INVOLVEMENT OF THE TRANSCRIPTION TERMINATION FACTOR, RHO, IN THE REGULATION OF DNA SUPERCOILING IN ESCHERICHIA COLI: AN EXTENDED MODEL FOR RHO-MEDIATED TRANSCRIPTION TERMINATION

GAIL FERSTANDIG ARNOLD, Purdue University

Abstract

Much remains unknown about how the Rho protein causes specific termination of transcription and release of the transcription complex and RNA in bacteria. I was intrigued by the suggestion that Rho might carry out these functions by the affecting the topology of DNA. The studies described indicate that most rho mutants examined had reduced supercoiling of plasmid DNA relative to rho $\sp+$ counterparts. Transductional studies with the strongly defective rho-15 mutant indicated that this phenomenon is a function of the rho allele (and not of a secondary mutation). Supercoiling of nucleoids was examined; rho-15 nucleoids apparently had relatively less folded genomic DNA than rho $\sp+$ nucleoids. The cell's topoisomerases, DNA gyrase and DNA topoisomerase I, were found to be expressed at normal levels in the rho-15 mutant. Levels of three other proteins examined all showed altered levels in rho-15 extracts. The altered levels of two of these, the SSB protein and DNA polymerase I, or of unexamined proteins could be responsible for the reduced plasmid supercoiling in rho $\sp-$ cells. Alternatively, the remarkable similarity seen between rho and gyrB mutants suggests a central relationship between these proteins. Furthermore, normalization of transcription in $rho\sp-$ strains with suppressor mutations was matched in every case by normalization of DNA supercoiling. This suggested a mechanistic connection between transcription termination and DNA supercoiling. To account for this, I propose the following model. When Rho protein is at the termination site, it can either directly or by stimulating RNA polymerase cause an unwinding of the DNA. The result would be to block further transcription and favor dissociation of the transcription complex and RNA from the DNA. Some of the time, DNA gyrase would be expected to respond to the transient unwinding (and resultant loss of local negative superhelicity) by introducing new superhelical turns. If mutant Rho could not induce Rho-mediated unwinding, there would be no need for a response by gyrase, possibly explaining why plasmid supercoiling in most rho mutants is reduced relative to wild type cells. The implications of this model are that the consequences of mutation in rho are derived from both transcription termination defects and reduced supercoiling.

Degree

Ph.D.

Subject Area

Biology

COinS