Despite their inherent potential, the use of [18F]-protein radiotracers in both preclinical and clinical settings is hampered by a lack of robust radiolabeling techniques.
We aim to address this by developing a fully automated radiolabeling methodology that is site-specific, occurs rapidly under mild aqueous conditions, and requires only small amounts of precursor peptide for high radiolabeling yields. The most frequent protein radiofluorination approach, conjugation of N-succinimidyl 4-[18F]-fluorobenzoate to lysine residues, is limited by a lengthy multi-step synthesis, low yields, poor control over labeling site (which can lead to reduced immunoreactivity), and the frequent requirement for large amounts of protein precursor. Other prosthetic groups have been reported, however to date none address all of these issues. The lack of an optimal, broadly applicable, radiolabeling strategy for proteins motivated us to investigate enzymes as radiolabeling catalysts. Preliminary, proof-of-concept data shows that the enzyme lipoic acid ligase can site-specifically ligate a [18F]- prosthetic to a model protein tagged with a 13-amino acid acceptor sequence (`LAP-tag'). The reaction was high yielding at neutral pH and ambient temperature, ensuring full retention of the protein's biological activity. In this proposal, we will build on these studies by developing a second generation prosthetic with a streamlined radiosynthesis and improved metabolic stability. Crucially, all radiosynthetic steps will be performed on an automated radiosynthesizer to enable reproducibility and to facilitate the immediate transfer of labeling protocols between different sites.
In Specific Aim 1 we will synthesize the second generation aryl [18F]-fluoride prosthetic, confirm its ligation to a model protein-LAP construct, and purify the resulting radiotracer.
In Specific Aim 2, we will transfer this protocol to UCSF where the radiolabeling will be reproduced on their radiosynthesizer. Finally, the radiotracer will be characterized for purity, specific activity, retention of biological activity, and metabolic stability. Upon completion of these studies we anticipate having a fully optimized radiofluorination protocol in hand along with proof-of-concept data for our model protein. In Phase II of this project, we will apply our methodology to several different classes of protein, including affibodies, nanobodies, diabodies, and Fab antibody fragments, with the goal of defining its full scope and benchmarking its performance against the current gold standard in the field in anticipation of producing a fully automated, commercial available, kit-based approach for protein radiofluorination.
The translation of protein-based radiotracers from the lab to the clinic for use in cancer diagnostics is currently hampered by a lack of suitable radiolabeling methodologies. This project will use enzymes to catalyze protein radiolabeling on a commercially available automated radiosynthesizer to facilitate widespread adoption of the technique. If successful, this approach will be highly efficient, require smaller amounts of protein precursor compared to other techniques, enable site-specific labeling, and will take place under mild, aqueous conditions.
