Laminar CBV-fMRI reflects dynamic regulation by synaptic input-driven activity in mouse somatosensory cortex

초록

Mapping brain activity across cortical layers is crucial for understanding laminar-level functional details, including thalamocortical (TC) or coticocortical (CC) pathways. BOLD (blood oxygenation level-dependent) contrast widely used in fMRI is not suitable for resolving layer-specific activity due to its bias towards venous site. In contrast, CBV changes, driven by arterial vessels and capillaries, provide higher specificity to the neuronal active site. However, its neurophysiological source remains unclear. Here, we investigated whether laminar fMRI responses reflect neuronal input- or output-driven activity by comparing somatosensory-evoked fMRI responses (BOLD vs. CBV) with electrophysiological data in mice. Furthermore, we performed ultrahigh resolution CBV-weighted laminar fMRI by modulating TC and CC inputs with stimulus onset asynchrony to determine whether layer response is regulated by initial synaptic input and/or synaptic strength. In cortical depth-dependent fMRI responses, somatosensory-evoked CBV peaked at L4 of S1, while BOLD response peaked at the cortical surface. CBV specificity to L4 increased over time, whereas early BOLD appeared in L4. In sensory-evoked neural activity, the highest current sink and LFP changes were observed at L4 depths, consistent with TC inputs, while early MUA responses occurred at L4 but were stronger in L5 later. Despite TC input preceding CC input, the highest CBV responses occurred in the middle layer of S1. Conversely, with reversed stimulus onset order, the CC inputs remained predominant. However, when the second input strength was larger, the second inputs also contribute to CBV response. Our data demonstrate that laminar CBV responses primarily reflect synaptic input-driven activity rather than output response, as indicated by LFP and MUA. Moreover, these responses are tightly regulated by microvessels coupled with earlier synaptic input, but potentially driven by more synchronous activity within neural circuits.