Of mice and men: topologically complex dendrites assemble uniquely human networks

Abstract

The organizational principles that distinguish the human brain from those of other species have been a long-standing enigma in neuroscience. Here, we leverage advances in algebraic topology to uncover the structural properties of the human brain at subcellular resolution. First, we reveal a much higher perisomatic branching density in pyramidal neurons when comparing homologous cortical regions in humans and mice. Traditional scaling methods consistently fail to interpret this difference, suggesting a distinctive feature of human pyramidal neurons. We next show that topological complexity leads to highly interconnected pyramidal-to-pyramidal and higher-order networks, which is unexpected in view of reduced neuronal density in humans compared to mouse neocortex. We thus present robust evidence that reduced neuronal density but increased topological complexity in human neurons ultimately leads to highly interconnected cortical networks. The dendritic complexity, which is a defining attribute of human brain networks, may serve as the foundation of enhanced computational capacity and cognitive flexibility.Abstract FigureGraphical abstractA. Human neural networks are different from mice due to their lower neuron density, resulting in increased distances between neurons, particularly among pyramidal cells. B. The topological analysis of layer 2/3 pyramidal cells in the cortex reveals an intriguing difference: human neurons exhibit a significantly larger number of dendritic branches, especially near the cell body compared to mice. This phenomenon is termed ”higher topological complexity” in dendrites. C. The combination of reduced neuron density and enhanced dendritic complexity results in greater network complexity within the human brain. Network complexity is defined by larger groups of neurons forming complex interconnections throughout the network. Our findings suggest that dendritic complexity wields a more substantial influence on network complexity than neuron density does, hinting at a potential strategy for enhancing cognitive abilities.

Human hippocampal CA3 uses specific functional connectivity rules for efficient associative memory

“…Previous studies observed high connectivity between neocortical layer 2-3 PNs in humans (10–15%; (Campagnola et al, 2022; Kanari et al, 2024; Peng et al, 2024, 2019; Planert et al, 2023; Seeman et al, 2018)). To validate that sparse synaptic connectivity in human CA3 resulted from circuit architecture, rather than tissue or recording quality, we performed additional multicellular recordings from human layer 2-3 neocortical neurons under identical experimental conditions, using temporal lobe tissue resected during hippocampal surgery ( Fig.…”

“…Therefore, not only does human CA3 employ much sparser connectivity than neocortical recurrent networks, but circuit architecture appears to sparsify across species with increasing brain size. Neocortical connectivity levels are comparable if not denser in humans than mice (Campagnola et al, 2022; Seeman et al, 2018), and formation of dense interconnectivity has been considered a mechanism for ‘uniquely human networks’ (Kanari et al, 2024). Our results suggest that the CA3 recurrent circuit uses a different mechanism to support human cognition.…”

“…Employing sparse but broad CA3–CA3 connectivity will improve pattern completion and allow the system to associate engrams in a distributed, hippocampus-wide manner. This is fundamentally different from the neocortex, in which dense connectivity within cortical modules or columns can enable fast and efficient local computations (Kanari et al, 2024). Together, we reveal distinct features of microcircuit function in the human hippocampus.…”

Heterogeneous and non-random cortical connectivity undergirds efficient, robust and reliable neural codes

“…Experimental research has demonstrated several ways in which neuronal connectivity is nonrandom with significant higher-order structure 30,33,42,43,45 . Moreover, small models with simple dynamics have shown the relevance of higher-order motifs in the observed activity [69][70][71] while in silico research has shown that the complexity is largely driven by neuronal morphologies 72,73,32 . At the same time, in the field of computational neuroscience, the bulk of research is conducted in models with simple connectivity based mostly on pairwise statistics that do not capture the entire complexity of the structure.…”