Sonogenetics for noninvasive and cellular-level neuromodulation in rodent brain

Abstract

Sonogenetics, which uses ultrasound to noninvasively control cells genetically modified with ultrasound-sensitive ion channels, can be a powerful tool for investigating intact brain circuits.Here we show that TRPV1 is an ultrasound-sensitive ion channel that can modify the activity of TRPV1-expressing cells in vitro when exposing to focused ultrasound. We also show that focused ultrasound exposure at the mouse brain in vivo can selectively activate neurons that are genetically modified to express TRPV1. We demonstrate that precise manipulation of neural activity via TRPV1-based sonogenetics can be achieved by spatiotemporal control of ultrasoundinduced heating. The focused ultrasound exposure is safe based on our inspection of neuronal integrity, apoptosis, and inflammation markers. This sonogenetic tool enables noninvasive, celltype specific, spatiotemporally controlled modulation of mammalian brain activity.

Sonogenetic stimulation of the brain at a spatiotemporal resolution suitable for vision restoration

“…Unfortunately, existing US neuromodulation strategies are restricted to low-frequency transmissions, resulting in poor spatial resolution (>3 mm) and long-lasting responses. Indeed, attempts at high-frequency neuromodulation have resulted in high levels of acoustic energy 26 , with a risk of thermal heating [8][9][10] and USmediated tissue damage 6,11 . This inevitable trade-off between spatial resolution and acoustic intensity has greatly limited the applicability of US neuromodulation for BMI.…”

“…The ultimate goal is the development of a technology that is less invasive than electrodes but capable of activating neurons in the visual cortex with a high spatial (~100 µm) and temporal (<50 ms) resolution. We propose 1) to boost neuronal sensitivity to US through the expression of USsensitive channels on cell membranes [8][9][10]27,28 , 2) to demonstrate that it is possible to target a locally defined subset of neurons by gene therapy, which is not currently possible with nonspecific US neuromodulation strategies, 3) to induce responses with a sufficiently high temporal precision and 4) to gain more than one order of magnitude in spatial resolution through the use of high-frequency US, which was previously considered impossible in vivo without considerably increasing acoustic intensities and potential adverse effects 29 . This activation of the brain with a unique combination of high spatial and temporal resolution, through the intact dura, could render sonogenetic therapy perfectly compatible with applications for vision restoration, which require video-rate patterns of stimulation.…”

“…US stimulation can overcome this limitation 2,[44][45][46] , but the optimal conditions for sonogenetic approaches, in terms of both genetic constructions and acoustic parameters, are far from clear. Most previous sonogenetic studies focused on the use of low-frequency US [8][9][10] as, in particular, in the recent demonstration of MscL-based sonogenetic activation in mice brain with long bursts of lowfrequency US 10 . However, such low-frequency US waves lead to limited centimetric spatial resolutions (typically around 5x5x45 mm 3 ) and an uncontrolled spatial beam distribution, due to the formation of standing waves 47 .…”

Vision Restoration by Optogenetic Therapy and Developments Toward Sonogenetic Therapy

“… 82 To reduce these unwanted effects, various proteins have been expressed ectopically in neurons to render them ultrasound sensitive. These ultrasound sensitive proteins include the auditory-sensing Prestin protein, 83 , 84 the TRP-4 ionic channel, 85 the TRPV1 ionic channel, 86 and the mechanosensitive large conductance ionic channel MscL. 87 – 89 All the in vivo studies using these proteins showed temporal resolution requiring several hundred milliseconds to a second for developing the response, which is too slow for visual restoration.…”

“… 87 – 89 All the in vivo studies using these proteins showed temporal resolution requiring several hundred milliseconds to a second for developing the response, which is too slow for visual restoration. 83 – 86 , 89 We therefore investigated if a mechanosensitive protein can generate a sonogenetic activation with a spatiotemporal resolution compatible with visual restoration. We have shown that the G22S mutated version of the MscL ionic channel displaying an increased sensitivity to ultrasound stimulation can be reliably expressed in neurons of the visual cortex providing a millisecond temporal resolution and a spatial resolution in the 100 µm range when applying 15 MHz ultrasound waves at safe energies.…”