An Immunological QSP Model Incorporating Peptide-HLA Binding to Explain Variability in CD4+ T-Cell Responses to a Subunit Varicella-Zoster Vaccine

Posted on by RoryMarriott

Background: T-cell response to vaccination is highly variable, making it difficult to predict vaccine efficacy from clinical trials. CD4+ T cell activation requires presentation of a pool of antigenic peptides to a diverse repertoire of T cell receptors via class II major histocompatibility complexes (MHCII) [1]. Human leukocyte antigen (HLA) molecules contain the peptide binding domains of MHCII and exhibit high allelic diversity, with allele-specific binding efficiencies. We aimed to develop an immunological QSP model with peptide-HLA binding parameters identifiable from experimental data, enabling mechanistic characterisation of the observed inter-individual variability in clinical CD4+ T-cell response to vaccination.

Methods: An immunological QSP model of CD4+ T-cell response was developed in MATLAB Simbiology (2024b), incorporating features from published models [2-4]. Model parameters were calibrated to literature data, including clinical CD4+ T-cell response to an adjuvanted glycoprotein E (gE) subunit vaccine (Shingrix®) [5, 6]. An effective binding model was implemented to describe interactions in a many-to-many network of peptide-HLA allele combinations, using an effective dissociation constant, which extends the work of Rodriguez Messan et al. [7]. This approach reduces model complexity by representing all peptide-HLA pairs with a single binding equation. Binding efficiency was calibrated using published in vitro peptide-HLA binding data for gE [8]. Allele-specific data was used to predict the variability in CD4+ T-cell response to Shingrix®. Global and local sensitivity analyses (GSA/LSA) were used to further characterise the binding effect.

Results: The model described the observed saturated CD4+ T-cell response across three doses of Shingrix®. This was attributed to the common adjuvant dose and saturated antigen presentation in the model. Calibrating the binding model to specific HLA alleles predicted the variability in CD4+ T-cell response to Shingrix® solely from peptide-HLA binding efficiency. GSA revealed that peptide-HLA and MHCII-TCR binding are primary mechanisms that drive CD4+ T-cell response. LSA demonstrated that a 50% increase in peptide-HLA binding efficiency leads to a >300% increase in CD4+ T Cell response, corresponding to the exponential growth of T-cells activated from the peptide signal. This work represents the first application of HLA binding data in an immunological QSP framework to mechanistically describe inter-individual variability in vaccine response, moving beyond predictive peptide-binding algorithms that capture only the average binding dynamics [7, 9].

Conclusion: The effective binding model represents a method for incorporating the binding of many peptide-HLA allele pairs into a single binding equation within an immunological QSP model. This model has the potential to be applied to clinical trials with covariate information of participant HLA backgrounds to predict individual response to vaccination.

References

  1. Soto, C., et al., High Frequency of Shared Clonotypes in Human T Cell Receptor Repertoires. Cell Rep, 2020. 32(2): p. 107882.
  2. Wang, H., et al., Quantitative systems pharmacology model predictions for efficacy of atezolizumab and nab-paclitaxel in triple-negative breast cancer. J Immunother Cancer, 2021. 9(2).
  3. Chen, X., T. Hickling, and P. Vicini, A Mechanistic, Multiscale Mathematical Model of Immunogenicity for Therapeutic Proteins: Part 1—Theoretical Model. CPT: Pharmacometrics & Systems Pharmacology, 2014. 3(9): p. 1-9.
  4. Lee, H.Y., et al., Simulation and prediction of the adaptive immune response to influenza A virus infection. J Virol, 2009. 83(14): p. 7151-65.
  5. Chlibek, R., et al., Safety and immunogenicity of three different formulations of an adjuvanted varicella-zoster virus subunit candidate vaccine in older adults: a phase II, randomized, controlled study. Vaccine, 2014. 32(15).
  6. Hastie, A., et al., Immunogenicity of the Adjuvanted Recombinant Zoster Vaccine: Persistence and Anamnestic Response to Additional Doses Administered 10 Years After Primary Vaccination. The Journal of Infectious Diseases, 2021. 224(12): p. 2025-2034.
  7. Rodriguez Messan, M., et al., Mathematical model of a personalized neoantigen cancer vaccine and the human immune system. PLoS Comput Biol, 2021. 17(9): p. e1009318.
  8. Voic, H., et al., Identification and Characterization of CD4(+) T Cell Epitopes after Shingrix Vaccination. J Virol, 2020. 94(24).
  9. Nielsen, M., et al., NetMHCpan-4.1 and NetMHCIIpan-4.0: improved predictions of MHC antigen presentation by concurrent motif deconvolution and integration of MS MHC eluted ligand data. Nucleic Acids Research, 2020. 48(W1): p. W449-W454.
2026