Figure 7. A Signaling Model for T Cell Ligand Discrimination based on Receptor Clustering and ZAP70 Recruitment.

(A) Model of for receptor clustering and phosphorylation. A single receptor-ligand interaction pins the two membranes in close apposition; unbound ligands that diffuse into this region are in closer proximity to and can more readily bind a receptor. Clustered receptors more effectively exclude the phosphatase CD45 and become phosphorylated; for lower-affinity ligands, receptor clusters and phosphorylation can be reversed by ligand dissociation, providing a mechanism for kinetic proofreading (see Discussion).

(B and C) A theoretical signaling model, incorporating experimentally measured parameters (Figure S7) was used to perform stochastic simulations of cells interacting with a ligand-functionalized SLB. Simulations were performed at ligand densities between 0.01 and 1,000 molecules/μm² and ligand dwell times (1–1000 s) for a fixed time interval of 500 s and repeated for 250 cells at each point in this parameter space. The simulation output was timeseries data of ligand binding, cluster formation, and ZAP70 recruitment, and heatmaps were generated showing the fraction of cells showing a defined characteristic of ZAP70-positive receptor-ligand clusters (see B and Figure S7K).

(B) Fraction of simulated cells with at least one receptor-ligand cluster containing eight ZAP70-positive receptors.

(C) Increasing ligand density and affinity results in an increasing number of ligand-receptor clusters. The 20% contour of these heatmaps was analyzed to demarcate regions within this parameter space where the simulation showed >20% of simulated cells forming the indicated number of clusters with a minimum of eight ZAP70-positive receptors. On this plot, the experimental data are shown for the ligand densities and dwell times at which the indicated DNA ligands elicited 20% phosphoErk-positive cells (Figures 2 and S2).