Poster CC15 - Lipid nanobubbles as delivery vehicles for tolerogenic hybrid insulin peptides in a mouse model of autoimmune diabetes

Introduction: Type 1 diabetes (T1D) is an autoimmune disease where autoreactive immune cells infiltrate pancreatic islets and kill insulin-producing beta cells resulting in severe hyperglycemia if untreated. Numerous beta cell antigens have been implicated in T1D, including T cells reactive to ‘hybrid insulin peptides,’ which are non-germline neo-antigens formed by transpeptidation in the beta cell. One such hybrid insulin peptide identified in non-obese diabetic mice is 2.5HIP, which is a fusion of insulin and chromogranin A peptides. A promising approach to address autoimmunity against hybrid insulin peptides is antigen-specific immunotherapy with tolerogenic peptides, in which exogeneous peptide is administered to promote expansion of anti-inflammatory regulatory T cells. Treatment of mice with nanoparticles containing 2.5HIP has been shown to delay or prevent diabetes onset in an adoptive transfer model of autoimmune diabetes. Here, we develop an image-guided nanoparticle-based delivery platform for 2.5HIP by incorporating the peptide into lipid-shelled ‘nanobubbles.’ Nanobubbles function as submicron ultrasound contrast agents, meaning they can be noninvasively visualized and burst using ultrasound in real time. Prior work has demonstrated that nanobubbles accumulate in pancreatic islets of non-obese diabetic mice as a result of inflammation-associated microvascular permeability. We modified 2.5HIP for incorporation in nanobubbles for delivery of 2.5HIP to islets in an adoptive transfer model of autoimmune diabetes.

Materials and

Methods: Nanobubbles were synthesized by mechanically agitating a lipid precursor solution (developed in-house) with perfluoropropane gas for 45 seconds and isolating the submicron portion (Fig. 1A). To incorporate peptide, 2.5HIP was modified to include an N-terminal palmitoyl tail linked by a cationic glycine-rich linker region (Fig. 1B) and was incorporated with lipids while formulating the lipid precursor solution. For fluorescent imaging of peptide and nanobubble colocalization, palmitoylated 2.5HIP was modified with FITC at the C-terminus and Liss Rhod PE (rhodamine B) was incorporated in the nanobubble lipid shell. To assess peptide function in vitro, splenocytes from a non-obese diabetic mouse (source of antigen presenting cells) were pulsed with 2.5HIP-nanobubbles and co-cultured overnight with CD4+ T cells from a NOD.BDC2.5 TCR mouse. IL-2 production was then measured via ELISA (Fig. 1D). To establish the antigen-specific adoptive transfer model of autoimmune diabetes, cells were isolated from the spleen and lymph nodes of NOD.BDC2.5 TCR mice, sorted for CD3+ cells, cultured for 3 days with CD3/CD28 stimulating beads and cytokines, and transferred into NOD.SCID mice via retroorbital injection. Adoptive transfer mice were then treated daily for 4 days with 2.5HIP-nanobubbles, administered via tail vein injection (Fig. 1F). For assessment of time-to-diabetes, blood glucose was monitored with a Contour Next glucometer until >250 mg/dL for 2 consecutive measurements. For assessment of islet-resident regulatory T cells, islets were isolated, dissociated, stained, and analyzed via flow cytometry. Cells were fixed, permeabilized, and stained for Foxp3 using a transcription factor staining kit (eBioscience).

Results, Conclusions, and Discussions: Peptide palmitoylation resulted in stable encapsulation of 2.5HIP in the nanobubble shell, with peptide and nanobubbles being well-colocalized (14% colocalization) (Fig. 1C). Inclusion of a six-residue cationic glycine-rich linker region was also necessary for stable encapsulation, with an 83% decrease in colocalization without it (p < 0.05). Palmitoylation and incorporation in nanobubbles did not substantially diminish ability of 2.5HIP to stimulate BDC2.5 T cell receptors, with significantly greater IL-2 being produced by BDC2.5 T cells in vitro than control NBs (p < 0.05) (Fig. 1E). T cells were also analyzed with flow cytometry and no differences were observed in the proportion of CD4+ T cells that were 2.5HIP-reactive, indicating that elevated IL-2 production resulted from T cell receptor stimulation rather than an elevated presence of 2.5HIP-reactive cells. These results indicate that 2.5HIP can be incorporated in nanobubbles and doing so does not affect peptide function. We then activated NOD.BDC2.5 TCR T cells in culture and transferred them into NOD.SCID mice. Activation of T cells was confirmed by observing changes in cell size and morphology and measuring cell expansion, with CD3+ cell count being 1.95 ± 0.17 times higher following activation. The cells were confirmed as being diabetogenic by measuring blood glucose daily in mice that had received an adoptive transfer but no treatment; all of these mice became diabetic within one week. Experimental mice were treated with 2.5HIP-nanobubbles, HEL-nanobubbles (irrelevant peptide), or 2.5 HIP alone. 2.5HIP-nanobubble treated mice had a significantly longer time-to-diabetes than other groups, with diabetes being delayed to at least 12 weeks-post-transfer in each mouse (Fig. 1G). Using survival analysis, time-to-diabetes was significantly greater for 2.5HIP-nanobubble-treated mice compared to 2.5HIP-treated mice (p < 0.0001) and HEL-nanobubble-treated mice (p < 0.0001). Additionally, there was a trend observed toward a higher proportion of islet-resident regulatory T cells with 2.5HIP-nanobubble treatment compared to controls (p = 0.0946) (Fig. 1H). These results indicate that nanobubbles can be applied as a highly effective vehicle for tolerance induction in the context of autoimmunity against hybrid insulin peptides and have the potential to improve preventative therapy in type 1 diabetes.

Acknowledgements (Optional): This project was supported through funding by the National Institutes of Health NIDDK (T32 DK120520, F31 DK136186, R01 DK102950, R01 DK106412, P30 DK116073). The authors would like to thank the University of Colorado Denver College of Engineering, Design, and Computing and Department of Bioengineering, the University of Colorado Anschutz Department of Immunology and Microbiology, and the Barbara Davis Center for Diabetes at the Anschutz Medical Campus. We thank the NIH Tetramer Core Facility at Emory University for providing MHC tetramers. We would like to acknowledge Biorender for the creation of schematics.