Back

Functional Continuum of GABAergic Synaptic Dynamics Reflects Genetic Identities

Poirier, J.; Beninger, J.; Toth, K.; Naud, R.

2026-06-10 neuroscience
10.64898/2026.06.09.731181 bioRxiv
Show abstract

The language of the brain is articulated by temporal patterns of neuronal activity, which individual synapses interpret through distinct forms of synaptic dynamics. While glutamatergic synapses have been proposed to use a handful of functionally distinct types of short-term plasticity (STP) that loosely align with genetic identities, the functional organization of GABAergic synaptic dynamics remains ambiguous. Here, we ask whether the diversity of GABAergic dynamics also clusters into discrete functional types or instead forms a continuum, and whether this functional organization also aligns with genetic identities. Using a combination of machine learning and synaptic modeling on the Allen Institute Synaptic Physiology Dataset, we present evidence that inhibitory dynamics organize into a functional continuum with overlapping modes corresponding to different genetic identities. This continuum spans Parvalbumin (Pvalb) to Vasoactive Intestinal Peptide (VIP) presynaptic neurons, with Somatostatin (Sst) neurons forming an intermediate class. Across the three classes, most inhibitory synapses showed either strong depression or a biphasic form of plasticity that is facilitating at higher stimulation frequencies and depressing at lower ones. However, a higher proportion of Sst and VIP synapses, compared to Pvalb synapses, exhibited either depression restricted to high frequencies or consistent facilitation. When paired with functional subtypes of excitatory synaptic dynamics, this inhibitory continuum completes a unified framework for understanding information flow in the cortical microcircuit, advancing long-standing efforts to explain circuit function by investigating the organization of its components.

Matching journals

The top 4 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.