PT - JOURNAL ARTICLE AU - Dinkar Wadhwa TI - Design Principle of Lysis/Lysogeny Decision vis-a-vis Multiplicity of Infection AID - 10.1101/146308 DP - 2017 Jan 01 TA - bioRxiv PG - 146308 4099 - http://biorxiv.org/content/early/2017/06/06/146308.1.short 4100 - http://biorxiv.org/content/early/2017/06/06/146308.1.full AB - Bacteriophage lambda possesses dual strategy of replication. Upon infecting its host, Escherichia coli, it can either choose lytic pathway, in which the host undergoes lysis, releasing hundreds of progeny viruses, or opt for lysogeny, in which the viral genome exists as part of bacterial chromosome known as prophage. Classic and molecular studies have shown that the lysis/lysogeny decision depends upon the number of coinfecting phages, viz. the multiplicity of infection (MoI): lysis at low MoI; lysogeny at high MoI. Here, by constructing an expression for quality of the lysis/lysogeny minimalist two-protein switch which, beside another thing, demands high equilibrium concentration of Cro-like protein (Lyt) and low equilibrium concentration of CI-like protein (Lys) - that is, lytic development - at MoI of 1, and vice versa - that is, lysogeny development - at MoI of 2, I demonstrate that positive feedback loop formed by activation of cI’s transcription by its own product in a cooperative manner underlies the switch’s design. The minimalist two-protein model, in which Lys performs exactly the same function as CI does in lambda phage’s genetic regulatory network (GRN), is justified by showing its analogy with the GRN responsible for lysis/lysogeny decision. Existence of another stable state at MoI of 1 is argued to be responsible for lysogen stability. Further, by comparing the minimalist model and its variants, possessing the positive feedback loop, with other models, without having the positive feedback loop, such as the mutual repression model, it is shown why lysis/lysogeny switch involving positive autoregulation of cI is evolved instead of one without it. A three-protein model, which is very close to lambda’s GRN, is shown to be equivalent to a close variant of the two-protein minimalist switch. Finally, only a fraction of parameter sets that produced switch deterministically were able to do so, if at all, under stochastic simulations more than 95% of the time. Additionally, another stable state at MoI of 1 was not found during stochastic simulation.