Ubiquitous Transcriptional Pausing Is Independent of RNA Polymerase Backtracking

Keir C. Neuman, Elio A. Abbondanzieri, Robert Landick, Jeff Gelles, Steven M. Block
2003 Cell  
that interact with the transcription elongation complex 1 Department of Biological Sciences (TEC) (e.g., N and Q; Richardson and Greenblatt, 1996; Stanford University Roberts et al., 1998; Weisberg and Gottesman, 1999; Stanford, California 94305 and references therein). Finally, frequent pausing is ob-2 Department of Applied Physics served during the transcription of genomic DNA (Kassa-Stanford University vetis and Chamberlin, 1981; Matsuzaki et al., 1994; Adel-Stanford, California 94305 man et
more » ... al., 2002). For bacterial RNAP, such pauses are 3 Department of Bacteriology thought to limit the overall rate of transcription, thereby University of Wisconsin synchronizing transcription with translation (translating Madison, Wisconsin 53706 ribosomes release paused RNAPs), and allowing Rho-4 Department of Biochemistry catalyzed termination of transcription should translation Brandeis University fail (Landick et al., 1985; Richardson and Greenblatt, Waltham, Massachusetts 02454 1996). Both bulk (Kassavetis and Chamberlin, 1981; Theissen et al., 1990; Matsuzaki et al., 1994; Wang et al., 1995) Summary and single-molecule (Davenport et al., 2000; Adelman et al., 2002; Forde et al., 2002) transcription experiments RNA polymerase (RNAP) transcribes DNA discontinu- demonstrate that, in most cases, only a fraction of RNAP ously, with periods of rapid nucleotide addition puncmolecules pause at any given DNA sequence. Thus, a tuated by frequent pauses. We investigated the mechpause is characterized by two empirical parameters: the anism of transcription by measuring the effect of both efficiency (equivalently, the pause probability or frehindering and assisting forces on the translocation of quency) and the lifetime (equivalently, the pause durasingle Escherichia coli transcription elongation comtion or half-life). The observation that pause efficiency plexes, using an optical trapping apparatus that allows is generally less than 100% is consistent with a branched for the detection of pauses as short as one second. reaction pathway where pauses occur by reversible We found that the vast majority of pauses are brief isomerization of the TEC into a catalytically inactive (1-6 s at 21؇C, 1 mM NTPs), and that the probability state that does not occur during the cycle of ordinary of pausing at any particular position on a DNA template elongation. is low and fairly constant. Neither the probability nor To date, only high-efficiency pauses with demonthe duration of these ubiquitous pauses was affected strated roles in transcription regulation have been charby hindering or assisting loads, establishing that they acterized in mechanistic detail (Roberts et al., 1998; do not result from the backtracking of RNAP along Artsimovitch and Landick, 2000; Palangat and Landick, the DNA template. We propose instead that they are 2001, and references therein). These pauses fall into at caused by a structural rearrangement within the least two classes, both of which arise from interactions enzyme. of RNAP with specific nucleic acid sequences, including those found in duplex DNA downstream of the RNAP Introduction active site, in the active site itself, or in the RNA:DNA heteroduplex upstream of the active site (Korzheva et The rate at which RNAP adds nucleotides to the 3Ј end al., 2000; Gnatt et al., 2001; Landick, 2001). of a growing RNA transcript is highly nonuniform. The In one class of pauses, a nascent RNA hairpin structime required for nucleotide addition increases by an ture is associated with the paused conformation of the order of magnitude or more over average during events enzyme. Here, pausing is thought to be generated by a that have been termed "transcriptional pauses" (Kassarearrangement in the active site that is stabilized by vetis and Chamberlin, 1981; Reisbig and Hearst, 1981; an allosteric interaction of the hairpin with the RNAP Kadesch and Chamberlin, 1982; Winkler and Yanofsky, (Toulokhonov et al., 2001). A second class of pauses has 1981; Levin and Chamberlin, 1987; Matsuzaki et al., been termed "backtrack" pausing, because it appears to 1994; Lyakhov et al., 1998). be associated with rearward motion of RNAP along the Elucidating the mechanisms of pausing is vital for DNA and RNA chains (Komissarova and Kashlev, 1997a; several reasons. First, pausing plays important roles in Nudler et al., 1997; Landick, 1997; Nudler, 1999; Artsitranscriptional regulation, for example, in synchronizing movitch and Landick, 2000). Backtracking has been interactions of ribosomes or transcription factors with demonstrated at sites of transcriptional arrest and when RNAP movements (reviewed in Chan and Landick, 1994; RNAPs are artificially halted by the removal of nucleo-Landick and Yanofsky, 1987; Richardson and Greenside triphosphates (NTPs) (Komissarova and Kashlev, blatt, 1996; Uptain et al., 1997). Second, pausing repre-1997b; Palangat and Landick, 2001; Reeder and Hawley, sents the first event in transcriptional termination (Kas-1996; Toulme et al., 1999). During backtracking, the 3Ј end of the nascent RNA protrudes from a pore in the enzyme thought to be the
doi:10.1016/s0092-8674(03)00845-6 pmid:14622598 fatcat:p4voehj6kfalhp7ja6ouznzcuy