Robert E. Webster

Abstract A nuclease with specificity for double-stranded RNA (RNase III) has been found in extracts of Escherichia coli. It remains within osmotically shocked cells attached to the ribosomes. It sediments with the ribosomes in less than 0.20 m NH4Cl, but is detached at higher concentrations. A purification is described which utilizes this property. The free enzyme, after diethylaminoethyl Sephadex and carboxymethyl Sephadex chromatography, shows an absolute specificity for polymers containing double helical polyribonucleotide regions—other polymers (single- and double-stranded DNAs, single-stranded RNAs) are not digested, nor do they inhibit the digestion of double-stranded RNAs when present in excess. RNase III shows an absolute requirement for divalent cations, including Mg++ and Mn++, and for monovalent cations, including NH4+, K+, and Na+. The optimal pH range for the reaction is from pH 7.6 to pH 9.75. RNase III is inactivated by exposure to many substances at monovalent cation concentrations below 0.2 m. The enzyme is not a factor required for protein synthesis in vitro. Its mode of action appears to be endonucleolytic.

Genetic studies have identified a number of genes whose products appear to be required for the transport of the group A colicins and the single-stranded DNA of certain filamentous bacteriophages into Escherichia coli. Mutations in these genes allow normal binding of the colicins to their outer-membrane receptors and of the bacteriophage of the tip of specific conjugative pili, but do not allow translocation of the macromolecules to their target. These mutations have been designed 'tolerant' (tol) mutations and the protein products specified by these genes appear to comprise part of a transport system known as the Tol import system. Some of these genes have been isolated, sequenced and their protein products localized to the membranes or periplasm of E. coli. Information is also available regarding the domains of the colicins or phage proteins which interact with the Tol proteins. A preliminary model of the location and possible interactions of the Tol proteins is presented.

Mutations in fii or tolA of the fii-tolA-tolB gene cluster at 17 min on the Escherichia coli map render cells tolerant to high concentrations of the E colicins and do not allow the DNA of infecting single-stranded filamentous bacteriophages to enter the bacterial cytoplasm. The nucleotide sequence of a 1,854-base-pair DNA fragment carrying the fii region was determined. This sequence predicts three open reading frames sequentially coding for proteins of 134, 230, and 142 amino acids, followed by the potential start of the tolA gene. Oligonucleotide mutagenesis of each open reading frame and maxicell analysis demonstrated that all open reading frames are expressed in vivo. Sequence analysis of mutant fii genes identified the 230-amino acid protein as the fii gene product. Chromosomal insertion mutations were constructed in each of the two remaining open reading frames. The phenotype resulting from an insertion of the chloramphenicol gene into the gene coding for the 142-amino acid protein is identical to that of mutations in fii and tolA. This gene is located between fii and tolA, and we propose the designation of tolQRA for this cluster in which tolQ is the former fii gene and tolR is the new open reading frame. The protein products of this gene cluster play an important role in the transport of large molecules such as the E colicins and filamentous phage DNA into the bacterium.

The group A colicins and the DNA of many single-stranded filamentous bacteriophage are able to use combinations of the Tol proteins to gain entrance into or across the membrane of Escherichia coli. The TolA protein is a 421-amino acid residue integral membrane protein composed of three domains. Domain I, consisting of the amino-terminal 47 amino acids, contains a 21-residue hydrophobic segment that anchors the protein in the inner membrane. The remaining 374 amino acids, containing the other two domains, reside in the periplasmic space. Domain III, consisting of the carboxyl-terminal 120 residues, is considered to be the functional domain based on the location of the tolA592 deletion mutation. The internal 262 amino acids comprise domain II, which connects domains I and III together via short regions of polyglycine. It contains a large number of 3- to 5-residue polyalanine stretches, many of which have a repeat of the sequence Lys-Ala-Ala-Ala-(Glu/Asp). Circular dichroism analysis of different portions of TolA show domain II to be predominantly alpha-helical in structure while domain III contains approximately 10% helical structure.