Optogenetics allow spatial and temporal control of the activity of defined neuronal populations in the living brain. Coupled to Blood Oxygen Level Dependent (BOLD) functional magnetic resonance magnetic resonance imaging (fMRI) this recent modality could achieve unprecedented understanding of neural circuit function. Nevertheless, questions remain regarding the maintenance of neurons in their physiological range when delivering strong laser light to specific regions. Here, experiments were conducted to investigate the effects of blue laser light on metabolites in the forelimb region of primary somatosensory cortex (S1FL) in rats using proton magnetic resonance spectroscopy (1H-MRS).

3 female Fisher rats (160±10g) underwent a craniotomy 4 mm lateral to the bregma at a depth of 0.1 mm in order to insert an optical fiber into the S1FL, according to the Paxinos and Watson atlas. Algnate was used to cover the craniotomy during MRI. 2 naive rats served as controls. Surgical procedures were performed under isoflurane anesthesia (2-2.5% in a mixture of air and oxygen). During MRI, body temperature and respiration rate were continuously monitored, using a rectal probe and a pressure sensor. Temperature was maintained at 36°± 1°C using warm water circulation tubing placed underneath the rat. Stainless steel electrodes were inserted in the right forepaw for electrical stimulations.

MRI and 1H-MRS were performed at 9.4T (Biospec 94/20, Bruker GmbH, Ettlingen, Germany) using a transmit volume coil and a 10mm receive surface coil placed over the S1FL and the optical fiber implant. Each animal was positioned in a dedicated cradle equipped with ear and bite bars. Experiments were conducted under medetomidine sedation (subcutaneous bolus injection of 0.04 mg/kg followed by a continuous infusion of 0.05mg/kg) 30 minutes after discontinuation of isoflurane anesthesia. After anatomical imaging using a 2D RARE sequence and shimming using MAPSHIM, BOLD fMRI was performed using a single shot gradient echo EPI sequence (TR/TE=1000/22 ms; BW=300KHz ; 9 slices ; Slice thickness=0.8mm ; FOV=228x26mm2 ; Matrix=80x80). Electrical stimulation was performed using a repeated 10sOFF- 10sON-10sOFF block paradigm. 5 ms pulses were delivered at 9Hz and 2mA by a stimulator controlled by an in-house written LABVIEW program. Blue light was delivered through an optic fiber and controlled with the same program, using 10ms light pulses above 22mW mm-2 at 9Hz. PRESS 1H-MRS was conducted using TR/TE=4000/13.55ms. NA=512 ; spectral width=4960.32Hz and 4096 points with carefully optimized OVS pulses and VAPOR pulses for water suppression. Spectra were acquired in a 1.5x1.5x3mm3 volume of interest localized over the activated S1FL region previously determined with fMRI and encompassing the tip of the optical fiber. The total acquisition time was 34 minutes including unsuppressed water signal acquisition in the same VOI. Prior to PRESS acquisitions, first and second order FASTMAP shimming were performed over the same VOI. fMRI series were analyzed using an in-house written ImageJ analysis script, performing voxel-wise t-test between stimulation and rest periods. Statistical significance level was set at pvalue<0.01. Average fids acquired with PRESS were Fourier-transformed, water scaled and fitted with LCmodel using a simulated basis set.

BOLD responses were observed in S1FL upon electric fore paw stimulation (Fig.1B) and upon light delivery in S1FL at the fiber tip, potentially representing heat-induced BOLD effects (Fig.1C). Despite the presence of the optical fiber within the VOI (Fig.1A), we were able to shim down to a water linewidth of 15±3 Hz,allowing good water suppression and acquisition of high-quality spectra (SNR=46±3) during both 34 minute- electrical forepaw and light stimulations. Fig. 2A shows superimposed labeled spectra acquired without any stimulation (black) and during electrical forepaw (blue) and blue laser (red) stimulations in the same rat. Cramer-Rao lower bounds were under 35% for all the quantified metabolites. Upon electrical forepaw stimulation, glucose, glutamate and aspartate concentrations dropped by 25.6, 8.6 and 4.6%, respectively, while glutamine and lactate increased by 7.1 and 22.7%, respectively, in presence of the optical fiber (Fig.2B). In addition, phosphocreatine (PCr) decreased by 22% while creatine (Cr) increased by 24%. Similar results were obtained in naïve animals without implanted fiber (Fig.3). These results were in agreement with previous findings obtained during cortical stimulation in the rat (1). During blue laser stimulation, most metabolites (glucose, glutamate, lactate) returned to their resting levels. PCr and Cr further decreased (-52%) and increased (+43%), respectively relative to resting levels.First results indicate that a superficially implanted optical fiber and low power laser light delivery in the rat cortex do not disturb 1H-MRS acquisitions. Nonetheless, the S1FL neurochemical profile measured during light stimulation differed from neurochemical profiles measured during rest and electrical stimulation, thus requiring further validation.