Modeling bacterial immune systems: strategies for expression of toxic - but useful - molecules

Protection of bacterial cells against virus infection requires expression of molecules that are able to destroy the incoming foreign DNA. However, these molecules can also be toxic for the host cell. In both restriction-modification (R-M), and the recently discovered CRISPR/Cas systems, the toxicity is (in part) avoided through rapid transition of the expression of the toxic molecules from "OFF" to "ON" state. In restriction-modification systems the rapid transition is achieved through a large binding cooperativity, and low translation rate of the control protein. On the other hand, CRISPR array expression in CRISPR/Cas systems involves a mechanism where a small decrease of unprocessed RNAs leads to a rapid increase of processed small RNAs. Surprisingly, this rapid amplification crucially depends on fast non-specific degradation of the unprocessed molecules by an unidentified nuclease, rather than on large cooperativity in protein binding. Furthermore, the major control elements that are responsible for fast transition of R-M and CRISPR/Cas systems from "OFF" to "ON" state, are also directly involved in increased stability of the steady states of these systems. We here discuss mechanisms that allow rapid transition of toxic molecules from the unproductive to the productive state in R-M and CRISPR/Cas systems. The main purpose of this discussion is to put relevant theoretical and experimental work in a perspective that points to general similarities in otherwise mechanistically very different bacterial immune systems.