Research article

LuxS-independent formation of AI-2 from ribulose-5-phosphate

Timothy J Tavender^1,3 , Nigel M Halliday¹ , Kim R Hardie¹ and Klaus Winzer^1,2

¹  Institute of Infections, Immunity, and Inflammation, University of Nottingham, Centre for Biomolecular Sciences, Nottingham, NG7 2RD, UK

²  School for Health and Medicine, Division of Biomedical and Life Sciences, Lancaster University, Bailrigg, Lancaster, LA1 4YQ, UK

³  Faculty of Life Sciences, Michael Smith Building, University of Manchester, Manchester, M13 9PT, UK

author email corresponding author email

BMC Microbiology 2008, 8:98doi:10.1186/1471-2180-8-98

Published:	18 June 2008

Abstract

Background

In many bacteria, the signal molecule AI-2 is generated from its precursor S-ribosyl-L-homocysteine in a reaction catalysed by the enzyme LuxS. However, generation of AI-2-like activity has also been reported for organisms lacking the luxS gene and the existence of alternative pathways for AI-2 formation in Escherichia coli has recently been predicted by stochastic modelling. Here, we investigate the possibility that spontaneous conversion of ribulose-5-phosphate could be responsible for AI-2 generation in the absence of luxS.

Results

Buffered solutions of ribulose-5-phosphate, but not ribose-5-phosphate, were found to contain high levels of AI-2 activity following incubation at concentrations similar to those reported in vivo. To test whether this process contributes to AI-2 formation by bacterial cells in vivo, an improved Vibrio harveyi bioassay was used. In agreement with previous studies, culture supernatants of E. coli and Staphylococcus aureus luxS mutants were found not to contain detectable levels of AI-2 activity. However, low activities were detected in an E. coli pgi-eda-edd-luxS mutant, a strain which degrades glucose entirely via the oxidative pentose phosphate pathway, with ribulose-5-phosphate as an obligatory intermediate.

Conclusion

Our results suggest that LuxS-independent formation of AI-2, via spontaneous conversion of ribulose-5-phosphate, may indeed occur in vivo. It does not contribute to AI-2 formation in wildtype E. coli and S. aureus under the conditions tested, but may be responsible for the AI-2-like activities reported for other organisms lacking the luxS gene.