Presentations

Abstract

Introduction:

Using Positron Emission Tomography (PET) to image bacterial infection is an emerging field that has celebrated several recent successes. Among them is carbon-11 labeled D-amino acids, but their 20 minute half-lives are challenging to work with, thus a major goal is to evolve this into fluorine-18 amino acid derivates. Due to the costly, inefficient serial design and radiosynthesis of various D-amino acid derived PET tracers, we sought a method to rapidly surveil tolerable changes to the D-amino acid scaffold, including variable side-chain structure and amino acid C-terminal modification. Using a biorthogonal chemistry approach, we delivered an exogenously derived D-amino acid bearing a side-chain azide for incorporation into bacterial peptidoglycan and subsequently ligated a detectable strained cyclooctyne via strained-promoted azide-alkyne cycloaddition (SPAAC). We developed a water-soluble and fluorine-18 labelled strained cyclooctyne ([18F]FB-sulfo-DBCO) that could be used to generate pathogen-specific PET signals. We believe that this strategy could be used to develop a "pre-targeting" infection imaging method for which a patient-delivered D-amino acid would be followed by a PET tracer used to detect it with the advantage of sensing only peptidoglycan-incorporated D-amino acids. Methods: Radiotracer synthesis [18F]SFB was prepared using an automated published method and reacted with the commercially available sulfo-DBCO precursor in the presence of DMF and base. [18F]FB-Sulfo-DBCO was purified and isolated using HPLC purification. Identity was confirmed by comparison to the cold standard, which was synthesized using analogous methods. In Vitro Uptake Assay Azide metabolites (final conc. 5mM) were added to 1% dilution of overnight cultures of Staphylococcus aureus, Listeria monocytogenes, Pseudomonas aeruginosa, Acinetobacter baumannii, and Escherichia coli and incubated for 1 hour. As biological controls, bacterial cultures were grown under similar conditions and treated with D-alanine following the aforementioned protocol. The cultures were pelleted and washed with washing buffer. 50µCi of [18F]FB- Sulfo-DBCO was added and reacted for 1 hour. The pellets were separated and washed with buffer. The pellets and filtrates were individually analyzed on a gamma counter. SPAAC ligation is reported as % of signal normalized by measured optical densities. Dynamic PET/CT imaging in healthy mice Healthy CBA/J mice (female, 9-11 weeks old, 20-24 g) were used for all experiments. 90-minute whole-body dynamic μPET acquisitions of animals (N = 4) were obtained with 34 frames with [18F]-Sulfo-DBCO (200 ± 21 MBq, 100 μL) via tail vein injection, followed by a 10 minute μCT scan. Region of interest (ROI) analysis was used to evaluate organ-specific tracer clearance at early time points, and ex vivo analysis (N = 5) at 90 minutes used to assess tracer retention in organs via harvesting and gamma counting.

Results:

In this study, we created a high-throughput assay to assess [18F]FB-sulfo-DBCO labeling of several azido-D amino acid metabolites in an array of pathogens and its biodistribution. From selected azido compounds, D-azido alanine presented the highest accumulation (5 fold to L-azido alanine) into the exposed surface of S. aureus. Increasing methylene moieties on the amino acid side chain negatively impacted the incorporation or presentation of the azide on the cell wall, although overall modification of the side chain length was tolerated. The incorporation of D-azido alanine is highly concentration dependent (65% signal reduction observed across 5 to 0.156mM range) and currently no discernable trend is observed with azide ligation in gram-positive and gram-negative pathogens. [18F]FB-sulfo-DBCO accumulated in the liver and was excreted through the gallbladder, spleen, and gastrointestinal system, suggesting a dominant biliary excretion pathway with excretion through the bladder over time of imaging.

Conclusion:

The new PET tracer, [18F]FB-sulfo-DBCO, readily labels azide-modified bacteria. Significant differences in labeling have been observed among the different metabolites and pathogens. The presented data demonstrates the utility of our developed high-throughput method and its promise to being a valuable tool for us in understanding the underlying biology of these metabolic pathways and in the development of new imaging probes for bacterial infection.