Sonogenetics Offers Promise to Startup Neurotech Firms

by Sharena Rice, contributing editor

August 2025 issue

In recent years, neuroengineers have devised a number of new modalities for interfacing with the nervous system. Among these are optical stimulation, vibrational stimulation, and optogenetics. A newer and perhaps more promising technology is sonogenetics.

Sonogenetics, the use of focused ultrasound to control cells that have been made ultrasound-responsive via gene delivery, is moving from compelling papers to a potential platform strategy. From a neurotech commercialization standpoint, the significance of sonogenetics is less about a single lab trick and more about the emerging convergence of three capabilities: precise genetic targeting, durable and safe delivery, and field-robust ultrasound systems that work the first time outside the origin lab.

One commercial firm that may be exploiting this technology is Merge Labs. The startup recently made a big splash with a $250 million investment from Open AI and Sam Altman. While the company has not yet released its website and the technical personnel behind the company have not been identified, it is rumored to be working with focused ultrasound implants and sonogenetics as gene therapy. If Merge and its peers can validate durable expression, predictable dose–response, and reliable outside-the-lab bring-up, a first wave of indications will likely sit at the intersection of neurology, psychiatry, and rehabilitation, with longer-term spillover into human-machine interaction.

Sound is a mechanical wave. Compared with optogenetics, sonogenetics trades light-sensitive channels for mechanosensitive channels and implants in brain tissue for transducers external to brain tissue. The resulting value proposition is clear for deep brain regions and repeat dosing: no cranial leads, no optical fibers, and compatibility with standard imaging and clinical workflows. Still, like the ultrasound transducers themselves, the expectations should be calibrated.

Sonogenetics involves targeting neurons based on which molecular markers they have. While all neurons have molecular markers associated with being neurons, there are different types of neurons with their own specialized molecular signatures. This does not mean that everything is one-and done: off-target expression is also a consideration in the development of this genetic-based technique. As synthetic biologist Logan Thrasher Collins (not affiliated with Merge Labs) stated, temporal resolution in sonogenetics “depends on the kinetics of the mechanosensitive ion channels chosen,” and “overall, most optogenetic channels have faster kinetics than sonogenetic channels.” Collins adds that spatial resolution is largely dictated by the system delivering the genes to the neurons in the region of interest, so “spatial resolution should be similar” across sonogenetics and optogenetics when comparable delivery is used.

Durability is the next test. Corporate diligence teams should ask how long expression lasts under realistic use and what re-dosing looks like if immunity rises. Collins noted that in neurons, while non-AAV viral approaches may persist around a year and nonviral methods are usually much shorter-lived unless genome editing is introduced, AAV expression “lasts for years.” The major unknown, he added, is whether the field can engineer a nonviral vector which persists in expression the same way as AAVs. Those answers will not only shape risk and safety narratives, but also manufacturability, pricing, and payer logic.

Safety and reversibility remain central to any neurotech system entering clinical pathways. Collins emphasized that if CRISPR or similar tools are not used, many vectors do not modify the genome and instead persist as episomes for years in the case of AAVs, with integration events considered rare. He also noted that neuromodulatory FUS itself “probably would not strongly affect the viral vector or gene expression,” though other FUS paradigms can be used intentionally to open the blood-brain barrier for delivery. For companies developing sonogenetics, that translates into a clearer reversibility story for some programs and a credible path to chronic use, paired with the obligation to monitor immunogenicity and to design re-dosing strategies upfront.

For teams interested in emerging research in potential neurotherapeutic tools, the takeaway is straightforward. Sonogenetics is no longer just a mechanistic curiosity; it is a systems-integration race. The technique holds both promise and complex practicalities: channel kinetics constrain temporal precision, delivery governs who gets hit, expression durability could define patient experience over time, and safety narratives hinge on episomal persistence and immunology rather than blanket “gene editing” fears. The companies that align molecular targeting, dependable field usability, and the evidence pipelines to substantiate claims, will set the pace as this interface class moves from preclinical promise to first-in-human studies to real-world products.