Guarding versus self-guarding in innate immunity

Hosts have evolved a variety of innate immune responses to pathogens. In many cases, hosts directly detect pathogen-associated molecular patterns (PAMPs) or pathogen effectors to trigger an immune response. However, hosts may also detect pathogens indirectly through 'guarding', whereby immune receptors ('guards') monitor the effects of pathogens (e.g., modification of target cells) rather than the pathogens themselves. Guarding poses a different evolutionary challenge for pathogens than direct recognition of PAMPs, as replication may necessitate the modification or disruption of guarded host proteins ('guardees'). Recently, self-guarding has been discovered, in which the host target functions as both guard and guardee. Self-guarding appears to present an intractable problem for pathogens: modification of the host target may benefit replication, but also triggers an immune response. If self-guarding creates an apparently inescapable detection mechanism, why has self-guarding only recently been discovered? Here, we use mathematical models of within-host pathogen and immune dynamics to compare guarding and self-guarding architectures. We show that self-guarding leads to a more rapid immune response and faster pathogen suppression, but is also more prone to false-positive immune responses, likely imposing greater costs through autoimmunity. We therefore hypothesise that the greater potential for false-positive immune responses may limit the conditions under which self-guarding evolves.