TRPV1-mediated sonogenetic neuromodulation of motor cortex in freely moving mice

Abstract Background. Noninvasive and cell-type-specific neuromodulation tools are critically needed for probing intact brain function. Sonogenetics for noninvasive activation of neurons engineered to express thermosensitive transient receptor potential vanilloid 1 (TRPV1) by transcranial focused ultrasound (FUS) was recently developed to address this need. However, using TRPV1-mediated sonogenetics to evoke behavior by targeting the cortex is challenged by its proximity to the skull due to high skull absorption of ultrasound and increased risks of thermal-induced tissue damage. Objective. This study evaluated the feasibility and safety of TRPV1-mediated sonogenetics in targeting the motor cortex to modulate the locomotor behavior of freely moving mice. Approach. Adeno-associated viral vectors was delivered to the mouse motor cortex via intracranial injection to express TRPV1 in excitatory neurons. A wearable FUS device was installed on the mouse head after a month to control neuronal activity by activating virally expressed TRPV1 through FUS sonication at different acoustic pressures. Immunohistochemistry staining of ex vivo brain slices was performed to verify neuron activation and evaluate safety. Results. TRPV1-mediated sonogenetic stimulation at 0.7 MPa successfully evoked rotational behavior in the direction contralateral to the stimulation site, activated cortical neurons as indicated by the upregulation of c-Fos, and did not induce significant changes in inflammatory or apoptotic markers (GFAP, Iba1, and Caspase-3). Sonogenetic stimulation of TRPV1 mice at a higher acoustic pressure, 1.1 MPa, induced significant changes in motor behavior and upregulation of c-Fos compared with FUS sonication of naïve mice at 1.1 MPa. However, signs of damage at the meninges were observed at 1.1 MPa. Significance. TRPV1-mediated sonogenetics can achieve effective and safe neuromodulation at the cortex with carefully selected FUS parameters. These findings expand the application of this technique to include superficial brain targets.


Introduction
The evolution of brain neuromodulation tools has provided unprecedented opportunities to probe neural circuits, understand brain function, and develop new treatment strategies for brain diseases. Transcranial neuromodulation tools, such as direct current, magnetic stimulation, and ultrasound stimulation, offer noninvasive ways to stimulate the brain and have contributed to the understanding of brain function [1,2]. The lack of cell-type specificity in these tools, however, limits their utility in understanding the brain at cellular resolution. Genetic-based neuromodulation tools, such as optogenetics and chemogenetics, encode stimulussensitive probes into a defined neuron population and have transformed fundamental neuroscience research [3,4]. Each method, however, suffers from its own limitations. Most commonly, optogenetics requires the invasive implantation of optical probes to deliver light to opsin-encoding neurons, limiting the ability to study the brain without the risk of ischemia and inflammation. Noninvasive optogenetics modulates the activity of opsin-encoding neurons via transcranial illumination, but light scattering in brain tissue limits its depth penetration in large animal models [5]. On the other hand, chemogenetics noninvasively activates neurons encoding designer receptors exclusively activated by designer drugs via minimallyinvasive systemic delivery of designer drugs, but the long residence time of circulating drugs sacrifices the temporal resolution of this technique. There is a clear need for techniques that can facilitate noninvasive, cell-type specific neuromodulation with high spatiotemporal resolution and the potential to be scaled up to large animals and humans.
Ultrasound propagation in tissue can generate not only mechanical effects but also thermal effects. The transient receptor potential vanilloid 1 (TRPV1) ion channel is extremely sensitive to temperature and has a thermal activation threshold of approximately 42 • C, which is only a few degrees above the physiological body temperature of many mammals [20]. Such an activation temperature allows TRPV1 to be closed at the physiological body temperature and open upon sufficient heating to ∼42 • C. Additionally, the single-channel conductance of TRPV1 is ∼1000-fold greater than that of a channelrhodopsin, a commonly used optogenetic ion channel [21]. TRPV1 can drive robust activation even at low expression levels, minimizing potential toxicity associated with the expression of exogenous proteins. Because of these unique features, TRPV1 has been used to develop genetics-based neuromodulation techniques, such as magneto-thermogenetics [22,23] and photothermal genetics [24]. TRPV1-mediated sonogenetics was recently developed to achieve noninvasive, cell-type specific neuromodulation. Our previous study demonstrated that TRPV1 is an ultrasoundsensitive actuator, and that TRPV1-mediated sonogenetics can control the motor behavior of freely moving mice by targeting a deep brain region, the striatum [25]. However, the capability of TRPV1mediated sonogenetics in controlling mouse behavior by targeting the superficial brain area has not been demonstrated. Targeting deep regions using a technique based on ultrasound-induced heating will induce heating at the target region and the skull, which are spatially separated [26]. Targeting superficial brain regions, however, is challenging for TRPV1mediated sonogenetics because the high absorption of ultrasound in the skull could contribute additional heating to the cortical regions directly underneath the skull [27,28]. This could increase the risk of undesirable neuromodulatory effects associated with heating and potential tissue damage when the temperature is high. Therefore, the objective of the current study was to assess the capability of TRPV1-mediated sonogenetics in evoking mouse motor behavior by targeting a superficial brain target-the motor cortex.

Stereotaxic injection of viral vectors
All animal procedures were performed under a protocol approved by the Washington University in St. Louis Institutional Animal Care and Use Committee. C57/BL6 mice (female, 6-8 weeks old) were purchased from Charles River and housed in an animal facility under a 12 h light-dark cycle. Adenoassociated viral vectors (AAV) were introduced to CaMKII-expressing neurons of the M2 cortex to overexpress TRPV1 ion channel. Cloning, packaging, purification, and viral titer calculations were performed by the Hope Center Viral Vectors Core at Washington University School of Medicine. The CaMKII-TRPV1 sequence was introduced into the AAV5 recombinant genome flanked by inverted terminal repeat sequences. The control sequence did not contain TRPV1. All surgeries were conducted under aseptic conditions. Mice were anesthetized with 2% isoflurane in oxygen in an anesthetic chamber for induction and 1.5% isoflurane for maintaining anesthesia. Anesthetized mice were then fixed onto a stereotaxic frame (Kopf Instruments) using a bite bar and ear bars. Buprenorphine SR (1.0 mg kg −1 ) was administered subcutaneously for pre-operative and post-operative pain management. The head was shaved and was rubbed with skin disinfectant (Hibiclens). An incision was made on the scalp, the skin was retracted, and the periosteum was removed. A small hole was drilled through the skull (−1.0 mm ML, +2.5 mm AP, −1.0 mm DV), and a micro-injector (Nanoject II, Drummond Scientific) was inserted into the motor cortex. 1200 nL of TRPV1 virus (1.4 × 10 12 vg ml −1 ) was introduced at a rate of 64 nL min −1 . 1000 nL of control virus (3.2 × 10 12 vg ml −1 ) was introduced to approximately match the viral genome copy numbers delivered to the motor cortex. After injection, the micro-injector was slowly removed, the hole was filled with bone wax, and the scalp was sutured. Mice were housed for at least 4 weeks to facilitate sufficient virus expression before further treatments were conducted.

Wearable FUS device
A wearable FUS device was used to stimulate the motor cortex of freely moving mice, and the design is described in our previous report of TRPV1-mediated sonogenetics [25]. In brief, the wearable FUS device consisted of two parts: a FUS transducer and a base plate. The FUS transducer was made of a lead zirconate titanate (PZT) ceramic resonator (DL-43, DeL Piezo Specialties) encapsulated by a 3D-printed housing. The PZT ceramic resonated at a frequency of 1.5 MHz and had an aperture of 10 mm and a radius of curvature of 10 mm. The wearable FUS transducer was plugged into the base plate, a 3D printed circular adapter attached to the mouse skull. When the FUS transducer was plugged into the base plate, the wearable FUS device was stabilized on the mouse head.
Each component of the wearable FUS device was specifically designed to target the motor cortex. The base plate was designed with a hole in its geometric center to facilitate the alignment of base plate to the medial-lateral and anterior-posterior coordinates of the motor cortex. The height of the FUS transducer housing was designed to align the FUS focus to the dorsal-ventral coordinates of the motor cortex. The entire wearable FUS device was calibrated by a hydrophone (HGL-200, Onda). The full width halfmaximum of the FUS focal region was approximately 0.9 mm and 2.5 mm in the lateral and axial directions, respectively.

Attachment of FUS transducer base plate to the mouse skull
Four to five weeks after virus injection, mice were again anesthetized with isoflurane (2% for induction, 1.5% for maintenance), fixed in a stereotaxic frame, and subcutaneously administered with Buprenorphine SR (1.0 mg kg −1 ). A piece of the scalp was removed, the periosteum was removed, and the drilled hole from the intracranial injection of AAV was identified and accentuated with a marker. The custom designed base plate was 3D-printed and glued onto the skull using dental adhesives (Metabond) with the center of the base plate aligned to the predrilled hole. The mice were housed for a week to facilitate sufficient recovery before performing behavior experiments.

FUS stimulation with behavior recording
Prior to the behavior test, mice were adapted to the behavior recording environment by placing the mouse in the behavior testing arena with the power amplifier turned on. During the behavior recording, mice were lightly anesthetized with isoflurane (1% induction and maintenance). The base plate on the mouse and the wearable ultrasound transducer were both sufficiently filled with degassed ultrasound gel (Aquasonics). The wearable transducer was then securely plugged into the base plate of the mouse, and the mouse was then placed in a circular arena on a heating pad for 30 min to allow the mouse body temperature to recover from any possible anesthesia effects. The heating pad was then removed, and the mouse was allowed to habituate for 15 min in the actual behavior test arena.
During the recording period, FUS was applied at a frequency of 1.5 MHz, duty cycle (DC) of 40%, pulse repetition frequency (PRF) of 10 Hz, and 15 s total sonication duration with 185 s inter-stimulation interval (ISI) for a total of five stimulations. The onset and offset of the ultrasound pulse was smoothed to avoid possible auditory effects [29]. The acoustic pressures used in the study were 0, 0.7, and 1.1 MPa (measured in a water tank) to investigate the effect of pressure on locomotor behavior outcomes. These parameters corresponded to mechanical indices of 0, 0.57, and 0.90 (in water), respectively. Custom MAT-LAB software was used to control when ultrasound was applied via an Arduino Uno. A red LED attached to the Arduino Uno would turn on when ultrasound was applied to precisely synchronize mouse behavior to each FUS stimulation. In each group, mice were given five consecutive FUS stimulations at one pressure.

Behavioral analysis
Mice were recorded using a camera (Logitech C920X, 30 fps) before, during, and after each FUS stimulation. During the recording session, each video is simultaneously processed using Bonsai to quantify the positional coordinates and the angular orientation of the mice. After conducting recordings, data were processed using a custom MATLAB script to compute the average angular velocities upon FUS stimulation at different acoustic pressures.

In vivo MR thermometry recording during FUS stimulation
Magnetic resonance (MR) thermometry was used to noninvasively quantify the temperature rise associated with FUS stimulation at the motor cortex. Wild-type, untreated C57/BL6 mice (10-12 weeks) were used in this experiment. The age of these mice was slightly older to match the age of the mice used in the locomotor behavior test. The FUS transducer base plate was attached to the skull (as previously described), and the mice were allowed to recover for at least a week.
At the time of MR imaging, mice were anesthetized with isoflurane (2% for induction, 1.5% for maintenance). Similar to the FUS stimulation during behavior recording, the base plate and wearable ultrasound transducer were both sufficiently filled with degassed ultrasound gel, and the wearable transducer was plugged into the base plate of the mouse. Mice remained under anesthesia for the duration of MR imaging at 1.5% isoflurane. Mice were fixed in a small animal cradle, coupled to an MRI saddle coil (Image Guided Therapy). The cradle was then inserted into a 4.7 T MRI system (Agilent/Varian Direct-Drive Console). Temperature maps were generated with a continuously applied gradient-echo imaging sequence with a flip angle of 20 degrees, T R of 10 ms, T E of 4 ms, slice thickness of 1.5 mm, and a matrix size of 128 × 128 for 60 × 60 mm field of view. Phase images were processed using ThermoGuide software (Image Guided Therapy). During MR imaging, the mouse rectal temperature was monitored throughout the duration of the experiment. The mouse body temperature was maintained at approximately 37 • C using warm air. During the ultrasound stimulation, six ultrasound stimuli were applied to the motor cortex with the same parameters used in the locomotor behavior assay.

Immunohistological analysis
Approximately 90 min after the last FUS stimulation, TRPV1− and TRPV1+ mice from each acoustic pressure stimulation group were sacrificed via transcardial perfusion with 1x PBS solution for the evaluation of TRPV1 expression, c-Fos expression, and safety of sonogenetics via inflammatory and apoptotic markers (GFAP, Iba1, and Caspase-3). A sacrifice time of 90 min post stimulation was chosen to visualize the peak expression of c-Fos [30]. This time was also suitable to visualize any rapid recruitment of inflammatory and apoptotic markers at the FUS stimulation site [31]. The brains were fixed in 4% w/v paraformaldehyde in 1x PBS solution overnight and were transferred to 15% and 30% w/v sucrose in 1x PBS for the following two days, respectively. The brain tissue was embedded in a cryomold with Optimal Cutting Temperature medium (Scigen) to generate 10 µm thick coronal brain slices affixed on a glass slide.

Statistics
Statistical tests were conducted using GraphPad (Prism). Data were analyzed using either a two-tailed t-test or repeated measures ANOVA with either Bonferroni's post-hoc test (to compare row-and columnwise groups) or Dunnett's post-hoc test (to compare to a control group). Statistical differences were considered significant whenever p < 0.05. All graphs presented results as the mean ± standard error of the mean (SEM).

Results
We intracranially injected AAV to the mouse motor cortex (M2) to express TRPV1 primarily in excitatory neurons under the CaMKII promoter ( figure 1(a)). These mice are referred to as TRPV1+ mice. Control mice were injected with TRPV1− virus, referred to as TRPV1− mice. After sufficient virus expression (4-5 weeks), a wearable FUS transducer was attached to the mouse head. Mouse locomotion was assessed before, during, and after FUS sonication in an openfield behavior test arena. FUS sonication was targeted at the motor cortex using the same coordinates as the virus injection by mechanically aligning the FUS device to the craniotomy from the virus injection. FUS was applied with a center frequency of 1.5 MHz, a PRF of 10 Hz, a DC of 40%, acoustic pressures of 0.7 and 1.1 MPa, and a burst duration of 15 s with an ISI of 185 s for a total of five stimulations ( figure 1(b)). The associated peak temperatures in the motor cortex using acoustic pressures of 0.7 and 1.1 MPa were respectively 38.57 ± 0.31 • C and 39.69 ± 0.59 • C, as measured by MR thermometry (figure 1(c)). These temperature rises were sufficient to evoke activation in neurons expressed with TRPV1 [24,25]. Mice were sacrificed after the behavior test to evaluate the expression of TRPV1, the activation of neurons (c-Fos), and safety of sonogenetics.

Characterization of exogenous TRPV1 expression in the motor cortex
We first describe the virus expression level of TRPV1 in the motor cortex. The brains of TRPV1+ mice were harvested, sectioned, and co-stained with anti-TRPV1 antibody and Nissl to evaluate the expression profile of TRPV1 in cortical neurons of the motor cortex after 4-5 weeks of viral transfection. A representative brain slice of a TRPV1+ mouse illustrated that TRPV1 expression was primarily confined to the motor cortex ( figure 2(a)). As expected, the contralateral non-injection site did not express any TRPV1 in cortical neurons of the motor cortex. The viral spread of TRPV1 in the cortex was 1.06 ± 0.07 mm 2 , and the density of neurons in the motor cortex that were virally transduced to express TRPV1 was 66.7 ± 4.0 cells mm −2 (figures 2(b) and (c)). The proportion of neurons that expressed TRPV1 in the motor cortex and viral spread region was 5.16 ± 0.43% and 17.21 ± 1.88%, respectively, which was consistent with other reported transduction efficiencies (figures 2(d) and (e)) [32,33]. Higher magnification of the virus transduction region showed that TRPV1 expression was largely confined to neurons, in which 83.4 ± 3.0% of the cells transfected with TRPV1 were neurons (figure 2(f)). These data demonstrate the feasibility of exogenous TRPV1 expression in the motor cortex region and lay the foundation to facilitate TRPV1-mediated sonogenetic control of motor cortex behaviors.

TRPV1-mediated sonogenetic neuromodulation within the motor cortex alters locomotor behavior
We recorded the locomotor behavior of TRPV1− and TRPV1+ mice with the application of FUS at the motor cortex to assess the ability of TRPV1mediated sonogenetics in modulating locomotor behavior. Representative locomotor behavior of TRPV1− and TRPV1+ mice with and without FUS are shown as position traces (figure 3(a); TRPV1-, movie S1; TRPV1+, movie S2). FUS sonication at 0.7 MPa did not evoke considerable motion in TRPV1− mice compared to that before FUS. In TRPV1+ mice, however, FUS stimulation did evoke rotational behavior around the behavior testing arena, which was not observed before FUS. During the FUS sonication periods (shown by the highlighted yellow bars), TRPV1+   stimulation at 0 MPa in TRPV1+ mice (1.1 MPa: 0.36 ± 0.53 rev min −1 ). However, sonogenetics stimulation at 1.1 MPa of TRPV1+ mice achieved a significantly higher angular velocity than that obtained by FUS sonication at 1.1 MPa of TRPV1− mice (p = 0.036). Although not statistically significant, FUS sonication at 1.1 MPa of TRPV1− mice evoked an increase in the average angular velocity in the ipsilateral direction (−0.88 ± 0.36 rev min −1 ) compared with those at 0 MPa and 0.7 MPa. These findings suggest that FUS stimulation at 1.1 MPa alone (without TRPV1) potentially induced neuromodulation effects and generated a confounding impact on TRPV1-mediated sonogenetics at this high-pressure level.

TRPV1-mediated sonogenetics activates cortical neurons on the cellular level
To provide a secondary readout for successful modulation of the motor cortex using sonogenetics, we sacrificed the mice 90 min after the final stimulation and used immunohistochemical staining to analyze c-Fos expression levels of TRPV1+ and TRPV1− mice. Representative fluorescent images of TRPV1− and TRPV1+ brains stimulated at different acoustic pressures demonstrate that sonication at both 0.7 MPa and 1.1 MPa elicited greater c-Fos expression levels in the motor cortex of TRPV1+ mice ( figure 5(a)). Group analysis found that TRPV1+ mice showed enhancement in the number of c-Fos cells at both

Inflammatory and apoptotic responses in the brain are not engaged by TRPV1-mediated sonogenetics
Gross pathology of the mice skull and brain stimulated at 0.7 MPa showed no signs of damage ( figure 6(a)). In contrast, bleeding was consistently observed in the meninges between the skull and the brain at 1.1 MPa. Furthermore, we used Nissl to stain for signs of neuronal damage, GFAP and Iba1 to stain for signs of inflammation, and Caspase-3 to stain for signs of apoptosis ( figure 6(b)). Using the non-injection and non-stimulated side of both TRPV1+ and TRPV1− mice as the control, there were no significant differences in any of the protein marker expression levels in the mouse brain at 0.7 MPa or 1.1 MPa (figure 6(c); one-way repeated measures ANOVA with Dunnett's post-hoc test). Both gross pathology and immunohistological analysis of inflammatory and apoptotic markers showed that TRPV1-mediated sonogenetics at 0.7 MPa enables safe neuromodulation, while damage at the meninges was associated with sonogenetics at 1.1 MPa.

Discussion
Sonogenetics is a rapidly emerging technique that enables noninvasive, cell-type specific neuromodulation with high spatiotemporal resolution. This study demonstrates the capability of TRPV1-mediated sonogenetics to modulate behavior in freely moving mice.
Previous studies have reported sonogeneticenabled neuromodulation in mice by activating mechanosensitive ion channels and proteins, such as prestin [17], MscL G22S [18], and TRPA1 [19], using FUS-induced mechanical effects. They observed neuron activation based on c-Fos staining, as well as motor responses in head-fixed anesthetized mice based on electromyography, but did not report induction of real-time behavior modulation in freely moving mice. Different from these studies, the current study used thermosensitive ion channel TRPV1mediated sonogenetics and achieved successful behavior modulation in freely moving mice. Our previous study demonstrated successful locomotor behavior modulation by TRPV1-mediated sonogenetics in freely moving mice by targeting a deep brain region, the striatum [25]. Here we demonstrated that this technique could modulate locomotor behaviors via a superficial brain target (motor cortex), expanding the application of TRPV1-mediated sonogenetics to include both superficial and deep brain targets.
Previous studies have also reported successful neuromodulation using TRPV1-mediated neuromodulation via combining TRPV1 with different external stimulation modalities. TRPV1-mediated magnetothermal-genetics targeting both superficial and deep brain targets were previously reported [22,23]; however, magnetic nanoparticles need to be injected into the brain to convert energy from an alternating magnetic field to heat for TRPV1 activation. Recently, TRPV1-mediated photothermal genetic stimulation was reported, which combines an injection of nanoparticles with near-infrared light to generate heat for TRPV1 activation [24]. TRPV1mediated magnetothermal and photothermal genetic modulation of the motor cortex induced increases in the angular speed of freely moving mice in the contralateral direction, which was consistent with the results of this study at 0.7 MPa sonication. However, both existing techniques require an additional component of 'energy-converting' nanoparticles that were directly injected into brain tissue. The injection process poses inflammation and ischemia risks, and the presence of nanoparticles in the brain possesses immunogenic and biocompatibility concerns. Since FUS-mediated heating does not require the injection of nanoparticles, it provides a powerful alternative approach to achieving TRPV1-mediated genetic neuromodulation.
We show that TRPV1-mediated sonogenetics successfully evoked motor behavior by targeting the superficial brain target with carefully selected ultrasound parameters. Ultrasound parameters must be selected to achieve successful behavior control without causing any detectable tissue damage. Based on behavior, c-Fos, and safety analyses, TRPV1mediated sonogenetics at 0.7 MPa met this requirement. However, increasing the pressure to 1.1 MPa did not evoke statistically significant changes in angular velocity in TRPV1+ mice relative to the sham sonication at 0 MPa. FUS sonication at 1.1 MPa in TRPV1− mice showed a trend, although not significant, to evoke ipsilateral rotations. Previous work showed that heating alone can suppress neuronal spiking activities, and that heating the circuit involved in locomotion evoked ipsilateral rotational behavior, which can explain the locomotor activity observed in TRPV1− mice at 1.1 MPa stimulation [34]. For TRPV1+ mice, the average angular velocity at 1.1 MPa stimulation was lower compared to that at 0.7 MPa. However, the angular velocity of TRPV1+ mice at 1.1 MPa was statistically higher than that of TRPV1− mice at the same pressure. The data suggest that FUS stimulation at 1.1 MPa alone could impact animal behavior, and that there are potential competing effects between FUS-heating alone and TRPV1-mediated sonogenetics. Future studies will investigate the mechanisms by which FUS-heating and FUS-heating combined with neuronal TRPV1 expression affect neural and circuit level activity, and how these mechanisms can be used to further understand and optimize TRPV1-mediated sonogenetics. It was also interesting to find that damage to the meninges was observed at 1.1 MPa, although no damage to the brain tissue was clearly detected. Damage to the meninges was due to its proximity to the skull. The high skull absorption of ultrasound at 1.1 MPa caused thermal-induced damage to the meninges. Therefore, ultrasound parameter selection must be carefully selected when performing TRPV1mediated sonogenetics to achieve effective and safe neuromodulation.
TRPV1-mediated sonogenetics has great potential for use in large animal models and even humans for clinical applications. Despite this great promise, the current study relies on expressing TRPV1 in the neurons of the motor cortex via intracranial injection, which is invasive and damages healthy brain tissue. There are studies exploring non-invasive approaches for AAV delivery, including FUS-induced Figure 6. TRPV1-mediated sonogenetics at 0.7 MPa did not show signs of inflammation or apoptosis. (a) Representative gross pathology images of the mice skull and brain 90 min after the last FUS sonication at 0, 0.7, and 1.1 MPa. The first column of images shows the intact skull of the perfused mouse, and the second column of images shows the intact brain of the perfused mouse (scale bar = 5 mm). (b) Representative immunofluorescence images of mice brains stained with Nissl dye (cyan), anti-GFAP antibody (yellow), anti-Iba1 antibody (magenta), and anti-Caspase-3 antibody (red) at 0, 0.7, and 1.1 MPa (scale bar = 100 µm). (c) Summary of neuron, astrocyte, microglia, and caspase-3 cell counts in the mice motor cortex after FUS sonication at 0, 0.7, and 1.1 MPa. Data are reported as mean ± SEM. Statistical analysis was conducted with one-way repeated-measures ANOVA with Dunnett's post-hoc test.
blood-brain barrier (BBB) disruption at the targeted brain region [35], FUS-mediated intranasal administration at the targeted brain region [36], or the systemic delivery of AAV variants (AAV-PHP.eB and AAV-PHP.S) that can bypass the BBB and target the central nervous system [37]. Future work will also explore the combination of these minimally invasive delivery approaches with TRPV1-mediated sonogenetics in small and large animal models towards incisionless, spatially targeted, and cell-type specific neuromodulation.

Conclusion
In conclusion, our findings demonstrated the feasibility and safety of using TRPV1-mediated sonogenetics to modulate locomotor behaviors by targeting the motor cortex. Combined with our previous report on TRPV1-mediated sonogenetics for behavior modulation by targeting the deep brain region, our present study indicates that this technique can facilitate neuromodulation at the whole depth of the mouse brain.

Data availability statement
The data that support the findings of this study are available upon reasonable request from the authors.