Target Audience

Scientists who are interested in high resolution fMRI, cerebrovascular imaging and optogenetics.

Introduction

Cortical states profoundly regulate many aspects of behavior, from states of consciousness to perception, learning and cognition¹. The thalamus contains ~40 nuclei, each innervating a different combination of cortical areas². It remains challenging to characterize the precise functional projections of individual thalamic subdivisions to the cortex despite the existing resting-state functional connectivity mapping schemes³. Optogenetic fMRI allows visualization of the causal effects of specific neural projection circuits defined not only by genetic identity and cell body location, but also by axonal projection target⁴. Therefore, an MRI-guided robotic arm (MgRA) positioning system with high targeting accuracy and precision is crucial to target optic fiber at different thalamic nuclei and map the BOLD-fMRI functional patterns globally.

Methods

An MRI-compatible MgRA positioning system was developed for the 14.1T horizontal MR scanner with 12cm inner diameter gradient (Fig.1a). Fig.1c illustrates the camera-based visual guidance of fiber optic positioning on the rat skull craniotomy at the iso-center of the horizontal bore magnet. Driven by MgRA, the optic fiber was inserted to target multiple coordinates in the thalamus. A 24mm-diameter custom-designed transmit/receive surface coil was attached on the rat head. Anatomical images for fiber location were acquired using 2D rapid acquisition with relaxation enhancement sequence: TR, 1200ms, TE, 7ms, 1.92cmX1.50cm FOV, 128X100 matrix, 150X150um in-plane resolution, 800um thickness. fMRI scans were performed using 3D Echo planar imaging sequence: TR, 1.5s, TE,11.5ms, 1.92X1.92X1.92cm³ FOV, 48X48X48 matrix, 400X400X400 um³ spatial resolution. Fig.1b demonstrates three continuous MRI images with step distance 400um. AAV5.CAG.ChR2.mcherry viral vectors were stereotaxically injected in the thalamus in 4-week old rats. The detailed surgical procedures for optogenetic fMRI were described previously⁵. The optical fiber(~200um) delivered blue light pulses (473nm) at 5Hz, 10ms width with 4s duration for the fMRI block design. The EPI-fMRI data were acquired in a stepwise manner when the fiber optic was inserted at different locations of the thalamus. MRI data analysis was performed using Analysis of Functional NeuroImages software(NIH, Bethesda).

Results

Fig 2 shows the broad cortical activation after the optogenetic stimulation of the thalamus. Although the fiber optic was located at the dorsal part of the thalamus, the BOLD signal spread through the major cortical regions in the ipsilateral hemisphere, as well as a small part of the contralateral hemisphere. In addition, the BOLD signal could be detected at the thalamic regions close to the fiber tip. Fig 3 shows the BOLD signal changes of the somatosensory cortex when the fiber optic was lowered at different depth of the thalamus. It clearly showed that the amplitude of the BOLD signal approached to peak when the fiber optic was at the depth (trial # 37) and gradually decreased when the fiber was inserted to deeper thalamic area ( trial # 55).

Conclusion

The optogenetically stepwise activation of different thalamic nuclei can be achieved by the MgRA positioning system. This work shows that MgRA provides a precise and effective method to target multiple deep brain nuclei and makes it possible to map the functional activation patterns consecutively along the fiber insertion trajectory.

Acknowledgements

We thank Mr. Shanyi Yu for building up the first prototype of the robotic arm and Mr. Johannes Boldt for helping to improve the MgRA system. The financial support of the Max-Planck-Society and the China Scholarship Council (PhD fellowship to Yi Chen) are gratefully acknowledged.