Speaker

Kira S. Makarova

 

Tuesday, January 22, 2008 11AM

B2 NCBI Library, Building 38A

 

Title

Comprehensive computational analysis of toxin-antitoxin systems in prokaryotes

 

Abstract

Toxin-antitoxin modules are widespread in prokaryotic genomes, highly mobile and diverse in molecular mechanisms of their action. 
Currently, dozens of unrelated systems of type 2 (where both toxin and antitoxin are proteins) and several distinct systems of type 1 
(where toxin is a protein and antitoxin is an antisense RNA) are experimentally characterized, and it is expected that many others 
will be found eventually. It is believed that these systems play major role in control for quality of cellular components, 
stress response and even management of prokaryotic population.

            The goal of this study is a comprehensive comparative-genomic analysis of toxin-antitoxin systems in prokaryotes. 
Using sensitive methods for remote sequence similarity search, genome context analysis and a new approach for detection of mobile
 two-component systems, we identified numerous, previously undetected genes that are homologous to toxins and antitoxins of known 
type 2 systems. In addition, we predict new families of toxins, antitoxins, and novel gene modules that were not previously suspected 
to function as toxins-antitoxins.  In particular we describe the HicAB cassette, a putative novel, RNA-targeting toxin-antitoxin 
system present in both archaea and bacteria. We also propose that two-gene system encoding minimal nucleotidyl transferase and 
the accompanying HEPN protein comprise a novel toxin-antitoxin system. The later system is peculiar by the fact that one s
ubfamily of HEPN domain is specific for thermophiles and therefore may be considered a genetic determinant of thermophily. 
We also report the prediction followed by experimental verification of kinase activity of a HipA toxin of HipAB toxin-antitoxin
 module which is known to be responsible for multidrug tolerance in Escherichia coli and some other pathogenic bacteria. 
Finally we show some evidences that the majority of the type I toxins belong to holin superfamily and predict numerous new toxins 
of this family in many bacterial genomes.