R N A Po l y mera s e
Tars K, Bundule M, Fridborg K and Liljas L (1997) The crystal structure of bacteriophage GA and a comparison of bacteriophages belonging to the major groups of Escherichia coli leviviruses. Journal of Molecular Biology 271: 759±773.
See also: Reverse Genetics; RNA Polymerase
RNA Polymerase J Parker Copyright ß 2001 Academic Press doi: 10.1006/rwgn.2001.1135
RNA polymerase is the name given to a class of enzymes which in vivo synthesize RNA molecules using double-stranded DNA as a template. Such enzymes are more properly known as DNA-dependent RNA polymerases. The copying of the information contained in a DNA sequence into an RNA sequence is termed `transcription,' a central step in biological information flow. RNA polymerase is the key enzyme involved in transcription. (Some RNA viruses encode enzymes which synthesize RNA from an RNA template. Typically such an enzyme is called an `RNA replicase,' but occasionally the term `RNA-dependent RNA polymerase' is used. These enzymes are distinct from the RNA polymerases discussed here.) All RNA polymerases synthesize an RNA chain from the 50 end to the 30 end; the template strand of the DNA is consequently read in the antiparallel 30 to 50 direction, since templating requires base-pairing. The substrates are ATP, GTP, CTP, and UTP (and magnesium ion is required). The RNA molecules are synthesized from specific starting sites on the DNA (called promoters), and RNA polymerase can initiate new chains without the requirement for a primer, unlike the case with DNA polymerase. However, DNA polymerase and RNA polymerase have essentially identical mechanisms of phosphodiester bond formation during chain elongation. Cellular RNA polymerases are multisubunit enzymes. Bacteria and Archaea each have a single RNA polymerase, while the eukaryotic nucleus contains three such enzymes: RNA polymerase I (RNAP I), RNA polymerase II (RNAP II), and RNA polymerase III (RNAP III). While there are profound differences between these multisubunit RNA polymerases, there are also significant similarities. Indeed, it is clear that these enzymes are all related and form a family. All members of this family have three different subunits, which are evolutionarily conserved to a greater or lesser extent. Bacterial RNA polymerase and the closely related chloroplast RNA polymerase contain
only these conserved subunits. The Archaea, like the Bacteria, have only a single RNA polymerase, but it is more complex than the bacterial enzyme and is more closely related to the eukaryotic RNA polymerase II. However, even in the complex eukaryotic RNA polymerases, conserved sequences make up over 50% of the enzyme mass, and therefore the simpler bacterial enzymes have provided an important model for RNA polymerase structure and function. This has been confirmed by structural analysis of the purified enzymes. However, not all RNA polymerases are multisubunit enzymes. The enzymes found in mitochondria (but encoded in the nucleus) and those encoded by some bacteriophages (for transcription) are singlesubunit enzymes. These single-subunit enzymes are not closely related to the complex cellular RNA polymerases but are more closely related to certain DNA polymerases.
Bacterial RNA Polymerases Bacteria have a single cellular RNA polymerase (RNAP), whose `holoenzyme' form has five subunits: two copies of the relatively small a-subunit (each about 36 kDa), one copy each of large b- and b0 -subunits (151 kDa and 155 kDa, respectively), and one copy of the s-subunit, also called the `sigma factor.' The `core' enzyme, of about 400 kDa, contains all the subunits except s and can carry out the elongation reaction of polymerization using a DNA template and the four substrates ATP, CTP, GTP, and UTP. The evolutionarily conserved subunits are those that make up the core. However, site-specific initiation requires the s subunit, which allows RNAP to recognize the promoter. Most bacteria encode several alternative s factors (Escherichia coli encodes seven, Bacillus subtilis encodes 17), which may vary widely in size and which allow the RNAP to recognize several different types (sequences) of promoters. If there are several different s factors in a cell, there must be several different holoenzymes and, therefore, one could say there are several different RNAPs in a given bacterium. However, this would be misleading, because the s factor (of whatever kind) is only bound to the enzyme during initiation. Also, in a given bacterium, the majority of genes typically require only a single species of sigma factor and, therefore, one form of the holoenzyme predominates. In E. coli the primary s factor, and the first discovered, has a mass of 70 kDa and is often referred to as s70. Initiation of transcription by RNAP at the promoter is a complex process involving many different steps. First, of course, the core enzyme must bind the appropriate s factor. The holoenzyme then binds to
RN A Pol y m erase 1747 promoter DNA upstream of the transcriptional start site. RNAP then interacts with the DNA, leading to melting of about 14 bp of the promoter DNA, including the transcriptional start site. There is also a conformational change of the RNAP during this process. RNAP can then begin RNA synthesis, but chain elongation often aborts, yielding short chains of less than 10 nucleotides. However, RNAP remains at the promoter and can undergo further rounds of abortive synthesis or true elongation. If the chain reaches about 10 nucleotides in length, s factor is released and the core RNAP begins moving along the DNA template, synthesizing the RNA chain. The antibiotic rifampicin specifically inhibits initiation by bacterial RNAP, at the first or second phosphodiester bond. The antibiotic binds to the b-subunit, and resistant mutants have mutations in the gene encoding this subunit. After initiation the s-subunit is released form RNAP and the elongation phase begins. Elongation by bacterial RNAP is inhibited by the antibiotic streptolydigin, which also binds to the b-subunit. During initiation the RNAP may span 70±90 bp of DNA (some of which is wrapped around the enzyme), but this is reduced to about 35 bp during elongation. The newly synthesized RNA forms base pairs with the DNA template for approximately 8 or 9 nucleotides. The newly synthesized chain exits the RNAP through a channel. The rate of elongation of an RNA chain in vivo may be about 50 nucleotides per second, but this rate is the mean of rapid elongation over some sequences and pauses at others. The elongating complex is quite stable (RNA molecules of over 10 000 nucleotides may be synthesized), but the RNAP also terminates at specific DNA sequences, termed `transcription terminators.' Some such sequences can be recognized by the RNAP itself, but others require specific accessory proteins, called `termination factors.'
Eukaryotic RNA Polymerases RNAP I, RNAP II, and RNAP III of the eukaryotic nucleus are quite different from each other structurally and each transcribes a different set of genes (other polymerases are located in the mitochondria and chloroplasts). However, all three have two large subunits that are related to each other and also to the two largest subunits of the bacterial RNAP. In addition, several of the smaller subunits are found in common among all three of these enzymes, or only between RNAP I and RNAP III. As with the bacterial RNAPs, there are special accessory factors necessary for transcription initiation.
However, unlike the case in Bacteria, the eukaryotic initiation factors (and those of the Archaea) recognize the promoter elements independently, not as part of a polymerase holoenzyme. Many different initiation factors are involved, particularly in genes transcribed by RNAP II, and some of the initiation factors are themselves very complex proteins. Purified eukaryotic RNA polymerases, then, cannot selectively initiate transcription at promoters. The term `holoenzyme' is sometimes used to refer to a eukaryotic RNAP, but in this case it refers to something more like the bacterial `core' enzyme and would not be able to initiate from promoters. However, unlike the bacterial core enzyme, the eukaryotic holoenzyme may contain a large number of other proteins involved in transcription or the processing of RNA. RNAP I is found in the nucleolus and transcribes only genes encoding large ribosomal RNAs, the majority of cellular RNA synthesized. In yeast the enzyme has 13 subunits (and a mass of almost 600 kDa). Five of the smaller subunits are also found in yeast RNAP II and III and two others in yeast RNAP III. RNAP II transcribes genes which encode proteins, the majority of genes in a cell. It also transcribes genes encoding most of the small nuclear RNAs (snRNAs). Most organisms seem to have a 12-subunit RNAP II (with a mass of about 550 kDa). However, several other proteins are required for complete activity and the RNAP holoenzyme may have a mass of 4000 kDa. RNAP II is inhibited by the fungal toxin a-amanitin, and thus eukaryotic mRNA synthesis is sensitive to this inhibitor. RNAP III primarily transcribes genes encoding transfer RNA and 5S RNA but also transcribes some genes encoding other small RNAs. RNAP III has 14 or more distinct subunits with a mass of almost 700 kDa. Although the promoters for RNAP I and RNAP II lie for the most part upstream of the transcription start site (as is the case for prokaryotic promoters), some promoters for RNAP III lie downstream of the start site. The overall elongation complexes formed by these enzymes seem similar to those of the bacterial RNAPs. Although the mechanisms by which these enzymes locate promoters are quite different from that used by bacteria, the overall mechanism of transcriptional initiation, including abortive cycles, is very similar. Less is known about termination in eukaryotes, however. See also: Promoters; Sigma Factors; Transcription