Charting the Chemical and Mechanistic Scope of Light-Triggered Protein Ligation

Abstract

The creation of discrete, covalent bonds between a protein and a functional molecule like a drug, fluorophore, or radiolabeled complex is essential for making state-of-the-art tools that find applications in basic science and clinical medicine. Photochemistry offers a unique set of reactive groups that hold potential for the synthesis of protein conjugates. Previous studies have demonstrated that photoactivatable desferrioxamine B (DFO) derivatives featuring a para-substituted aryl azide ($ArN_3$) can be used to produce viable zirconium-89-radiolabeled monoclonal antibodies ($^{89}Zr-mAbs$) for applications in noninvasive diagnostic positron emission tomography (PET) imaging of cancers. Here, we report on the synthesis, $^{89}Zr$-radiochemistry, and light-triggered photoradiosynthesis of $^{89}Zr$-labeled human serum albumin (HSA) using a series of 14 different photoactivatable DFO derivatives. The photoactive groups explore a range of substituted, and isomeric $ArN_3$ reagents, as well as derivatives of benzophenone, a para-substituted trifluoromethyl phenyl diazirine, and a tetrazole species. For the compounds studied, efficient photochemical activation occurs inside the UVA-to-visible region of the electromagnetic spectrum (∼365–450 nm) and the photochemical reactions with HSA in water were complete within 15 min under ambient conditions. Under standardized experimental conditions, photoradiosynthesis with compounds 1–14 produced the corresponding $^{89}ZrDFO-PEG_{3}-HSA$ conjugates with decay-corrected isolated radiochemical yields between 18.1 ± 1.8% and 62.3 ± 3.6%. Extensive density functional theory (DFT) calculations were used to explore the reaction mechanisms and chemoselectivity of the light-induced bimolecular conjugation of compounds 1–14 to protein. The photoactivatable DFO-derivatives operate by at least five distinct mechanisms, each producing a different type of bioconjugate bond. Overall, the experimental and computational work presented here confirms that photochemistry is a viable option for making diverse, functionalized protein conjugates.