Dependence of post-segregational killing mediated by Type II restriction–modification systems on the lifetime of restriction endonuclease effective activity

Abstract

Plasmid-borne Type II restriction–modification (RM) systems mediate post-segregational killing (PSK). PSK is thought to be caused by the dilution of restriction and modification enzymes during cell division, resulting in accumulation of unmethylated DNA recognition sites and their cleavage by restriction endonucleases. PSK is the likely reason for stabilization of plasmids carrying RM systems in the absence of selection for plasmid maintenance. In this study, we developed a CRISPR interference-based method to eliminate RM-carrying plasmids and study PSK-related phenomena with minimal perturbation to the Escherichia coli host. Plasmids carrying the EcoRV, Eco29kI, and EcoRI RM systems were highly stable, and their loss resulted in SOS response and PSK. In contrast, plasmids carrying the Esp1396I system were poorly stabilized; their loss led to a temporary cessation of growth, followed by full recovery. We demonstrate that this unusual behavior is due to a limited lifetime of the Esp1396I restriction endonuclease activity, which, upon Esp1396I plasmid loss, disappears approximately after two cycles of cell division, i.e., before unmethylated sites appear in significant numbers. Our results indicate that whenever PSK induced by a loss of RM systems, and, possibly, other toxin–antitoxin systems, is considered, the lifetimes of individual system components and the growth rate of host cells shall be taken in account. Mathematical modeling shows, that unlike the situation with classical toxin–antitoxin systems, RM system-mediated PSK is possible when the lifetimes of restriction endonuclease and methyltransferase activities are similar, as long as the toxic restriction endonuclease activity persists for more than two chromosome replication cycles. IMPORTANCE It is widely accepted that many Type II restriction–modification (RM) systems mediate post-segregational killing (PSK) if plasmids that encode them are lost. In this study, we harnessed an inducible CRISPR-Cas system to remove RM plasmids from Escherichia coli cells to study PSK while minimally perturbing cell physiology. We demonstrate that PSK depends on restriction endonuclease activity lifetime and is not observed when it is less than two replication cycles. We present a mathematical model that explains experimental data and shows that unlike the case of toxin–antitoxin-mediated PSK, the loss of an RM system induced PSK even when the RM enzymes have identical lifetimes.