Biomolecular Engineering of Gas Vesicles with Thiol Functionality

Citation

Schrunk, Erik Tao (2025) Biomolecular Engineering of Gas Vesicles with Thiol Functionality. Dissertation (Ph.D.), California Institute of Technology. doi:10.7907/49mr-3744. https://resolver.caltech.edu/CaltechTHESIS:05292025-004444146

Abstract

Therapies involving the administration of engineered cells, such as CAR-T cell therapy and the delivery of genetically modified gut microbes, have enjoyed clinical success and increasing interest in recent years. While these therapies continue to show great promise, the opacity of tissue precludes the use of light in the observation and potential manipulation of these engineered cells as they carry out their functions within the body. To access these cells non-invasively in deep tissue requires the use of imaging modalities that do not involve light, of which ultrasound (US) is especially appealing due to its relatively low cost, safety, and widespread availability. Engineered cells can exhibit US contrast by expressing gas vesicles (GVs), air-filled polymeric proteinaceous nanostructures; GVs have already been used as acoustic reporters for gut colonization, tumor cell activity, and more.

Whereas GVs are most notable for their acoustic properties, we set out to further expand the function of GVs by chemically modifying them at the genetic level. Our goals were twofold: we wished to equip the external, solution-facing side of GVs with a unique chemical handle; and we wished to hide a reactive group within the internal, air-facing side of GVs that could only be revealed when the GV structures are irreversibly collapsed. To accomplish both these goals, we chose to incorporate cysteine into the shell of GVs because cysteine’s thiol side chain is chemically unique among all natural amino acids and because wild-type GVs do not contain cysteine in their shells. We set up a cysteine scanning mutant library of the GV shell protein, GvpA/GvpA1, and screened for cysteine-tolerant mutations in the gene. Through this process, we discovered cysteine substitutions that furnished thiol groups facing both the GV exterior and interior.

The GV-exterior-facing cysteines were leveraged to develop a modified GvpA that contains the bioorthogonal six amino acid tetracysteine tag, or TC tag. The TC tag reacts with the membrane-permeable molecule FlAsH, which becomes fluorescent upon reaction. We used TC-tagged GvpA, or tcGvpA, to express GVs in HEK 293T cells, and used confocal microscopy of FlAsH to study those GVs. Notably, we only substituted a small percentage of GvpA to tcGvpA, leaving the rest of the GvpA as wild type; to our knowledge, this is the only report of a polymeric proteinaceous structure that employs this chimeric assembly approach being successfully expressed and labeled with FlAsH. The microscopy results from this study were used to generate three-dimensional renderings that provided insights into the size and positioning of GV clusters expressed within HEK 293T cells.

Second, we identified several interior-facing cysteine mutants to the GV shell protein GvpA1, which we used to develop “SonoCages”: chemical entities whose reactivity is gated by US. We purified GVs with one mutation from our screen, V47C, and reacted them with monobromobimane (mBBr), a fluorogenic, thiol-reactive molecule. The mutant GVs only reacted with mBBr after treatment with US, which collapsed the GVs and exposed their hydrophobic interiors to the bulk solution. Thus, we had developed thiol-bearing SonoCages whose cysteines could only engage in reactions after US-mediated collapse of the GVs—a process we call “sono-uncaging” in analogy to photo-uncaging. We further demonstrated the utility of SonoCages by preparing a hydrogel containing SonoCages and mBBr and using US to create fluorescent patterns corresponding to regions of GV collapse.

The work presented in this thesis not only demonstrates the functionalization of the GV interior and exterior, but also establishes a framework through which further modifications can be performed. Whereas we used cysteine as our reactive chemical of choice, other amino acids (including non-canonical amino acids) could be used to explore a much wider library of reactivities. The vast potential of GV chemical modification, along with the amazing results from the rest of the Shapiro Lab and in labs across the world, serves as a reminder that GVs and GV-based technologies are not just a bubble (pun intended)—they are going to be around for a long time.

Thesis Files