Introduction

The single-vessel fMRI method has been developed to map the hemodynamic signal from individual penetrating vessels in deep cortical layers¹. The advantage of the single-vessel fMRI method is to detect the vessel-specific hemodynamic responses at the deeper brain regions beyond the penetrating depth of the conventional optical methods (e.g. two photon imaging), e.g. hippocampus, noninvasively². By combining single-vessel fMRI with optogenetics and fiber optic calcium recording methods, i.e. an multi-modal fMRI platform, we could examine the underlying neurovascular coupling events in the subcortical regions of animals at both normal physiological and pathophysiological states³. Several studies reported the optogenetically driven seizure-like events in the hippocampus of anesthetized rats with fMRI^{4, 5}. However, it remains unclear how the seizure-like or long-term depolarization events couple to hemodynamic signaling in the hippocampal vasculature. Here, we applied the multi-model fMRI platform to acquire concurrent single-vessel fMRI and calcium signal upon optogenetic stimulation, showing distinct neurovascular coupling efficacy between normally evoked and long-term depolarization/seizure-like calcium events in hippocampus.

Methods

fMRI were performed in alpha-chloralose anesthetized rats. The detailed surgical procedures were described previously⁶. Briefly, all images were acquired with a 14.1T/26 cm horizontal bore magnet (Magnex), interfaced to an AVANCE III console (Bruker) and equipped with a 12 cm gradient set, capable of providing 100 G/cm with a rise time of 150 μs (Resonance Research). A transreceiver surface coil with 7.5 mm diameter was used to acquire fMRI images. To detect the individual arterioles and venules, a 2D-MGE sequence was used with the following parameters: TR, 100 ms; TE, 2.5, 5.5, 11.5, 14.5, 17.5, 17.5, 20.5, 23.5 ms; flip angle, 58°; matrix, 256x192; in-plane, 67x67 μm; slice thickness, 500 μm. For fMRI, the steady state free precession (SSFP) imaging method was applied to map the fMRI signal with the following parameters: TE, 5.85 ms (BOLD) TR, 11.7 ms; matrix, 128x128 (BOLD); flip angle, 22°; TE, 5.2 ms (CBV); TR, 10.4 ms; matrix, 96x96 (CBV); flip angle, 17°; slice thickness, 500 μm. CBV-weighted signals were obtained after intravenous administration of 15-20 mg Fe/kg dextran coated iron oxide (Biopal, MA). The ChR2 and Gcamp6f were expressed by AAV5 viral vectors into the hippocampal CA1 region with CaMKII and Syn promoters. Fiber optic (200 μm) was inserted into the hippocampus for optical stimulation, the light pulse was delivered through the 473 nm laser (10 ms, 3 Hz, 8 s). AFNI software was used for image analysis. The amplitude modulated response model (AM1) based on the AFNI script was used to perform the linear regression analysis.

Results

We first established the multi-modal fMRI platform to record the single-vessel fMRI and calcium signal in the hippocampus upon optogenetic stimulation. Fig 1A shows the schematic plan for the opto-fMRI with calcium signal recording in the hippocampal region, where GCaMP6f was co-expressed with Channelrhodopsin-2. Individual penetrating venules (black dots) and arterioles (white dots) in the hippocampus can be specifically in a 2D A-V map (Fig 1B)². Two fiber optics were inserted into the hippocampus to concurrently deliver light pulses for optogenetic activation and record the evoked GCaMP-mediated calcium spikes. Fig 1C shows that the BOLD signal was primarily located at the venules and the CBV signal was mainly located at the arterioles. The corresponding calcium signal driven by 10 ms light pulse was recorded simultaneously. As previously reported^2,4,5, seizure-like events could be initiated by optogenetic activation in the hippocampus. Here, we observed long-term depolarization (LT-D)-induced large-scale calcium signal during the inter-stimulus intervals and the corresponding increased BOLD signal(Fig 2A). An AFNI-based amplitude modulated response model was applied to perform the linear regression analysis between the calcium and BOLD signal for both normally evoked and LT-D events. The calcium-based neurovascular coupling coefficient is significantly higher for normally evoked events than the LT-D event (Fig 2B, C). Fig 2D shows the vessel-specific coefficient scatter plot for both events, demonstrating largely spread coefficient values for normally evoked events and uniformly distributed LT-D coefficients. By plotting the coefficients from individual vessels as the function of venule distance, the LT-D coefficients were similar in venules penetrating the whole hippocampal structure, but the peak coefficients for normally evoked events were mainly located at the activated hippocampal regions.

Conclusion

We detected distinct neurovascular coupling efficacy between normally evoked and Long-term depolarization-induced calcium events. The significantly reduced coefficients of the LT-D event indicates the lack of sufficient blood supply to the hippocampal regions. This multi-modal hippocampal fMRI mapping scheme makes it possible to better decipher the vascular dynamic basis of hippocampal function during the seizure-like states.

Acknowledgements

This research was supported by the internal funding from Max Planck Society.