Introduction:: Ultrasound imaging and ultrasound-mediated gene/drug delivery are minimally invasive diagnostic and therapeutic methods. However, the use of microbubbles and nanobubbles is limited by their sizes, which prevents them from crossing biological barriers and reaching deep tissue cells. To address this, we introduce a novel sub-50 nm gas-filled protein nanostructures, referred to as S50GVs, derived from genetically modified gas vesicles. These diamond-shaped nanostructures have smaller diameters than commercially available 50-nm gold nanoparticles and can efficiently traverse biological barriers. S50GVs can be produced in bacteria, purified through simple centrifugation, and remain stable for months. We have shown that S50GVs can extravasate lymph drainage into lymphatic tissues and access critical immune cells. In a parallel effort, we aim to equip our acoustic protein nanostructures with the targeting moieties. To this end, targeting GVs to specific tissues has proven challenging, primarily due to the lack of effective methods to modify the targeting moiety on the GV surface while maintaining a homogenous preparation. This obstacle is a significant hindrance to engineering these nanoparticles for targeted theragnostic applications. In this presented work, we create a dual-functional acoustic protein nanostructures by conjugating therapeutic antibodies via a so-called p-Click method using specifically designed non-canonical amino acid (ncAA)-containing antibody-binding peptide. In summary, our work addresses two key challenges, the size and the targeting capability, of the biogenic gas-filled protein nanostructures for its ultrasound theranostic applications.
Materials and Methods:: First, we engineered new genetic variants of GVs that have hydrodynamic diameter smaller than the commercially available 50-nm gold nanoparticle. These ultrasmall GVs are formed by a novel genetic sequence of the major shell protein, which renders them into diamond shape (thus much shorter length) and smaller diameter. We used dynamic light scattering and electron microscopy to characterize their size, shape and surface charge. We then conducted rodent animal experiments to assess their biodistribution into lymph node, asking the question of whether they extravasate out of the lymph drainage and penetrated through the lymphatic endothelial cell barrier, which is known to exclude nanoparticles above 100 nm in diameter. Furthermore, we have characterized their sub-cellular localization using thin-sectioned electron microscopy and quantified their ultrasound imaging contrast. Secondly, for the antibody-targeted gas vesicles, we employed a proximity-induced antibody conjugation strategy, which allows for the precise covalent bonding between functional moieties and native antibodies without the need for antibody engineering or any supplementary chemical treatments. We used this method to conjugate anit-HER2 antibodies to gas vesicles, and in order to assess the efficacy of our modified anti-HER2 GVs, we conducted in vitro and in vivo ultrasound imaging experiments to access the tumor targeting of anti-HER2 GVs.
Results, Conclusions, and Discussions:: For ultrasmall GVs, the hydrodynamic radius of these GVs was measured by dynamic light scattering to be 57.88 ± 11.89 nm, smaller than 72.9 ± 17.3 nm measured on commercially available 50-nm gold nanoparticles (Fig. 1, top panels). Next, tissue histology of the lymph nodes from rodent models revealed these ultra-small GVs can extravasate out of the lymph drainage into the lymphatic tissues, which cannot be achieved by the normal wildtype GVs (hydrodynamic radius ~ 200 nm). Lastly, we are the in the progress to append targeting moieties to these GVs to enable their targeting of specific cell types in the lymphatic tissue.
To produce targeted gas vesicle (GVs), we used the proximity-induced site-specific method to conjugate the reactive azide group on the fragment crystallizable (Fc) part of the anti-Her2 antibody3. To achieve the site-selective formation of a covalent bond between adjacent native amino acids, we introduced a proximal lysine residue into a specific site of a well-characterized Fc targeting peptide. Upon binding to the antibody, the cross-linking non-canonical amino acid (ncAA)-containing peptide will enable proximity-induced covalent attachment of the cross-linking ncAA to the nearby lysine residue of the antibody. We chose a clinically used therapeutic antibody, an anti-Her2 antibody, to demonstrate this process. Next, we evaluated the in vitro targeting ability of HER2_GVs by comparing its targeting efficiency to HER2+ cells and HER2- cells. We used confocal microscopy to visualize the binding of the HER2_GVs to the HER2+ cancer cells, and we found that the modified GVs showed a significantly higher binding affinity to HER2+ cancer cells compared to the HER2- cancer cells. Furthermore, we assessed the acoustic signal of HER2_GVs under ultrasound (Fig1. c). We used a novel ‘BURST’ ultrasound sequence to generate images of the modified GVs, and we found that the HER2_GVs had the unique acoustic signal under ‘BURST’ sequence. Finally, we demonstrated HER2_GVs enabled highly specific detection of HER2+ cancer cells in vivo (Fig1. d).
