Monday, January 14, 2008

How bacteria avoid dying alone

Programmed cell death, PCD or apoptosis, was discovered as the way individual cells altruistically remove themselves during development of a multi-cellular organism (e.g., C. elegans). However, single celled bacteria also undergo PCD to prevent the spread of a bacteriophage infection, for example. Stressed E. coli produce the stable endoribonuclease toxin MazF as well as the labile antitoxin MazE; even a transient interruption of protein synthesis can lead to PCD.

This group previously suggested that mazEF-mediated PCD depends on cell density. Here they show that indeed the mazEF system is only effective above ~3 million cells/ml. Adding supernatant from a dense culture to a diluted culture along with a stress-inducing antibiotic quickly induced PCD in wild-type (WT) cells but not cells in which mazEF is deleted (del-mazEF). This observation led them to identify an “extracellular death factor” (EDF) that is produced during log-phase growth but not during the stationary phase. Here they show that EDF is a pentapeptide (NNWNN) and that adding synthetic EDF to supernatants from stationary phase cultures renders them capable of inducing mazEF-induced PCD (Figure). High EDF (> 200 ng/ml) reduced viability of even del-mazEF strains, which the authors ascribe to perhaps inducing other PCD systems or inactivating essential components. Using mutagenesis, they demonstrated that NNWNN is the optimal sequence for EDF activity. No E. coli genes encode NNWNN but zwf-encoded NNWDN could be amidated to yield NNWNN and deletion of both zwf and ygeO, which encodes a similar NNWN peptide, prevented EDF induction. The authors propose a quorum-sensing role for EDF and point out that synthetic EDF “may be a lead for a new class of antibiotics that specifically trigger bacterial cell death".
Kolodkin-Gal et al. Science. 2007 Oct 26;318(5850):652-5. “A linear pentapeptide is a quorum-sensing factor required for mazEF-mediated cell death in Escherichia coli”.

1 comment:

Anonymous said...

The authors claim that NNWNN is the "optimal sequence for EDF function". However, the term "optimal" seems inappropriate as data for only 10 deviant peptides is reported, and a "superagonist" of quorum sensing that provokes the same response as the natural signal but at 10-fold lower concentration has been identified by a library synthesis approach (Pomianek and M.F. Semmelhack 2007). A similarly comprehensive study may identify peptides with more potent EDF activity that NNWNN. Also, the term quorum sensing may be not be a fitting description for the phenomenon described in this paper, as EDF activity increases with cell population density only until mid-log phase, then diminishes substantially as the population becomes extremely dense. With bona fide quorum sensing, autoinducer concentration would not decrease with increasing cell population density.