FIGURE 3.

Antigen independent mechanisms can have varying affects across antigen specific CD8+ T cell populations, producing memory populations with a range of characteristics. (A) A single LN, wherein all Ag-independent parameters are constant, can generate a variety of EM/CM ratios and memory population sizes across Ag-specific T cell populations. Shaded area represents region of CD8+ Memory Design Space reached after simulating parameter sets sampled by sensitivity analysis. (B) Ag-independent mechanisms may have differing affects across Ag-specific populations. We can compare differences in Ag-specific responses across LNs or Ag-independent responses from distinct LNs across Ag-specific populations. (C) Ag-specific binding probability can influence the direction of correlation with P_tot. For CD8+ T cell populations with high binding probability P_tot is anti-correlated with EM/CM ratio. Leftmost panel plots Ag-specific memory populations in Memory Design Space. Each T cell population simulated had high P_i(bind) with binding threshold a_i = 30, and was generated from one of five LNs, with increasing P_tot (light to dark). All error bars represent SEM (n = 10). For T cells with medium binding probability (a_i = 150, middle panel, blue), P_tot is positively correlated with both EM/CM ratio and size of memory population. As on left, each population was generated from a separate LN with increasing P_tot, but Ag-specific populations with medium P_i(bind) are shown here. For T cells with low binding probability (a_i = 300, right panel, purple), P_tot is positively correlated with EM/CM ratio and size of memory population. For all panels, P_tot values were 100, 200, 300, 400, 500. (D) With all Ag-independent parameters held constant, the parameter set that generated the greatest diversity of EM/CM ratios across Ag-specific populations is plotted in Memory Design Space. P_i(bind) varied from high to low across T cell populations (purple = low, green = high). Error Bars represent SEM (n = 10).