Simple experiments modifying the beginning and middle parts of genes show that most often it is just the sequence at the end of the transcribed region that specifies transcription termination. One simple mechanism for transcription termination in bacteria utilizes just the RNA polymerase and requires only a special sequence near the 3’ end of the RNA. The termination signal consists of a region rich in GC bases that can form a hairpin loop closely followed by a string of U’s. This class of terminators functions without the need for auxiliary protein factors from the cell. Frequently termination in eukaryotic systems occurs in a run of U’s as in the prokaryotic case, but no hairpin upstream is apparent.

Most likely as the RNA is elongated past the region rich in GC, it base pairs with itself to form a hairpin (Fig. 1). This hairpin may fit so poorly in the transcript groove or canyon in the polymerase (Fig. 2), that it weakens the binding of polymerase to the transcription bubble and also causes the RNA polymerase to pause in this region. Release then occurs from the run of U’s. Direct physical measurements have shown that oligo (rU:dA) hybridizes with exceptional weakness compared to other oligonucleotides. The combination of these factors changes the RNA elongation complex from being extremely stable to being so unstable that transcription termination usually occurs.

Fig1. Typical sequence and the likely hairpin structure at the 3’-OH end of the RNA whose transcription terminates without the assistance of rho protein.

Fig2. The melted DNA and RNA likely fit within canyons on RNA polymerase. The walls, however, may close in after polymerase binds to DNA to help hold onto the DNA.

A second class of prokaryotic terminators is much different. This class requires the presence of the rho protein for termination. The termination activity is further stimulated by a second protein, the nusA gene product. During transcription, RNA polymerase pauses near a termination sequence, most likely aided in pausing by the NusA protein, and then rho terminates the transcription process and releases the RNA and RNA polymerase. Analysis of the 3’ ends of rho-dependent transcripts reveals them to contain no discernible sequence patterns or significant secondary structures. They have even less secondary structure than expected for completely random sequences.

Most likely, when the nascent RNA extending from the RNA polymerase is free of ribosomes and lacks significant secondary structure, the rho protein can bind and move with the consumption of energy along the RNA up to the polymerase. When it reaches the polymerase, it separates the growing transcript from the template and terminates transcription. The separation of the two strands is accomplished by an RNA-DNA helicase (Fig. 3).

Fig3. RNA hybridized to a circular DNA molecule is a substrate for rho protein plus ATP which separates the two nucleic acids.

The discovery of rho factor was accidental. Transcription of lambda phage DNA in an in vitro system produced a large amount of incorrect transcript. This inaccuracy was revealed by hybridizing the RNA to the two separated strands of lambda phage. Correct transcripts would have hybridized predominantly to only one strand. Apparently the conditions being used for transcription did not faithfully reproduce those existing within the cell and the rho factor somehow reduced the amount of incorrect transcription. This is a biochemist’s dream for it means that something must exist and is waiting to be found. Therefore Roberts looked for and found a protein in cell extracts that would enhance the fidelity of in vitro transcription. Upon completing the purification and in studying the properties of his “fidelity” factor, he discovered that it terminated transcription. Rho factor shares suggestive properties with a DNA helicase required in DNA replication, DnaB. Both bind to nucleic acid and move along the nucleic acid with the consumption of ATP. In the process of this movement, a complementary strand can be displaced. Further, both helicases are hexameric, and both hydrolyze significant amounts of ATP when in the presence of a single-stranded oligonucleotide.

Although transcription termination and its regulation is usually determined only by events occurring near the 3’ end of the transcript, sometimes the 5’ end is also involved. The transcription in E. coli of ribosomal RNA and of some of the genes of phage lambda depends upon modifying the transcription complex shortly after the polymerase crosses a special sequence near the promoter. At this point several proteins bind to the RNA copy of the sequence and bind to the RNA polymerase as well. After such a modification, elongation will proceed to the end of the transcription unit and ignore some opportunities for termination that would be utilized by an unmodified RNA polymerase. It would be amazing if eukaryotic cells did not also utilize this as a mechanism for gene regulation.