Senior Research Investigator Bristol Myers Squibb Hillsborough, New Jersey, United States
Objective: Antibody mediated cell-cell contacts can trigger events, including immune cell suppression, cytokine release, cellular cytotoxicity, phagocytosis or T cell activation. We present a novel shared-pool modeling approach, which minimizes the number of differential equations necessary to describe the system by pooling repeated reactions occurring on different cell states that extends upon prior insights from proper dimensional representation of surface reactions^1. The model depicts the formation of ternary complexes comprised of effector-receptor, antibody, and target-receptor at immunological synapses which provides a unified framework to describe cell-cell interactions and receptor dynamics in a mass-balanced manner.
Methods: The dynamic variables in our model are the numbers of cell states (effector cells, target cells, and effector-target complexes), receptor pools on each cell state and the number of crosslinks on effector-target complexes. Receptor pools are defined by the total number of receptors in a specific state regardless of the cell state they reside in. For example, the pool of free target receptors consists of free target receptors on target cells and on effect-target complexes and similarly for the antibody-bound target receptors. In contrast, crosslinks (or ternary receptor complexes) only exist on cell-cell complexes, which necessitates to define algebraic relations between receptor numbers on specific cell states and the pool variables in order to preserve receptor densities on free cells and cell-cell complexes. To model downstream effects such as antibody-dependent cellular cytotoxicity (ADCC), antibody-mediated phagocytosis (ADCP) or cytokine release, we use an indirect response model characterized by three parameters (k_Max, EC_50 and n_Hill), which uses the number of crosslinks per cell-cell complex as an input.
Results: We illustrate the shared pool modeling approach using an ADCP model of target cells expressing an antigen of interest. The model is informed using antibody binding data to the target receptors as well as cell-cell interaction assays with target and effector cells^2. The antibody binding on the bivalent arms shows bi-phasic shape explaining proper dimensional representations^1. The model is able to recapitulate the observed bell-shape of the cell-cell interaction assays^3, which is further utilized to estimate the EC50 value of phagocytosis. The model clearly exhibits how key parameters such as cell-cell contact fraction, target and effector receptor expression levels and binding parameters affect location and extent of the optimal antibody concentration.
Conclusions: Our novel modeling approach for antibody-mediated cell-cell interactions allows describing mechanisms such as ADCP, cell-cell contact mediated cytokine release or bi-specific T cell engagers to support clinical programs through dose selection and optimization.
Citations: 1. Schmidt, B. J., Bee, C., Han, M., Jing, Y., Cheng, Y., Tenney, D. J., & Leil, T. A. (2019). Antibodies to Modulate Surface Receptor Systems Are Often Bivalent and Must Compete in a Two‐Dimensional Cell Contact Region. CPT: Pharmacometrics & Systems Pharmacology, 8(12), 873.
2. Campbell, J. R., McDonald, B. R., Mesko, P. B., Siemers, N. O., Singh, P. B., Selby, M., ... & Lan, R. Y. (2021). Fc-optimized anti-CCR8 antibody depletes regulatory T cells in human tumor models. Cancer Research, 81(11), 2983-2994.
3. Ryall, R. G., Story, C. J., & Turner, D. R. (1982). Reappraisal of the causes of the “hook effect” in two-site immunoradiometric assays. Analytical Biochemistry, 127(2), 308-315.
