Degeneracy Explains Diversity in Interneuronal Regulation of Pattern Separation in Heterogeneous Dentate Gyrus...

Pattern separation, the ability of a network to distinguish similar inputs by transforming them into distinct outputs, was postulated by the Marr-Albus theory to be realized by divergent feedforward excitatory connectivity. Yet, there is evidence for strong but differential regulation of pattern separation by local circuit connectivity. How do we reconcile the conflicting views on local-circuit regulation of pattern separation in circuits receiving divergent feedforward connectivity? Here, we quantitatively examined a population of heterogeneous dentate gyrus (DG) spiking networks where identically divergent feedforward connectivity was enforced. We generated 20 000 random DG networks constructed with thousands of functionally validated, heterogeneous single-neuron models of 4 different DG neuronal subtypes. We recorded network outputs to morphed sets of input patterns and applied quantitative metrics that we developed to assess pattern separation performance of each network. Surprisingly, only 47 of these 20 000 networks (0.23%) manifested effective pattern separation showing that divergent feedforward connectivity alone does not guarantee pattern separation. Instead, our analyses unveiled strong contributions from the 3 interneuron subtypes toward granule cell sparsity and pattern separation, with pronounced network-to-network variability in such contributions. We traced this variability to differences in local synaptic weights across pattern-separating networks, highlighting synaptic degeneracy as a key mechanism that explains diversity in interneuronal regulation of pattern separation. Finally, we found heterogeneous DG networks to be more resilient to synaptic jitter compared to their homogeneous counterparts. Together, our findings reconcile conflicting evidence by revealing degeneracy in DG circuits, whereby similar pattern separation efficacy can arise through diverse interactions among granule cells and interneurons.