Poster M9 - Mathematical Modeling of the T cell Response to Vaccination

Introduction: Our ongoing research is driven by a goal to enhance vaccine development by using computational systems biology. In the context of vaccine development, there are growing demands for safer and more effective vaccines, particularly for vulnerable populations. Our approach combining laboratory experiments and mathematical modeling is essential for translating research findings into clinical therapies. Vaccines capable of producing T cell responses that protect against certain viral or parasitic infections are increasingly important, but efforts have been largely unsuccessful. Our overall aim is to develop a mathematical model that can predict the generation of protective memory T cell responses using data from time points early after immunization. In this work, a stochastic mathematical model has been developed to accurately recapitulate the T cell response in mice to a single dose vaccination with an attenuated parasite. By tracking individual T cells from priming to memory development, this model captures the stochasticity of this process. This research aims to bridge this gap by developing an accurate and efficient mathematical model capable of predicting the immune response to vaccination. Having such a model would accelerate the transformation of basic vaccination research into clinical therapy.

Materials and

Methods: A comprehensive stochastic dynamic model is used to simulate the priming of naïve T cells by antigen presenting cells (APCs) and their subsequent expansion, differentiation, and contraction within the mouse after a single intraperitoneal dose vaccination with an attenuated parasite. Key tissues were sampled at different time points during the vaccination to measure the kinetics of the total T cell count and phenotype. An agent-based stochastic model was developed using MATLAB where each event that occurs, such as two cells binding, cells dividing, or cells dying, is governed by a probability distribution that is either taken from the literature or is unknown and estimated through an optimization algorithm. A particle swarm optimization algorithm is used to estimate the unknown model parameters based on the experimental data aiming to minimize the sum of squared errors between the simulation results and experimental data.

Results, Conclusions, and Discussions:

Results:
The original STORE model (STochastic Omentum Response) model was developed by ​(Christian, 2022)​. We have expanded the STORE model to track T cell counts up to 47 days comparing to the previous limit which was 8 days. This expansion also required adding new functions to the model to track T cell counts across the entire mouse and distinguish between the differentiation of memory and effector T cells. Modeling the generation of memory T cells will allow for the prediction of whether a vaccine formulation could provide long-lived immunity. Understanding the dynamics of these cells helped us to improve our prediction of the immune response for a longer time period. Our model simulation results are given in Figure 1, which shows that the model aligns closely with experimental results for total T cell number as well and effector and memory T cell differentiation.

Conclusions:
In conclusion, the new STORE model has significantly improved its capability to simulate immune responses and accurately track different T-cell types up to 47 days. This achievement aligns ​(Christian, 2022)​closely with experimental data, confirming the model’s functionality.

Discussion:
Moving forward, we would like to further validate our model using experimental data from mouse genotypes that fail to generate protective T cell responses using this immunization model. If the model can accurately predict how the T cell response differs compared to wild-type (WT) mice, it will serve to both confirm the model’s accuracy and provide important biological insights on the conditions required to generate a protective T cell response from vaccination. Finally, we aim to apply the model to mRNA-LNP vaccination strategies in mice with the goal of translating the model to human patient data.

Acknowledgements (Optional):