Transgenic mice
Mice (CD-1, male, 35 – 50 g, N = 64 total, n = 15 for each mouse line) used for experiments in this study were heterozygous for Drd3-Cre (L2/3), Scnn1-Cre (L4), Rpb4-Cre (L5), or Ntsr1-Cre (L6). All transgenic mouse lines were obtained from the Adesnik Lab, UC Berkeley9,23-25. Mice were housed in cohorts of five or less with a light:dark cycle of 12:12 h with access to food and water ad libitum. All experimental procedures and animal husbandry were performed in strict accordance with the NIH and Stanford University Institutional Animal Care and Use Committee guidelines. For each transgenic mouse line, 12 were used for ofMRI, 1 for histology and immunohistochemistry, and 4 for electrophysiological recordings.
Viral Expression and Stereotaxic Surgery
A double-floxed inverted open reading frame (DIO) recombinant AAV5 virus was used to express ChR2-EYFP in Cre-expressing neurons. The recombinant AAV vector was packaged by the University of North Carolina viral vector core (titer of 4 × 1012 particles/mL). All stereotactic surgeries were performed with mice at 8 – 10 weeks of age. Animals were anesthetized with isoflurane (induction 5%, maintenance 1.5% – 2%; Sigma-Aldrich) and secured in a stereotactic frame with nonrupturing ear bars (Kopf Instruments). A heating pad was used to maintain body temperature, and sterile ocular lubricant was applied to the eyes to prevent desiccation during surgery. Buprenorphine (1 mg/kg) was injected subcutaneously for analgesia. After a midline incision along the scalp, a small craniotomy and viral injection/cannula implantation were performed at the primary motor cortex (+ 0.75 mm AP [anterior-posterior], − 1.5 mm ML [medial-lateral], injection at + 0.15 mm, + 0.35 mm, + 0.6 mm, or + 0.8 mm DV [dorsal-ventral] for L2/3, L4, L5, and L6 respectively). 1.5 μL of the AAV5/DIO-ChR2–EYFP virus was delivered using a 10 μL syringe with a 34G metal needle (World Precision Instruments Inc.) at a 75 nL/min flow rate driven by a micro-syringe pump controller. Following injection, the injection needle was held in place for 10 min before slowly retracting it from the brain. A custom-designed fiber-optic cannula was mounted and secured on the skull using light-cured dental cement (Kuraray Inc.), with the optical fiber extending from the cannula’s base to the desired depth (0.1 mm above the injection site). Following surgery, mice were kept on a heating pad until recovery from anesthesia and were given carprofen (5 mg/kg, subcutaneously [s.c.]) daily for 2 days to minimize post-operative discomfort. All experiments were conducted at least 4 weeks following virus injection to ensure optimal ChR2 expression. Probe locations were validated in all animals used for ofMRI experiments with T2-weighed structural MRI images.
Histology and immunohistochemistry to validate expression of ChR2
To confirm the precise targeting of ChR2 to L2/3, L4, L5 or L6, a cohort of mice injected with the DIO-recombinant virus was deeply anesthetized with pentobarbital and transcardially perfused with 0.1 M PBS followed by ice-cold 4% paraformaldehyde (PFA) in PBS. Brains were extracted and fixed in 4 % PFA overnight at 4 °C. The brains were equilibrated in 20 % sucrose in PBS at 4 °C overnight. Axial sections of 50 μm thickness were prepared on a freezing microtome (Micron HM550 Microtome, Thermo Scientific Inc.). For immunohistochemistry, free floating sections were washed with 0.1 M PBS for 20 mins at room temperature. Sections were then exposed to 200 ng/ml of DAPI in PBS at room temperature for 20 mins. Slices were then washed and mounted using Fluoromount-G (SouthernBiotech, Birmingham, AL). Immuno-fluorescence was assessed with a laser confocal microscope (LSM 880 inverted confocal with Airyscan, Zeiss Inc.) at the Cell Sciences Imaging Facility at Beckman Center for Molecular and Genetics Medicine. The images were then analyzed to validate the expression of ChR2. Since the primary motor cortex is 1000 µm thick, and since the AAV5 spread is roughly 500 µm, region of interest (Figure 1A) was defined as a 500 µm × 1000 µm (width × height) rectangular box centered at the injection site. Anatomical landmarks including the midline, corpus callosum, cortex and striatum were identified and the ROI was defined based on the mouse brain atlas and injection site location to ensure consistency. For sensitivity, 12%, 15%, 23% and 15% of cells identified within the injection area were ChR2-EYFP-positive for Drd3-Cre (L2/3), Scnn1-Cre (L4), Rpb4-Cre (L5), or Ntsr1-Cre (L6) mouse lines, respectively. All mouse lines have 100% specificity.
ofMRI Experiments
All MRI experiments were carried out on a 7 Tesla Bruker-converted Biospec small animal MRI system at the SCI3 facility at Clark Center using a transmit-only birdcage coil in combination with a custom-designed actively decoupled receive-only surface coil. Animals were initially anesthetized in an induction chamber with 5% isoflurane before placement into the restraining apparatus. The restrained animal was then placed onto the MRI-compatible cradle with ears, teeth, and head secured. To maximize signal-to-noise ratio, the receiver coil was first placed on top of the head and centered over the brain and then placed into the iso-center of the magnet. The animals were lightly anesthetized and sedated using a combination of isoflurane and dexmedetomidine (0.25% isoflurane mixed with O2 and medical air; an initial bolus of 0.1 mg/kg dexmedetomidine; continuous infusion at 0.1 mg/kg/hr during scanning). During all fMRI experiments, continuous physiological monitoring was performed using an MRI-compatible system (SA Instruments). Vital signs were within normal physiological ranges (rectal temperature: 36.5 – 37.5 °C, heart rate: 260 – 420 beat/min, breathing: 80 – 120 breath/min, oxygen saturation: > 90%) throughout the duration of the experiments.
Fourteen contiguous 1.0-mm slices were positioned in the transverse orientation according to the mouse brain atlas to cover the whole brain. T2-weighted high-resolution anatomical images were acquired prior to fMRI to check for brain damage and confirm accurate probe location. These anatomical images were acquired using a rapid acquisition with relaxation enhancement (RARE) sequence with field of view (FOV) = 20 × 20 mm2, matrix = 256 × 256, RARE factor = 8, echo time (TE) = 33 ms, repetition time (TR) = 4,000 ms. ofMRI measurements were obtained for the same slices using a single-shot Gradient-Echo Echo-Planar-Imaging (GE-EPI) sequence with FOV = 20 × 20 mm2, matrix = 75 × 75, flip angle = 40°, TE = 14 ms, TR = 1,000 ms.
The MRI scanner and laser for optogenetic stimulation were synchronized using a Master pulse stimulator system (A.M.P.I.). The light delivery system was kept outside the magnet and long optical cables (~5 m) were used to deliver light into the scanner bore. Blue light was delivered using a 473-nm DPSS laser measured before scanning as 3 mW at the fiber tip (250 μm) corresponding to a light intensity of 60 mW/mm2 for optogenetic stimulation. To determine the frequency-dependent spatiotemporal characteristics of evoked M1 layer-specific responses, four frequencies were used (5 Hz, 10 Hz, 20 Hz, and 40 Hz with 30% duty cycle). M1 layer-specific neurons were stimulated with a standard block design paradigm that consisted of 10-s light-on and 30-s light-off periods. Three to four trials were recorded for each frequency in a pseudo-random manner in each animal.
fMRI Data Analysis
For each animal, all fMRI images were corrected for slice timing differences, realigned to the first image of the first fMRI session, and spatially smoothed with a Gaussian kernel (FWHM = 2 pixels) using SPM12 (Wellcome Department of Imaging Neuroscience, University College, London). The fMRI data were then temporally linear detrended and temporally band-pass filtered (0.01-0.25 Hz) to eliminate baseline drift caused by system instability and physiological noise. For detrending, linear trend was first calculated using linear regression of the global temporal signal obtained from the whole brain. Then, detrending was applied to the temporal signal of each voxel using the obtained linear trend. T2-weighted images from each animal were registered to a custom-made brain template acquired with the same settings. Registration was performed by affine transformation and Gaussian smoothing to maximize normalized mutual information (SPM12). Only then were the fMRI images resliced correspondingly.
A general linear model (GLM) was applied. Student’s t test was performed to identify activated voxels using the threshold p < 0.001. The mapped responses were then compared between each animal to assess result quality before group level analysis. At the group level, realigned, registered, and resliced images corresponding to the same fMRI session were averaged across all animals. To quantify the activations from the activation maps, t-values within the ROI were extracted and averaged, while fraction of ROI positively and negatively modulated were calculated with a voxel-wise threshold of p < 0.05 as positively or negatively modulated. BOLD temporal profiles for each stimulation frequency were extracted from identical ROIs delineated from the mouse brain atlas by regional averaging across the whole ROI. Area under the curve of the BOLD profiles was calculated to quantify the strength of the BOLD responses for each individual animal, and then averaged across animals within the same group.
In Vivo Electrophysiology
In vivo electrophysiology was performed to directly measure the neuronal activity of the ipsilateral M1, CPu and VL. Similar to ofMRI experiments, the animals were lightly anesthetized and sedated using a combination of isoflurane and dexmedetomidine (0.25% isoflurane mixed with O2 and medical air; an initial bolus of 0.1 mg/kg dexmedetomidine; continuous infusion at 0.1 mg/kg/hr during experiment). Throughout the procedure, body temperature was maintained at 37 °C using a thermoresistive heating pad (FHC, Inc., ME, USA). After securing the animal within a stereotactic frame, an optrode composed of an optical fiber glued to the 16-channel linear-array electrode (NeuroNexus Technologies, MI, USA; A1x16-5mm-150-703-A16 electrode array) was inserted at the stimulation site (+ 0.75 mm AP [anterior-posterior], − 1.5 mm ML [medial-lateral], covering between +0.00 mm and +2.40 mm DV [dorsal-ventral]) and a 16-channel microelectrode array (NeuroNexus Technologies, MI, USA; A1x16-5mm-50-703-A16 electrode array) was inserted at the ipsilateral VL (-1.35 mm AP, +1.00 mm ML, covering between +3.30 mm and +3.70 mm DV). Identical optogenetic stimulation paradigms were applied. Recordings were acquired at 40 kHz sampling rate using the Plexon OmniPlex system with PlexControl software (Plexon Inc., TX, USA). The Plexon multichannel acquisition processor was used to amplify and low-pass filter (cutoff frequency: 200 Hz) the acquired signal for local field potential (LFP) recordings or high-pass filter (cutoff frequency: 300 Hz) for spike recordings.
Electrophysiology Data Analysis
For LFP recordings, raw data were initially imported into custom-written Matlab software and a notch filter centered at 60 Hz was applied to remove power-line noise. Individual animal LFPs were averaged across stimulation blocks to generate a single LFP trace for each stimulation frequency, which was then averaged across animals. For spike recordings, two-tailed paired t-tests were used to identify significant changes in firing rate within each unit between pre-stimulation and stimulation periods (10 seconds each). Only units with significant change (either increase or decrease) were included in the following analysis. Peri-event time histograms were averaged across stimulation blocks and units to generate an average peri-event time histograms for each stimulation frequency. One-way ANOVA followed by Bonferroni’s post hoc test was applied to compare pre-stimulation, stimulation and post stimulation periods.
Although BOLD-fMRI and LFP responses reflects the overall signal within a region, which all recorded units should be considered when exploring the association between spiking, BOLD-fMRI and LFP, we still examined different spiking parameters namely percentage of positively modulated neurons, percentage of negatively modulated neurons, percentage of modulated neurons, average percentage spike rate change of positively modulated neurons,, average percentage spike rate change of negatively modulated neurons, average percentage spike rate change of all modulated neurons, average absolute value of percentage spike rate change of all modulated neurons, average percentage spike rate change of all neurons, average absolute value of percentage spike rate change of all neurons, total percentage spike rate change of positively modulated neurons, total percentage spike rate change of negatively modulated neurons, total percentage spike rate change of all modulated neuron, total absolute value of percentage spike rate change of all modulated neurons, total percentage spike rate change of all neurons, and total absolute value of percentage spike rate change of all neurons. The summary of the R2 values of the scatter plots between these spiking parameters and brain-wide BOLD-fMRI (Supplementary Table 7), between these spiking parameters and long-range LFP recordings (Supplementary Table 8), between these spiking parameters and laminar BOLD-fMRI (Supplementary Table 9), and between these spiking parameters and laminar LFP recordings (Supplementary Table 10) generally showed that the spiking parameter, total absolute value of percentage spike rate change of all neurons, tends to yield the highest correlation.