Lack of genomic evidence of AI-2 receptors suggests a non-quorum sensing role for <it>luxS </it>in most bacteria

Abstract Background Great excitement accompanied discoveries over the last decade in several Gram-negative and Gram-positive bacteria of the LuxS protein, which catalyzes production of the AI-2 autoinducer molecule for a second quorum sensing system (QS-2). Since the luxS gene was found to be widespread among the most diverse bacterial taxa, it was hypothesized that AI-2 may constitute the basis of a universal microbial language, a kind of bacterial Esperanto. Many of the studies published in this field have drawn a direct correlation between the occurrence of the luxS gene in a given organism and the presence and functionality of a QS-2 therein. However, rarely hathe existence of potential AI-2 receptors been examined. This is important, since it is now well recognized that LuxS also holds a central role as a metabolic enzyme in the activated methyl cycle which is responsible for the generation of S-adenosyl-L-methionine, the major methyl donor in the cell. Results In order to assess whether the role of LuxS in these bacteria is indeed related to AI-2 mediated quorum sensing we analyzed genomic databases searching for established AI-2 receptors (i.e., LuxPQ-receptor of Vibrio harveyi and Lsr ABC-transporter of Salmonella typhimurium) and other presumed QS-related proteins and compared the outcome with published results about the role of QS-2 in these organisms. An unequivocal AI-2 related behavior was restricted primarily to organisms bearing known AI-2 receptor genes, while phenotypes of luxS mutant bacteria lacking these genes could often be explained simply by assuming deficiencies in sulfur metabolism. Conclusion Genomic analysis shows that while LuxPQ is restricted to Vibrionales, the Lsr-receptor complex is mainly present in pathogenic bacteria associated with endotherms. This suggests that QS-2 may play an important role in interactions with animal hosts. In most other species, however, the role of LuxS appears to be limited to metabolism, although in a few cases the presence of yet unknown receptors or the adaptation of pre-existent effectors to QS-2 must be postulated.