Cellular heterogeneity in immune responses helps us fight diverse pathogenic threats. Here, we present a resource-constrained, optimization-based flux-allocation paradigm by which phenotypic heterogeneity emerges in macrophage polarization. We build a six-dimensional mechanistic model of arginine metabolism and study how constrained resource allocation shapes polarization and population heterogeneity. In a baseline formulation without explicit resource budgets, the system is effectively monostable under fixed cytokine inputs, and heterogeneous populations collapse onto a single polarization trajectory, consistent with well-resolving wounds. Introducing cybernetic variables that actively distribute finite resources between competing metabolic alternatives enables ultrasensitive resource partitioning that amplifies feedback and produces robust bistability under fixed cytokine cues. Embedding this intracellular model in a structured population balance framework, we show that in the advection-dominated regime, an initial macrophage cloud splits along two cybernetic equilibria, yielding approximately symmetric M1-like (iNOShigh/Arg1low) and M2-like (Arg1high/iNOSlow) peaks. Adding small isotropic diffusion in trait space, its coupling to NO-dependent apoptosis skews mass toward the Arg1high/iNOSlow branch, producing persistent asymmetric bimodality with a denser M2-like subpopulation. The model matches reported dynamics of iNOS expression, including gradual convergence of iNOSlow and iNOShigh subpopulations under ±IFNγ treatment. Such models can accelerate discovery of therapies for chronic wounds by predicting population-level responses to treatment.
Gupta et al. (Sat,) studied this question.