Engineering Acoustic Protein Nanostructures for Non-Invasive Molecular Imaging using Ultrasound

• Main Author: Lakshmanan, Anupama
• Published: 2019
• Online Access: Lakshmanan, Anupama (2019) Engineering Acoustic Protein Nanostructures for Non-Invasive Molecular Imaging using Ultrasound. Dissertation (Ph.D.), California Institute of Technology. doi:10.7907/ASX5-KB62. https://resolver.caltech.edu/CaltechTHESIS:06062019-165907194 <https://resolver.caltech.edu/CaltechTHESIS:06062019-165907194>

<p> Visualizing biomolecular and cellular processes in real time within deep tissues is fundamental to our understanding of the normal and pathological activity underlying health and disease. Ultrasound provides the ability to non-invasively image deep inside biological tissues with high spatial and temporal resolution. However, this technology has limited capacity to monitor molecular and cellular processes, due to the lack of appropriate intra-cellular and endogenously producible nanoscale contrast agents, which can directly couple sound waves to the activity or concentration of physiologically relevant molecules. This problem could in principle be solved by developing genetically encodable ultrasound sensors – biomolecules that can get illuminated in ultrasound imaging in response to specific cellular or molecular activity. This thesis describes the engineering and characterization of acoustic protein nanostructures called 'gas vesicles', or 'GVs', to accomplish this task. </p> <p> GVs are protein-shelled gas-filled nanostructures produced by buoyant microbes, and were recently shown to be capable of scattering sound waves to produce ultrasound contrast. Owing to this property, they were initially conceptualized as a new class of ultrasound contrast agents. However, little was known about their tunability to enable molecular ultrasound imaging for a wide range of applications. In this thesis, we leveraged the genetic encodability of GVs to modify them at the level of their DNA sequence and constituent proteins, and thereby tune their mechanical, acoustic, surface and targeting properties. We accomplished this by establishing a facile and modular molecular engineering platform, to produce GVs that provide enhanced nonlinear signals for sensitive and specific detection in deep tissues, target specific cell types such as cancer and immune cells, and also provide distinct acoustic collapse spectra for multiplexed imaging. We then extended this platform to build GV-based biosensors that modulate their nonlinear ultrasound signals in response to changes in the activity or concentration of specific molecules in their environment. Specifically, we engineered acoustic sensors for three different types of enzymes and for calcium – whose activity or flux underlie a wide range of important cellular processes. Furthermore, we succeeded in transferring the genetic code of gas vesicles from their species of origin into a variety of other microbes that do not naturally produce them, in order to unlock their potential as ultrasound reporter genes. Our results establish GVs as reliable acoustic biomolecules, and thereby extend the capabilities of ultrasound for molecular and cellular imaging in a manner analogous to green fluorescent protein (GFP) and its derivatives in optical microscopy. When combined with the advantages of ultrasound for non-invasive imaging, this work facilitates novel technology to significantly enhance our understanding of molecular and cellular processes in basic biology, as well as enable improved diagnosis, monitoring and treatment of diseases. </p>