Responses to stimuli is a key characteristic of life. In nature, various organisms are capable of sensing distinct environmental inputs such as light, heat, chemicals, and magnetic field and therefore can react to them accordingly. Discovery of key proteins or molecular machinery required for the organisms’ sensing of various stimuli has benefited a lot both to basic science and to translational applications. It provides molecular mechanisms with how organisms sense and interact with the environment^1–6. Moreover, heterogeneously expressing of stimulus-responsive proteins endows targeted cells with the ability to sense the stimuli, allowing researchers to remotely control cellular activities with chemicals, light, or magnetic field in vitro and in vivo^1,2,5. Although these tools are versatile, several drawbacks exist regarding either slow on/off-rates or poor penetration to dense and deep tissues. To overcome such technical limitations and also extend the toolkit, the goal of our study is the identification of proteins that sense ultrasound, a focused and non-invasive stimulus with good depth of penetration.

To explore the genetically-encoded ultrasound-responsive protein from mother nature, we have focused on the membrane protein Prestin, which resides in the mammalian auditory systems and is important for high-frequency hearing. Inspired by the assumption that several parallel amino acid substitutions of Prestin may be involved in adaptive ultrasound hearing, we introduced two evolution-based mutants N7T and N308S into Prestin protein of non-echolocating mice. Heterogeneous expression of this construct (hereafter mPrestin(N7T, N308S)) allowed the mammalian cells to sense ultrasound stimulation, which evoked a calcium influx from the extracellular space into their cytosol under a low-frequency and low-pressure ultrasound condition. To the best of our knowledge, this is the first case that uses low-frequency ultrasound to control genetically-modified mammalian cells. Owing to great penetration of low-frequency ultrasound (~400 mm in depth), our sonogenetic system is applicable to deep tissues in large animals like primates. This success was made possible by elaborating a genetically-encoded ultrasound-responsive protein mPrestin(N7T, N308S) (Huang et al., Manuscript submitted; Patents are pending).

The echolocating ability of bats was found by Dr. Robert Galambos in the 1940s. However, thus far, the molecular mechanisms of how echolocating species sense ultrasound are still unclear. In our recent study, we identified two parallel amino acid substitutions of Prestin that frequently occurred in sonar species but not in their non-sonar counterparts. Using cell culture as an experimental model, introduction of N7T and N308S into mouse Prestin significantly improved the ultrasound sensitivity of mammalian cells. N7T and N308S enhanced self-association of Prestin in the punctate regions where they highly associate with cytoskeletons. A short pulse of 0.5 MHz ultrasound robustly induced oscillation of mPrestin(N7T, N308S) puncta and evoked calcium responses in cells. Our discovery proposes the first molecular mechanism of how an engineered auditory-sensing protein assists mammalian cells to sense ultrasound stimulation. Besides serving an experimental model for investigating how cells sense ultrasound stimulation, our sonogenetic system will offer far-reaching implications for non-invasive therapy in large animals like primates owing to great penetrability of low-frequency ultrasound (~400 mm in depth).