Development of caspase-3-selective activity-based probes for PET imaging of apoptosis

Background

The cysteine-aspartic acid protease caspase-3 is recognized as the main executioner of apoptosis in cells responding to specific extrinsic and intrinsic stimuli. Caspase-3 represents an interesting biomarker to evaluate treatment response, as many cancer therapies exert their effect by inducing tumour cell death. Previously developed caspase-3 PET tracers were unable to reach routine clinical use due to low tumour uptake or lack of target selectivity, which are two important requirements for effective treatment response evaluation in cancer patients. Therefore, the goal of this study was to develop and preclinically evaluate novel caspase-3-selective activity-based probes (ABPs) for apoptosis imaging.

Results

A library of caspase-3-selective ABPs was developed for tumour apoptosis detection. In a first attempt, the inhibitor Ac-DW3-KE (Ac-3Pal-Asp-βhLeu-Phe-Asp-KE) was ¹⁸F-labelled on the N-terminus to generate a radiotracer that was incapable of adequately detecting an increase in apoptosis in vivo. The inability to effectively detect active caspase-3 in vivo was likely attributable to slow binding, as demonstrated with in vitro inhibition kinetics. Hence, a second generation of caspase-3 selective ABPs was developed based on the Ac-ATS010-KE (Ac-3Pal-Asp-Phe(F₅)-Phe-Asp-KE) with greatly improved binding kinetics over Ac-DW3-KE. Our probes based on Ac-ATS010-KE were made by modifying the N-terminus with 6 different linkers. All the linker modifications had limited effect on the binding kinetics, target selectivity, and pharmacokinetic profile in healthy mice. In an in vitro apoptosis model, the least hydrophilic tracer [¹⁸F]MICA-316 showed an increased uptake in apoptotic cells in comparison to the control group. Finally, [¹⁸F]MICA-316 was tested in an in vivo colorectal cancer model, where it showed a limited tumour uptake and was unable to discriminate treated tumours from the untreated group, despite demonstrating that the radiotracer was able to bind caspase-3 in complex mixtures in vitro. In contrast, the phosphatidylethanolamine (PE)-binding radiotracer [^99mTc]Tc-duramycin was able to recognize the increased cell death in the disease model, making it the best performing treatment response assessment tracer developed thus far.

Conclusions

In conclusion, a novel library of caspase-3-binding PET tracers retaining similar binding kinetics as the original inhibitor was developed. The most promising tracer, [¹⁸F]MICA-316, showed an increase uptake in an in vitro apoptosis model and was able to selectively bind caspase-3 in apoptotic tumour cells. In order to distinguish therapy-responsive from non-responsive tumours, the next generation of caspase-3-selective ABPs will be developed with higher tumour accumulation and in vivo stability.