Enhanced Noninvasive Imaging of Acoustic Biomolecules

The extensive scientific interest in cellular and biomolecular processes is due in large part to the importance of such processes deep inside living organisms, in the context of both health and disease. However, most methods for imaging cellular processes such as gene expression have relied on fluorescent proteins and other optical reporters that, while providing a direct optical readout of the biomolecular environment in cells readily exposed to light, have greatly limited performance in large animals due to the poor penetration of visible light beyond 1 mm of biological tissue. In contrast, ultrasound is widely used to noninvasively image tissue deep inside living organisms but has rarely been used to investigate cellular function due a lack of acoustic reporters whose production and properties are coupled to biomolecular events. Recently, the first acoustic reporter genes (ARGs) were developed for ultrasound imaging of a unique class of air-filled protein nanostructures known as gas vesicles, or GVs, which scatter sound waves when expressed in bacterial and mammalian cells. ARGs allow gene expression to be visualized with ultrasound similar to how green fluorescent protein (GFP) allowed gene expression to be visualized with light. However, ARGs will have limited utility in practical applications involving living organisms without ultrasound imaging methods providing the specificity to reliably distinguish GVs from surrounding tissue and the sensitivity to detect GVs at low concentrations. In this thesis, we present two novel ultrasound imaging methods that exploit the unique nonlinear physical properties of gas vesicles to enhance image quality in situations that pose challenges for conventional imaging methods. In Chapter 1, we address the problem of distinguishing GVs from tissue with cross-Amplitude Modulation (xAM), an ultrasound pulse sequence that uses X-waves to isolate the signal generated by reversible buckling of the GV shell while cancelling scattering and artifacts from tissue. In Chapter 2, we present an application of xAM to imaging of dynamic biomolecular processes. We show that, when GVs are engineered such that buckling is induced by enzyme activity, xAM can visualize enzymatic processes deep inside living animals. In Chapter 3, we address the problem of detecting very low concentrations of ARG-expressing cells with Burst Ultrasound Reconstructed with Signal Templates (BURST), an imaging method that exploits the strong, transient signals generated during sudden GV collapse under acoustic pressure by unmixing the temporal dynamics of such signals from background scattering. BURST imaging improves cellular sensitivity by more than 1000-fold and, in dilute cell suspensions, enables the detection of gene expression in individual bacteria and mammalian cells. In Chapter 4, we present an application of an early formulation of BURST to imaging gene expression in mammalian cells. We use this imaging method to visualize vascularization patterns in tumors containing mammalian cells expressing acoustic reporter genes.