P032 Universal coarse-to-fine transition across the developing neocortex
Lorenzo Butti*1, Deyue Kong1, Nathaniel Powell2, Bettina Hein1, Jonas Elpelt1, Haleigh Mulholland2, Gordon Smith2, Matthias Kaschube1
1FIAS (Frankfurt Institute for Advanced Studies), Frankfurt am Main, DE
2Department of Neuroscience, University of Minnesota, Minneapolis, USA
*Email: butti@fias.uni-frankfurt.de
Introduction
How cortical representations emerge in development is an unresolved problem in neuroscience. Recent work in ferret shows that during early development, spontaneous activity exhibits a modular organization that is highly similar across diverse cortical areas, from sensory cortices to higher-order association areas [1]. Moreover, this modular organization persists in all areas after eye and ear-canal opening (approx. postnatal day 30), but the organization also changes, suggesting a considerable refinement over development, part of which may be area-specific [2]. It is currently unclear how this refinement unfolds on the level of local neural circuits and what mechanisms might guide this maturation.
Methods
We examine the development of network organization across diverse cortical regions (V1, A1, S1, PPC, PFC ), before (P21-24), around (P27-32) and after (P39-43) eye opening using both widefield and 2-photonin vivocalcium imaging of spontaneous activity in the ferret.
To gain mechanistic insight, we employ Local Excitation/Lateral Inhibition (LELI) network models, following [3]. These models can both reproduce the modular structure of early cortical activity and account for the ability of developing cortical circuits to transform unstructured input into modular output.
Results
We find that in both sensory and association areas, networks exhibit a highly similar pattern of changes over development: spontaneous activity is initially highly modular, i.e. strongly correlated and low-dimensional in local populations, becoming less correlated and higher-dimensional with age.
These in vivo changes can be explained by a developmental increase in lateral inhibition strength in a LELI model. This allows feedforward inputs to engage a larger number of network states, consistent with the transition of cortical networks to external sensory activity during this period. Moreover, the increase in inhibition predicts a decrease in modular wavelength over this same developmental time, which we confirm in our experimental data.
Discussion
Our findings indicate that the spontaneous activity in ferret cortex undergoes a developmental reorganization from coarser to finer-scaled organization, accompanied by a transition to more high-dimensional activity in both sensory and association areas. We propose that an increase in lateral inhibition serves as a common mechanism underlying cortical network refinement, and that this maturation leads to the expansion of representational capacity throughout the developing cortex.
Acknowledgements
We also thank the members of the Kaschube lab and the Smith lab for the useful discussions.
References
[1] N Powell, B Hein, D Kong, J Elpelt, HN Mulholland, M Kaschube, GB Smith. (2024).https://doi.org/10.1073/pnas.2313743121
[2]N Powell, B Hein, D Kong, J Elpelt, HN Mulholland, R Holland, M Kaschube, GB Smith.(2025).https://doi.org/10.1093/cercor/bhaf007
[3]HN Mulholland, M Kaschube, GB Smith.(2024).https://doi.org/10.1038/s41467-024-48341-x
