Proteus mirabilis Employs a Contact-Dependent Killing System against Competing Enterobacteriaceae

Many bacterial species employ systems for interference competition with other microorganisms. Some systems are effective without contact (e.g., through secretion of toxins), while other systems (e.g., type VI secretion system [T6SS]) require direct contact between cells. Here, we provide the initial characterization of a novel contact-dependent competition system for Proteus mirabilis. In neonatal mice, a commensal P. mirabilis strain apparently eliminated commensal Escherichia coli. We replicated the phenotype in vitro and showed that P. mirabilis efficiently reduced the viability of several Enterobacteriaceae species but not Gram-positive species or yeast cells. Importantly, P. mirabilis strains isolated from humans also killed E. coli. A reduction of viability occurred from early stationary phase to 24 h of culture and was observed in shaking liquid media as well as on solid media. Killing required contact but was independent of T6SS, which is the only contact-dependent killing system described for P. mirabilis. Expression of the killing system was regulated by osmolarity and components secreted into the supernatant. Stationary-phase P. mirabilis culture supernatant itself did not kill but was sufficient to induce killing in an exponentially growing coculture. In contrast, killing was largely prevented in media with low osmolarity. In summary, we provide the initial characterization of a potentially novel interbacterial competition system used by P. mirabilis. IMPORTANCE The study of bacterial competition systems has received significant attention in recent years. These systems are important in a multitude of polymicrobial environments and collectively shape the composition of complex ecosystems like the mammalian gut. They are also being explored as narrow-spectrum alternatives to specifically eliminate problematic pathogenic species. However, only a small fraction of the estimated number of interbacterial competition systems has been identified. We discovered a competition system that is novel for Proteus mirabilis. Inspired by an observation in infant mice, we confirmed in vitro that P. mirabilis was able to efficiently kill several Enterobacteriaceae species. This killing system might represent a new function of a known competition system or even a novel system, as the observed characteristics do not fit with described contact-dependent competition systems. Further characterization of this system might help understand how P. mirabilis competes with other Enterobacteriaceae in various niches.