Ecological determinants of altruism in prokaryote antivirus defense

Prokaryote evolution is driven in large part by the incessant arms race with viruses. Genomic investments in antivirus defense can be coarsely classified into two categories, immune systems that abrogate virus reproduction resulting in clearance, and programmed cell death (PCD) systems. Prokaryotic defense systems are enormously diverse, as revealed by an avalanche of recent discoveries, but the basic ecological determinants of defense strategy remain poorly understood. Through mathematical modeling of defense against lytic virus infection, we identify two principal determinants of optimal defense strategy and, through comparative genomics, we test this model by measuring the genomic investment into immunity vs PCD among diverse bacteria and archaea. First, as viral pressure grows, immunity becomes the preferred defense strategy. Second, as host population size grows, PCD becomes the preferred strategy. We additionally predict that, although optimal strategy typically involves investment in both PCD and immunity, investment in immunity can also result in antagonism, increasing the likelihood that a PCD-competent cell will lyse due to infection. Together, these findings indicate that, generally, PCD is preferred at low multiplicity of infection (MOI) and immunity is preferred at high MOI. Finally, we demonstrate that PCD, which is typically considered to be an altruistic trait, is in some cases neutral and can be maintained in an unstructured population over an evolutionary timescale. Our work shows that the landscape of prokaryotic antivirus defense is substantially more complex than previously suspected.