Synthetic Analogs Tailor Native AI-2 Signaling Across Bacterial Species

Abstract

The widespread use of antibiotics and the emergence of resistant strains call for new approaches to treat bacterial infection. Bacterial cell-cell communication or "quorum sensing" (QS) is mediated by "signatures" of small molecules that represent targets for "quenching" communication and avoiding virulent phenotypes. Only a handful of small molecules that antagonize the action of the "universal" autoinducer, AI-2, have been reported. The biological basis of antagonism, as well as the targets for these select few AI-2 antagonists, have not been clearly defined. We have developed C-1 alkyl analogs of AI-2 that quench the QS response in multiple bacterial species simultaneously. We also demonstrate the biological basis for this action. Like AI-2, the analogs are activated by the bacterial kinase, LsrK, and modulate AI-2 specific gene transcription through the transcriptional regulator, LsrR. Interestingly, addition of a single carbon to the C1-alkyl chain of the analog plays a crucial role in determining the effect of the analog on the QS response. While an ethyl modified analog is an agonist, propyl becomes an antagonist of the QS circuit. In a trispecies synthetic ecosystem comprised of E. coli, S. typhimurium, and V. harveyi we discovered both cross-species and species-specific anti-AI-2 QS activities. Our results suggest entirely new modalities for interrupting or tailoring the network of communication among bacteria.

Exploiting Quorum Sensing To Confuse Bacterial Pathogens

“…First, as anticipated, numerous DPD derivatives were found to be AI-2 QS antagonists in one or multiple bacterial species (209)(210)(211)(212)(213). Second, some of the analogues were found to be synergistic agonists despite having no agonist activity on their own (209)(210)(211)213).…”

“…For both types of QS-modulating activity, species specificity was apparent. For example, propyl-DPD and other analogues with longer alkyl chain substitutions were all antagonists of AI-2 QS in E. coli, but only butyl-DPD had antagonistic activity in S. Typhimurium (212). Likewise, cyclic DPD analogues generally inhibited AI-2 QS in E. coli, whereas they either had no activity or functioned as synergistic agonists in S. Typhimurium (209).…”

“…Biochemical and biological analyses indicate that phosphorylation of AI-2 analogues is necessary but not sufficient for antagonism in E. coli and S. Typhimurium (212) and that both the shape and flexibility of the analogue are likely important for the generation of productive interactions between the analogue and AI-2 receptor (209). Also, given the species specificity exhibited by many of the analogues tested to date, disruption of AI-2 signaling in multispecies environments will likely require multiple AI-2 analogues used in combination that can inhibit AI-2 QS in various species simultaneously (209,212) or in combination with inhibitors of other QS pathways that contribute to virulence alongside the AI-2 pathway. Finally, although AI-2 derivatives do not appear to be cytotoxic toward mammalian cells (211), work will need to be done to assess the stability and efficacy of AI-2 analogues in vivo.…”

“…Likewise, cyclic DPD analogues generally inhibited AI-2 QS in E. coli, whereas they either had no activity or functioned as synergistic agonists in S. Typhimurium (209). Additionally, the activity of a given analogue in a given strain may be context dependent, as multiple C1-alkyl DPD derivatives were found to function as either antagonists or synergistic agonists in V. harveyi under different assay conditions (210)(211)(212)(213).…”

“…The best inhibitor identified from these screens was pyrogallol (Table 3), with an IC 50 of 2 M. More recently, there have been multiple reports of the generation and screening of C1-and C2-alkyl DPD derivatives. C1-alkyl derivatives have included linear, branched, cyclic, and aromatic DPD analogues (209)(210)(211)(212)(213), whereas the C2 derivatives tested have thus far been limited to linear substitutions of various alkyl chain lengths (214). None of these C1-or C2-alkyl analogues were more potent inhibitors than the previously identified pyrogallol, but some did have AI-2-modulating activity.…”