398 Nature Vol. 272 30 March 1978 capacity for ribosome binding of However, binding efficiency was RNA or by a stable helical region in different size fragments prepared from reduced by removal of either the cap the DNA template. However, it is clear the 5' end of reovirus mRN A. Frag­ or of sequences comprising the begin­ that such stable regions are by them­ ments up to 30 nucleotides in length, ning of the coding region. Their con­ selves probably insufficient for chain which contained the cap but not the clusion is that, as with prokaryotic release to occur. Within the tRNATF initiation codon, did not form stable ribosome binding sites, the formation of gene a 178 base pair unit is completely complexes with ribosomes whereas the initiation complex is primarily de­ repeated three times. Each of these every fragment which did contain the pendent on the interaction between the repeats contains both the proximal AUG was able to bind ribosomes. AUG and the initiator tRNA. D and distal GC-rich sequences in their entirety. Nevertheless only two out of the three such regions can apparently serve as terminators in vitro. The region that is inactive differs by a C~T transition from the other two, a change that occurs in a sequence CAATCAA common to both the tRNAT y,. and t111 terminators. Ktipper Mechanism of et al. suggest that this sequence may be the recognition site for p-induced transcription termination chain release. In phage "A. the control of terminator from Andrew Travers function is believed to have a crucial role in the lytic and lysogenic pathways of development. Thus expression of the The regulation of transcription ter­ et al. (page 423) describe the entire genes distal to tm is dependent on the mination is a crucial aspect of the sequence of the tRNA?Y• gene distal function of the N gene product, a control of gene expression in bacteria to the promoter. Termination at both protein whose postulated role is that and bacteriophage. The termination tRJ and the major tRNAT,,. terminator of an anti-terminator at a majority of an RN A chain requires first that is absolutely p-dependent. The most of termination sites (Roberts Nature the transcribing polymerase ceases its striking difference between these 224, 1168; 1969). pN will however only movement along the DNA template, terminators and the p-independent mediate antitermination when tran­ then that both the completed RNA terminators so far studied is the absence scription originates from either of the molecule and RNA polymerase be of a terminal run of uridine residues. two early promoters suggesting that released. In vitro the DNA sequences However tRJ does contain a sequence recognition sites for pN (nut) are signalling termination differ in their for a comparatively abbreviated stem­ situated in reasonable proximity to effectiveness. At some sites RNA loop structure just before the 3' -end these promoters. One such site nutn, polymerase can terminate by itself, at of the transcript. The functional has been identified by Rosenberg et al .. others RNA release requires an significance of this region is made as a region of dyad symmetry im­ additional protein factor. However, apparent by mutations which alter the mediately preceding the terminator tn1. polymerase can still pause at such p• effectiveness of the tR1 terminators. The Thus the regulator recognition signal dependent sites even in the absence of cin and enc mutations both act in cis and its site of action are, in this the factor. to enhance or eliminate respectively p• instance, in close juxtaposition. What is the nature of the termination dependent termination at this site. The The sequence analysis of the "A. signals? Analysis of several p -indepen­ single base change in the cin mutant genome provides tantalising hints about dent terminators has revealed that in extends the length and stability for the mechanism of the choice between every case termination occurs distal to the potential stem-loop structure while, the lytic and lysogenic modes of a GC-rich region within a run of by contrast, those of the enc mutants development. The cin and enc mutants uridine residues. In addition all such reduce its stability. Accordingly, favour the lysogenic and lytic modes terminators display a region of dyad Rosenberg et .al. speculate that the respectively, suggesting that reduction symmetry in the DNA just proximal stability of this structural feature is of transcription beyond tm allows to the termination point. In terms of an important determinant of ter­ transcription to proceed in the reverse RNA structure this implies that the mination efficiency although they point direction from the promoter PnE whose molecule has the potential to form a out that the cin mutation could exert transcript acts as a messenger for the stem-loop configuration close to the 3' - its effect in another way by generating production of cl repressor. Rosenberg terminus. The stability and size of these a sequence believed to be important in et al. and Schwarz et al. have also structures is variable, termination of promoter function and thus stabilising defined the sites of the c Y mutations, those RNA molecules possessing the polymerase bound at tRJ. whose phenotypic consequence is the least stable stems being in general By contrast at the main tRNA Tyr reduction of repressor synthesis during more dependent on p. terminator no such stable stem-loop the establishment of lysogeny. These The advent of modern techniques structure is apparent. Instead a cluster cis acting mutations apparently define has considerably enhanced the direct of four G·C base pairs precedes the a regulatory sequence annd again occur sequence analysis of DNA regions termination point by one turn of the in regions of dyad symmetry. However with particular regulatory functions. DNA double helix, a configuration it remains unclear whether the c Y In this issue of Nature, (page 414 and believed to be sufficient to halt region is the promoter site for Ptt E, a 410), Rosenberg et al. and Schwarz et transcription temporarily (Maizels terminator site for a transcript initiated al. present DNA sequences from phage Proc. natn Acad. Sci. U.S.A . 70, from Pni,;, or a regulator recognition encompassing the first terminator, tu1 , 3585; 1973). In addition a more ex­ site. The identity of one section of the for RNA molecules initiated at a major tensive GC-rich region follows the c Y region with part of the tR1 region early promoter, PR, while Ktipper termination point. Consequently may indicate a common function but Ktipper et al. suggest that a transcribing clearly the precise elucidation of the Andrew Travers is at the MRC Laboratory polymerase can be halted either by a function of these sequences awaits a of Molecular Biology. Cambridge. stem-loop structure in the product rigorous biochemical analysis D

0028-0836/78/0272-0398$01.00 © Macmillan Journals Ltd 19111