Optogenetic Tractography for anatomo-functional characterization of cortico-subcortical neural circuits in non-human primates

Abstract

Dissecting neural circuitry in non-human primates (NHP) is crucial to identify potential neuromodulation anatomical targets for the treatment of pharmacoresistant neuropsychiatric diseases by electrical neuromodulation. How targets of deep brain stimulation (DBS) and cortical targets of transcranial magnetic stimulation (TMS) compare and might complement one another is an important question. Combining optogenetics and tractography may enable anatomo-functional characterization of large brain cortico-subcortical neural pathways. For the proof-of-concept this approach was used in the NHP brain to characterize the motor cortico-subthalamic pathway (m_CSP) which might be involved in DBS action mechanism in Parkinson's disease (PD). Rabies-G-pseudotyped and Rabies-G-VSVg-pseudotyped EIAV lentiviral vectors encoding the opsin ChR2 gene were stereotaxically injected into the subthalamic nucleus (STN) and were retrogradely transported to the layer of the motor cortex projecting to STN. A precise anatomical mapping of this pathway was then performed using histology-guided high angular resolution MRI tractography guiding accurately cortical photostimulation of m_CSP origins. Photoexcitation of m_CSP axon terminals or m_CSP cortical origins modified the spikes distribution for photosensitive STN neurons firing rate in non-equivalent ways. Optogenetic tractography might help design preclinical neuromodulation studies in NHP models of neuropsychiatric disease choosing the most appropriate target for the tested hypothesis.

Figure 1

Experimental design. (a) At D0 a bilateral MRI-compatible chamber was chronically implanted above frontal cortex of NPH. Double channels MRI compatible cannulas were implanted ending typically 1 mm above the dorsolateral aspect of the left and/or right STN. (b) At D15, baseline fluorescence within STN was detected in vivo with a fibre-coupled minicube (NHP 3). Photostimulation above STN and electrophysiological recordings within the STN were performed as control experiments. Finally, an EIAV-Rabies-CaMK2-ChR2-eYFP lentiviral vector with retrograde transfer properties was injected unilaterally inside the STN. (c) At D75, ChR2-eYFP associated fluorescence was detected in vivo with a fibre-coupled minicube in STN (NHP 3), in order to confirm transduction (green). (e) At D75, in NHP 1 & 2, a fibre optics was lowered through one channel and set typically 1 mm above STN. An electrode was lowered through the other channel. The distance between the two channels main axes was 0.8 mm. High frequency blue photostimulation was performed above the dorsolateral STN, and STN neurons were recorded along a dorso-ventral axis. (f) At D75, in NHP 3, a fibre optics was lowered in the motor cortex through a craniotomy, guided by tractography reconstructions, and electrodes were lowered inside the ipsilateral STN through both channels. High frequency blue photostimulation was performed in the prmary motor cortex while STN neurons were recorded along a dorso-ventral axis at the same time. D: Day; MRI: magnetic resonance imaging; NHP: non-human primate; STN: subthalamic nucleus.

Figure 2

Histological characterization of the lentiviral vector enabling retrograde transport of ChR2. (a) EIAV-Rabies-CaMK2-ChR2-eYFP and EIAV-VSVg-Rabies-CaMK2-ChR2-eYFP lentiviral vectors with retrograde transfer properties were injected unilaterally in the STN of NHP 1 & 2, and NHP 3 respectively. Cannula traces were visualized on histological slices and confirmed to end above the dorsolateral aspect of STN of NHP 1, 2 & 3. (b) Injection sites for NHP 1 & 2 are 3D-reconstituted in a common STN after affine and elastic transformations between the two STN for NHP 1 & 2. (c) i/ high magnification on a coronal slice showing right STN, thalamus and putamen. ii/ white rectangle around STN on i/ is zoomed. White dots dlimit the STN. Yellow dashes indicate the recording electrode trajectory in the prolongation of the cannula. iii/ the biggest white rectangle in ii/ is zoomed: axons expressing ChR2-eYFP arriving to STN. iv/ the smallest white rectangle in ii/ is zoomed: VgluT1 positive cortico-subthalamic axon terminals expressing ChR2-eYFP. (d) i/ zoom on a coronal motor cortex slice. ii/ white rectangle in i/ is zoomed: neurons of cortical layer V express ChR2-eYFP on their somas, dendrites and initial portion of their axons. iii/ dendrites expressing ChR2-eYFP. (e) Double immune-staining GFP/NeuN and GFP/GFAP in the layer V of motor cortex. White arrows indicate GFP-positive cells. White triangles indicate GFAP-positive cells NHP: Non-human primate; STN: subthalamic nucleus; Thal: thalamus; Put: putamen; VgluT1: vesicular glutamate transporter 1.

Figure 3

Three dimensional reconstruction of ChR2-eYFP associated fluorescence at the cortical origins of the motor-cortico-subthalamic tract. (a) Whole brains of NHP 1 & 2 were sliced. Every eight 40-µm thick coronal slice was photographed, and a whole brain histological block was reconstituted. ChR2-eYFP fluorescence detected with a microscope after immunohistochemistry was segmented on each slice. (b) ChR2-eYFP motor cortex fluorescence were 3D-reconstructed for NHP 1 (green) and NHP 2 (blue) and were expressed in a common space. (c) Six motor cortex functional regions were found to express fluorescence. Distribution of ChR2-eYFP fluorescence in the motor cortex was computed. (d) For each of these six regions, the proportion of the region expressing fluorescence was calculated. CMA c-v: caudo-ventral cingular motor cortex; CMA c-d: caudo-dorsal cingular motor cortex; CMAr: rostral cingular motor cortex; M1: primary motor cortex; NHP: non-human primate; PreM: premotor cortex; SMA-PreSMA: supplementary motor area-presupplementary motor area.

Figure 4

Histologically optimized Tractography in non-human primate. (a) In the same two non-human primates (NHP 1 & 2), in vivo tractography reconstructions after high angular resolution diffusion imaging (HARDI) acquisitions were performed for the pathway connecting the motor cortex to the lentiviral injection sites for various sets of reconstruction parameters. For these same animals, ChR2-eYFP associated cortical fluorescence was mapped in the motor cortex. Tractography and histological results were projected on cortical surface and experessed in a common space. (b) The optimal parameters (Δ, θ) for tractography reconstruction for each motor cortical area were those for which the average across NHP1 & 2 of the surface within the considered motor cortical area with local correlation was maximal (NHP 1 & 2). θ is the angulation between the normal to the cortical surface and the axis of fibers leaving the cortex. Δ is the depth from cortical surface for cortical origins of the studied fiber pathway. The probablity map of connection figure and the table illustrate the connectivity between primary motor cortex and the lentiviral injection sites in STN. The table displays values of the primary motor cortex surface (in mm², (sem)) for which the correlation coefficient between tractography and histological cortical maps of connectivity is superior to 0.5, for various sets of reconstruction parameters. The values of Δ νδ θ for which the surface is maximal are considered to be the optimal reconstruction parameters between STN and the considered area of motor cortex, M1 in this case. Such parameters are determined for each considered cortical motor area and detailed in Table 1. (c) Determination of the optimal tractography reconstruction parameters between STN injection sites and motor cortical areas in NHP 1 & 2 thanks to comparison with histological findings enabled accurate determination of the connectivity between whole STN and motor cortical areas for both hemispheres of NHP 1, 2 & 3 on one hand, and to guide photo-detection and -stimulation of the primary motor cortex area with the highest density of connections with STN injection sites in NHP3 on the other hand. M1: primary motor cortex; m_CSP: motor corticosubthalamic pathway; NHP: non-human primate; sem: standard error of the mean; STN: subthalamic nucleus.

Figure 5

Electrophysiological validation of the lentiviral vector and functional characterization of the motor cortico-subthalamic tract in non-human primate. (a) Multi-unit recordings in the subthalamic nucleus without (upper trace) or with (lower trace) high frequency blue light photostimulation (NHP 2). (b) Identified after spike sorting by thresholding and principal component analysis, a unit was considered photosensitive by assessing the goodness of fit between the inhomogeneous poisson processes modelling the time series of its spikes occurrences during the OFF-prestimulation, ON-stimulation periods and post-stimulation periods (Supplementary Methods, section 6). Here the fit between poisson processes models for the OFF-prestimulation and the ON-stimulation periods of an STN unit is bad according to the four Ogata’s tests of goodness of fit. (c) The percentage of STN neurons that were photosensitive was determined during axon terminals (NHP 1 & 2) or m_CSP cortical origins (NHP 3) photostimulation. (d) Effects of cortico-subthalamic tract cortical origins (NHP 3, blue diamonds) or axon terminals (NHP 1 & 2, red triangles) photoexcitation on firing rate of STN photosensitive neurons. m_CSP: motor corticosubthalamic pathway; NHP: non-human primate; STN: subthalamic nucleus.