Engineering Protein-Drug Conjugates for Solid Tumor Immunotherapy

Abstract

Antibody therapies have dominated biologics for many years. The innate specificity and long half-lives of these proteins provide a sustained response in many disease indications, including cancer. However, solid tumors can often evade treatment by antibodies alone which has led to the development of antibody-drug conjugates (ADCs). Although successful development of ADCs is a recent endeavor, the concept of a “magic bullet” or an agent that could kill unhealthy cells while leaving healthy cells unharmed was discussed in the early 1900’s. ADCs have experienced a rocky start but are currently experiencing unprecedented success. To better understand the mechanisms behind these clinical trends, this work first improved upon previously existing near infrared fluorophore techniques to increase the limit of detection of overall antibody uptake and cellular degradation for use in low expression target systems. Conjugating proteins of interest to a non-residualizing and residualizing fluorophore provided total protein uptake, tumor distribution, and cellular degradation in vivo. Utilizing similar fluorescent techniques, the tumor distribution for a panel of protein-drug conjugates with varying binding kinetics and size was tracked in vitro and in vivo. Computational simulations with a Krogh cylinder model along with in vitro and in vivo assays highlighted the internalization rate as a key feature to balance tumor penetration with cellular uptake and potency. The results demonstrated that rapid internalization rates induced by biparatopic antibodies with a highly potent payload against a highly expressed target (Prostate Specific Membrane Antigen, PSMA) increase the potency against cell lines in vitro but lead to greater heterogeneity in delivery and lower efficacy in vivo. In contrast, a slower internalization rate allowed for greater tumor penetration and higher efficacy even with decreased payload delivery per cell. Therefore, greater tumor penetration was beneficial for tumor efficacy with subsaturating doses. However, more recent ADCs utilize moderate potency payloads that can deliver saturating doses to the tumor due to their greater tolerability. Next, the impact of internalization rate under saturating conditions for anti-CEA ADCs. In contrast to the subsaturating PSMA ADCs, the more rapidly internalized anti-CEA cross-linking ADC delivered more payload to each cell in vitro and in vivo, resulting in greater efficacy when compared to a slower internalizing ADC. This work demonstrated the need to design ADCs to match the key tumor parameters associated with cell delivery to achieve a therapeutic dose in the largest fraction of tumor cells. Finally, a model system was developed to determine the relative importance of the three known mechanisms of ADC action: 1) receptor modulation, 2) payload delivery, and 3) Fc-effector function. In a genetically engineered mouse model, each mechanism played a role in the overall efficacy. In conclusion, this work developed improved methods for tracking ADCs in vitro and in vivo and demonstrated how internalization rate can be modified, depending on the target and therapeutic/payload properties, to provide maximum effect in vivo. Importantly, these results can differ from maximum efficacy in vitro, but computational methods were shown to be able to predict the optimal properties to aid in effective ADC design.