Laminar profiles of hippocampal subfields are differentially associated with navigation strategies

963 

Abstract Submission 

Khazar Ahmadi¹, David Stawarczyk¹, Viktor Pfaffenrot², Carlos Gomes¹, Sriranga Kashyap³, Zita Patai¹, David Norris⁴, Nikolai Axmacher⁵

¹Department of Neuropsychology, Institute of Cognitive Neuroscience, Ruhr University Bochum, Bochum, North Rhine Westfalen, Germany, ²Erwin L. Hahn Institute for Magnetic Resonance Imaging, University of Duisburg-Essen, Essen, North Rhine Westfalen, Germany, ³Krembil Brain Institute, University Health Network, Toronto, Ontario, Canada, ⁴Donders Institute for Brain Cognition and Behaviour, Radboud University, Nijmegen, Gelderland, Netherlands, ⁵Department of Neuropsychology, Institute of Cognitive Neuroscience, Ruhr University Bochum, Bochum, North Rhine Westfalen, , Germany

Khazar Ahmadi  
Department of Neuropsychology, Institute of Cognitive Neuroscience, Ruhr University Bochum
Bochum, North Rhine Westfalen, Germany

David Stawarczyk  
Department of Neuropsychology, Institute of Cognitive Neuroscience, Ruhr University Bochum
Bochum, North Rhine Westfalen, Germany

Viktor Pfaffenrot  
Erwin L. Hahn Institute for Magnetic Resonance Imaging, University of Duisburg-Essen
Essen, North Rhine Westfalen, Germany

Carlos Gomes  
Department of Neuropsychology, Institute of Cognitive Neuroscience, Ruhr University Bochum
Bochum, North Rhine Westfalen, Germany

Sriranga Kashyap  
Krembil Brain Institute, University Health Network
Toronto, Ontario, Canada

Zita Patai  
Department of Neuropsychology, Institute of Cognitive Neuroscience, Ruhr University Bochum
Bochum, North Rhine Westfalen, Germany

David Norris  
Donders Institute for Brain Cognition and Behaviour, Radboud University
Nijmegen, Gelderland, Netherlands

Nikolai Axmacher  
Department of Neuropsychology, Institute of Cognitive Neuroscience, Ruhr University Bochum
Bochum, North Rhine Westfalen, , Germany

The Hippocampus (HP) is a critical region for core cognitive functions including memory and spatial navigation. Although these functions have been extensively studied at macroscale (1), the underlying circuit-level mechanisms are yet to be determined. Harnessing the current advancements in functional magnetic resonance imaging (fMRI) at submillimeter scale, we aimed to investigate the laminar organization of HP subfields during spatial navigation and their associations with distinct navigation strategies.

39 healthy volunteers underwent a 2-session fMRI study using a 7T MAGNETOM-Terra scanner. During the first day, an anatomical image was acquired with 3D-MP2RAGE (2) sequence (isotropic spatial resolution of 0.75 mm3). This was followed by the acquisition of two fMRI runs using a 3D GRE-EPI sequence (3) with BOLD contrast (isotropic voxel size of 0.8 mm3, phase-encoding direction = A-P, TR = 2500 ms). On day 2, six additional fMRI runs were obtained. During each run, the participants navigated to six hidden objects randomly distributed in a virtual arena and completed 18 trials per run, each consisting of a retrieval and a subsequent re-encoding phase (see Fig.1A). Following preprocessing of the fMRI data and subsequent alignment with the anatomical images, HP was segmented into subfields including cornu ammonis (CA 1-4), stratum radiatum/lacunosum-moleculare (SRLM) and dentate gyrus (see ref.4). Furthermore, three folded surfaces were generated using an equivolume model, representing inner/outer sections and mid-thickness of the HP gray matter (Fig.1B). These surfaces were then transformed into the space of the corresponding anatomical image and equidistantly sampled into 30 depth bins using an in-house MATLAB script. This allowed for extraction of the BOLD signal across the bins during the navigation phase, that were later grouped into the inner, middle and outer depths. We further assessed trial-specific performance, i.e., drop error (see Fig.1A). Afterwards, different navigation strategies were quantified with two metrics: straightness index (SI), and median deviation to the boundary (MDB; see Fig.2A). Subsequently, we performed linear mixed-effect models to assess the association of subregional laminar profiles with drop error and navigation strategies. The extraction of laminar profiles in subfields is still ongoing (N = 13 until now).

Drop error decreased across runs and trials, indicating a gradual improvement of the participants' behavioral performance (see Fig.2B). During the navigation phase, an inverse U-shaped pattern of the BOLD signal was observed throughout the layers of CA1 and the adjacent segment of SRLM whereas the laminar profile of CA3 demonstrated a largely monotonous decrease from inner to outer depths (Fig.2C). Further, higher drop error was associated with increased BOLD activity in inner, middle and outer depths of CA1, CA3 (β = 3.02, 4.46, p = 0.0002, 0 | β = 2.53, 4.96 p = 0.003, 0 | β = 1.77, 5.27, p = 0.04, < 0.0001, respectively) and the neighboring SRLM regions (β = 2.34, p = 0.002 SRLM-CA1; β = 3.75, p = 0.0001 SRLM-CA3). While SI was positively correlated with laminar activity only in the middle and outer layers of CA1 (β = 2.43, 3.59 p = 0.005, 0.0001, respectively), lower SI was associated with elevated BOLD signal across all three layers of CA3 (inner: β = -3.74, p = 0.0003, middle: β = -4.26, p = 0.0001, outer: β = -4.45, p = < 0.0001) and SRLM-CA3 (β = -2.99, p = 0.001). By contrast, an increase in MDB was accompanied by a higher BOLD signal specifically in CA3 (inner: β = 2.59, p = 0.01, middle: β = 2.67, p = 0.01, outer: β = 2.76, p = 0.007) and SRLM-CA3 (β = 2.66, p = 0.004).

Our preliminary results demonstrate the feasibility of laminar-resolution recordings in human HP. They suggest that distinct navigation strategies are differentially related to subregion-specific laminar profiles, and that higher BOLD responses in CA1 and CA3 are negatively related to navigation performance.

Higher Cognitive Functions Other ¹

Learning and Memory Other

Activation (eg. BOLD task-fMRI) ²

FUNCTIONAL MRI

Memory

NORMAL HUMAN

Statistical Methods

Other - navigation; hippocampus; ultra-high field