Primacy of vision shapes behavioral strategies and neural substrates of spatial navigation in the hippocampus of the common marmoset

Abstract

The mammalian hippocampus has been compared to a Global Positioning System (GPS) that enables spatial navigation. This notion has been primarily drawn from studies conducted in nocturnal mammals, such as rats; that lack many adaptations to daylight vision compared to diurnal primates. Here we demonstrate that during foraging in a 3D maze, the common marmoset, a new world diurnal primate with foveal, stereo-color vision, predominantly uses rapid head-gaze shifts to visually explore their surroundings while remaining stationary, and then minimizes head movements to navigate towards goals. On the other hand, rats, mainly move their head at low velocities while locomoting to explore the environment using their whiskers. These differences in exploration-navigation strategies reflect the species' sensory adaptations to different ecological niches. In the marmoset hippocampus CA3/CA1 regions putative pyramidal neurons show selectivity for 3D view, head direction, and less for place, but mainly mixed selectivity for combinations of these variables. Despite weak place selectivity, the spatial position of the animal in the maze can be decoded from the activity of small ensembles of mixed selective neurons. Inhibitory interneurons are tuned to 3D angular head velocity and translation speed, with most cells showing mixed selectivity for both variables. Finally, marmosets lack the rhythmic theta oscillations of local field potentials seen during locomotion in rats. Instead, they show resetting of theta oscillations triggered by head-gaze shifts that co-occurred with the activation of inhibitory interneurons, followed by various modulations in the activity of pyramidal cells. Our results show that the marmoset visual exploration/navigation strategies and the hippocampal neuronal specializations supporting them diverge from those observed in rats, reflecting the far-sensing capabilities of the marmoset visual system adapted to diurnal lifestyle.

Circuit dynamics of superficial and deep CA1 pyramidal cells and inhibitory cells in freely-moving macaques

“…In this study, we sought to forge a link between the extensive understanding of CA1 microcircuit func�on from rats and mice 64,65 to the network dynamics and brain-behavior states seen during naturalis�c experiences in primates 36,37,42,66,67 . Building on previous work 58,59,68 , the semi-supervised clustering of waveforms and ISI distribu�ons of isolated CA1 single units iden�fied func�onal cell groups with diverse physiological characteris�cs.…”

“…Extensive research on the organiza�on of rodent CA1 has iden�fied state-dependent roles for specific func�onal cell types in regula�ng synap�c dynamics, network oscilla�ons, and cell assembly dynamics 24,25,123,124 . Despite some degree of mismatch in hippocampal brain-behavioral states across rodent and primate clades 36,42,44,66,67 , it is nonetheless temp�ng to iden�fy possible correspondences between the present groups and iden�fied cell classes in rodent CA1. During sharp-wave ripple (SWR) events, most cell groups exhibited increased firing rates, however, inhibitory cell groups showed greater par�cipa�on probabili�es but lower ripple modula�on.…”

“…In addi�on to refinements to waveform classifica�on 59,90,99,100 , cells' firing paterns may prove vital for more accurate and reliable classifica�on [101][102][103] . The firing paterns and �ming of neuron ac�vity summarized by the inter-spike interval (ISI) distribu�ons have previously been applied to differen�ate inhibitory from excitatory cells across brain regions and species 52,60,66,[104][105][106][107][108][109] . Informed by prior studies 58,68 , our method addi�onally merges unsupervised analysis of spike waveforms and firing paterns (ISI distribu�ons) with supervised refinements, to maximize the use of available informa�on in extracellular electrophysiology studies for cell group classifica�on.…”