GABAergic interneurons · two-photon calcium imaging · DANDI 000219 · developmental connectomics · structure-function · change-point model · reanalysis