Map optogenetically driven hippocampal activity with simultaneous fMRI and GCaMP-mediated calcium recording

Objective: This work is to develop a multi-modal fMRI platform so as to identify the neurovascular coupling events in the hippocampus. The neural activity and vascular signal could be simultaneously recorded with GCaMP-mediated calcium recording and fMRI in the rat hippocampus. This work provided a basic platform to better characterize the hippocampal vascular dynamics in the normal condition and possibly in the diseased states, e.g. focal stroke of CA1 for transient global amnesia. Results: A single-vessel fMRI method was developed to map hemodynamic signal of penetrating vessels in the deep cortical layers beyond the penetration depth of conventional optical methods1. It provided a unique advantage to map the subcortical vascular dynamics in the hippocampus. The hippocampal vascular network showed parallel alignment of interleaved arterioles and venules2, 3. In this work, a slice orientation trajectory was set to map individual arterioles and venules of the hippocampal vascular network. First, the multi-gradient-echo (MGE) images were acquired with the high flip angle at 45° and 100ms TR. At the short TE, both arterioles and venules were detected as bright dots due to in-flow effects, but at the longer TE, the venules were shown as dark dots due to the fast T2* decay1, 4. By integrating the MGE images acquired at different TE, we could identify individual penetrating vessels in the hippocampus from the arteriole-venule (A-V) map (Fig 1A). Channelrhodopsin-2 (ChR2) was expressed in the hippocampus, which was reported to initiate BOLD fMRI signal with fiber optic-mediated optical stimulation5, 6, 7. The steady state free precession (SSFP) imaging method 8, 9 was implemented to characterize the optogenetically driven single-vessel hemodynamic signal from the hippocampus, showing the BOLD fMRI signal located primarily at penetrating venules with on/off block design paradigm (Fig 1A). In addition, the optogenetically-driven hippocampal local field potential was recorded with in vivo electrophysiology (Fig 1B). GCaMP6f was co-expressed with ChR2 in the hippocampus (Fig 1C). Two channel fiber optics were inserted to target hippocampus for both optical stimulation and calcium recording. The optogenetically driven local field potential (LFP) was recorded in both rats with or without GCaMP6 expression (Persuasive data file). In the calcium recording channel, the optical stimulation light pulse was detected as sharp spikes during pulse delivery period in both rats. In contrast, the delayed calcium signal peaked at 50-70 ms following the light pulse was only detected in rats with GCaMP6 expressed in the hippocampus (Fig 1C). The calcium signal could be detected simultaneously with fMRI, showing optogenetically driven BOLD signal at the hippocampus. Conclusion: We have successfully established the multi-modal fMRI platform to simultaneously record optogenetically driven fMRI and calcium signal in hippocampus. This work makes it possible to further investigate the signaling events through the hippocampal neurovascular network to better understand vascular dynamic basis of hippocampal function