Parameters affecting transcription termination by Escherichia coli RNA polymerase : I. Analysis of 13 rho-independent terminators

Abstract

Escherichia coli RNA polymerase can terminate transcription efficiently at rho-independent terminators in a purified transcription system in the absence of accessory factors. This process of “intrinsic termination” involves direct recognition of the terminator by the core RNA polymerase, and provides an important model system for the study of the molecular interactions involved in the switch between elongation and termination. We have analyzed the intrinsic termination efficiency (%T) of 13 rho-independent terminators, under a variety of in vitro reaction conditions. Although all of these sites share the general sequence features of typical rho-independent terminators, we find a wide range of %T (2% to 90%) for the different sites under our standard transcription conditions. While %T for a particular site is characteristic of that site, the efficiency can be altered considerably by the nature and concentration of salts in the reaction, by alteration of the concentrations of the nucleoside triphosphate substrates, or by transcription from supercoiled rather than linear templates. Surprisingly, different conditions can alter %T to a different extent for different terminators. For neutral salts such as potassium chloride or potassium glutamate, changes in the range from 0.1 to 1 m affect %T for different terminators in a distinct manner, depending on the terminator and the anion involved. At some sites, %T is greatly increased by Cl − concentrations up to 1 m, while at other sites %T is reduced or unaffected by these conditions. At some sites K + concentrations up to 1 m give a modest increase in %T, while at other sites %T is slightly reduced under the same conditions. Thus the actual values of %T, as well as the order of terminator sites ranked according to %T, can be altered greatly according to the choice of reaction conditions. Reduction of the Mg 2+ concentration below 1 m m has a dramatic and quite different effect, enhancing termination to approximately 100% for all terminators tested. Transcription of supercoiled DNA templates gives somewhat reduced %T as compared with linear DNA templates. However, the effect is no greater than twofold. Our results are not consistent with those expected for models in which %T is determined by the differential stability of DNA, RNA and hybrid duplex structures at the melted region in the transcription complex. Thus, the Cl anion does not affect the stability of nucleic acid duplexes even at 1 m concentrations, but can enhance termination tenfold. Also, the alterations of monovalent cation concentration that affect % T are not expected to have a differential effect on T m for DNA, RNA and hybrid duplexes. Rather, it appears that effects of salts on protein-protein and/or protein-nucleic acid interactions must account for these changes in %T. Finally, the modest reduction in termination with supercoiled templates is inconsistent with expectations of models in which the stability of a long (7 to 12 base-pair) DNA-RNA hybrid determines %T. We suggest that in vitro termination is a multi-step process of unexpected complexity, in which RNA polymerase actively controls the efficiency of the overall reaction.