Bet-hedging in innate and adaptive immune systems

Abstract

Immune system evolution is shaped by the fitness costs and trade-offs associated with mounting an immune response. Costs that arise mainly as a function of the magnitude of investment, including energetic and immunopathological costs, are well-represented in studies of immune system evolution. Less well considered, however, are the costs of immune cell plasticity and specialization. Hosts in nature encounter a large diversity of microbes and parasites that require different and sometimes conflicting immune mechanisms for defense, but it takes precious time to recognize and correctly integrate signals for an effective polarized response. In this perspective, we propose that bet-hedging can be a viable alternative to plasticity in immune cell effector function, discuss conditions under which bet-hedging is likely to be an advantageous strategy for different arms of the immune system, and present cases from both innate and adaptive immune systems that suggest bet-hedging at play.

Keywords: B cells; T cells; adaptive immunity; evolutionary medicine; immune system evolution; innate immunity; macrophages; plasticity.

Figure 1.

Contrasting the efficacy of immunological bet-hedging (left plot) and polarization (right plot) under uncertain infection conditions. The polarization of immune responses (e.g. by helper T cells) relies on accurate recognition of parasite antigens, which stimulate the production of cytokines that coordinate immune responses to quickly and effectively clear viruses (facilitated by Th1 cells), extracellular microbes and parasites (facilitated by Th2 cells), and other invaders. Polarization and irreversible plasticity of polarized cells may pose an issue, however, if the host is susceptible to infection by multiple types of parasites at once. In cases like these, a polarized response aligned against one parasite type (e.g. Th1 cells against viruses) will result in an initially exponentially growing population of immune cells that effectively clear that parasite type, and hence produce an exponentially increasing clearance rate, but are ineffective at clearing or even impede the clearance of a different type of parasite. This creates substantial variance in pathogen clearance rate where some subpopulations of cells are highly effective, and others are not (right plot). On the other hand, responses that hedge their bets, in terms of producing and maintaining a subpopulation of the ‘wrong’ helper T-cell subtype, may not achieve maximum clearance efficiency against the any single infection but can avoid catastrophically slow responses against a second parasite, reducing overall variance in clearance efficacy. As a result, a bet-hedging strategy (left plot) that has a lower arithmetic mean clearance rate (dashed line) than a polarized response (right plot) can produce a higher geometric mean rate (thick line) due to its lower variance. Assuming clearance rates affect host fitness or cell subtype replication rates within a host, then host genotypes that rely on polarization will have lower geometric mean fitness than those relying on bet-hedging under these conditions. Illustrative simulations were created with a branching process whose growth rate is given by a gamma distribution. The arithmetic mean growth rate and variance in growth rate are lower in the left plot than in the right plot. Gray lines in the plots are sample trajectories and red regions denote 95% intervals.