The mechanism of transcription initiation begins when the TBP domain of TFIID binds the TATA box. The TBP then bends the TATA box DNA to facilitate the CTD of TFIIB to bind to the DNA. The CTD of TFIIB binds sequences both upstream and downstream of the TATA box, and the NTD binds to the dock domain, saddle, and active center cleft of Pol II. The promoter DNA is moved toward and above the active center of Pol II, passing the Tfg2 subunit of TFIIF and passing TFIIE and TFIIH at the far end of the cleft, this is all performed by interactions with the TFs, without binding to the Pol II. TFIIE interacts with the Pol II-TFIIB complex to facilitate TFIIH helicase activity (3).
The helicase activity opens the DNA, and the nontemplate strand binds to Tfg2. The template strand binds to the TFIIB finger domain (3). The clamp region of Pol II, which is formed by the NTD and CTD regions of Rpb1 and the CTD region of Rpb2, is guided into the closed state by three zinc ions and five “switches”, and swings about 30o, and moving up to 30 Angstroms, to facilitate Pol II binding to the template strand (2). The clamp is held in the closed state by interactions with the Rpb4/Rpb7 heterodimer (4). This action opens up an RNA binding groove. The downstream DNA then interacts with a proline residue in the jaw domain of Rpb5, and then passes between the lobe region of Rpb2 and the clamp head region of Rpb1. The DNA now moves over the bridge helix into the active site. The DNA at the +1 site in the active site is rotated 90o away from the other bases on the strand (2). The DNA also interacts with the rudder and zipper domains that form the transcription bubble at the upstream end, and with fork loop 2 at the downstream end . The enzyme is now ready for transcribing the first ribonucleotides.
To start adding the nucleotides, the transcription bubble must be maintained. The enzyme would clash if the +3 nucleotide of the nontemplate strand was still paired with the template strand. To alleviate this, a Tyr at residue 836 of Rpb1 helps align the DNA at the +2, +3, and +4 sites. The K1109 and N1110 residues of Rpb1 fray the end of the downstream DNA at the +5 site. Rpb2 fork loop 2 also helps maintain the downstream end of the bubble (2). Maintaining the bubble upstream is done by fork loop 1, which also unwinds the RNA-DNA hybrid, which is discussed later.
RNA-DNA hybrids of up to 9 base pairs can be stabilized in the complex before competition, by means of steric hinderance between the RNA and the zinc finger domain of TFIIB, begins with the NTD of TFIIB. A pause happens as a strand/loop network forms, which can displace TFIIB and allow the enzyme to continue into transcription (7). If TFIIB is not released from the initiation complex, the RNA-DNA hybrid dissociates along with the entire initiation complex as an abortive transcript. If TFIIB is released from the initiation complex, RNA transcript continues into elongation, and the entire initiation complex dissociates (3).
The active site consists of one magnesium ion, bound by three aspartate residues at 481, 483, and 485. A second metal ion is required for transcription to occur, and it enters the enzyme coupled with an NTP (6). The second metal ion, another magnesium ion, is located 5.8 Angstroms away from the first metal ion. The second metal ion is located near three acidic residues, D481 in Rpb1, and E836 and D837 in Rpb2 (1). The NTP enters the polymerase through pore 1 and binds to the E site of the active site. The NTP rotates around the second metal ion to sample base pairing in the A site. If the NTP is mismatched, the enzyme only interacts with the protein residues and active site metals and does not change into the A state and form a base pair with the template strand. If the NTP does match the desired nucleotide, the NTP base pairs with the template strand at the +1 residue, and forms a phosphodiester bond with the growing transcript strand. This process also can prove useful for discriminating between rNTPs and dNTPs (6).
After base pairing and phosphodiester bond formation, translocation is necessary to move to elongate the RNA strand. The bridge helix is necessary for translocation in all organisms, but in prokaryotes it is bent and it is straight in eukaryotes. It seems likely that the bridge helix bends and translocates the DNA strand to continue elongation (1).
Backtracking is an important process in transcription. It is used for proofreading and avoiding obstacles in the template strand. The polymerase moves backwards along the template strand, and the hybrid moves like a zipper. The 3’ end of the transcript becomes single stranded and the 5’ end rehybridizes with the template strand. Sometimes cleavage of the RNA transcript is necessary, whether due to a mistake in transcription, or an arrested complex due to overbacktracking. This is performed by TFIIS, which enters through pore 1. TFIIS has a zinc-binding domain with a beta-hairpin at one end. This hairpin contains two residues that reach the active site and are critical for RNA cleavage. Two more residues bind the nucleic acid and can bring the extruded RNA down the pore bound to the TFIIS (5).
The RNA-DNA hybrid is maintained from the +1 position to the -5 position in the transcription bubble. Three domains of Rpb1 are responsible for separating the RNA from the DNA, the lid, the rudder, and fork loop 1. The lid wedges the DNA and RNA apart at the -8 through -10 residues of RNA. Phe252 splits the hybrid at the -10 position, using its aromaticity to contact the DNA base. Phe264 of Rpb1 similarly works on DNA base -10 or -11. The rudder does not split the strand, but rather interacts with DNA at bases -9 to -11 to prevent reassociation with RNA. S318 and R320 contact the sugar and 5’ phosphate at -10. The rudder also stabilizes unwound DNA after the hybrid region. Fork loop 1 is located on Rpb2, and interacts with positions -5 to -7 on the RNA. Rpb2 residues K471 and R476 contact RNA phosphates (7).
The lid is also essential in RNA exit. RNA exits through an arch formed over the saddle by the interactions between the lid and protein elements. The DNA exits above the arch and RNA exits below the arch, preventing the reformation of the hybrid. RNA exits through a positively charged groove leading towards Rpb8 (7). After RNA exit begins and the transcript is outside the polymerase, it is subject to attack by nuclease. So capping and other processing enzymes have an affinity for the phosphorylated CTD of Rpb1, which begin processing as soon as the transcript is available. While the 5’ end of the transcript exits the polymerase at the exit groove, the 3’ end is “extruded during retrograde movement of the enzyme” (1).
Termination occurs when a combination of events happens on the 3' end of the DNA. Poly (A) signals can promote termination, but is most effective when coupled with a termination element. There are two different mechanisms for termination element-mediated termination in eukaryotic cells (11).
The first involves a pause domain. This domain, which is located downstream from the poly(A) site, causes a protein to bind to the corresponding transcript, pausing or slowing elongation. This allows the CTD of the Rpb1 subunit to cleave the transcript upstream of the poly(A) site transcript by TFIIS, followed by polyadenylation. The second mechanism involves a cotranscriptional cleavage site (CoTC). This site on the transcript is recognized by the CTD of Rpb1, with cleavage downstream of the poly(A) site transcript by TFIIS, followed by polyadenylation (11).
After cleavage both transcripts leave a free 5' end attached to the polymerase. This free 5' end is subject to 5'-3' exonuclease activity, which reaches the polymerase and leads to termination and freeing of the polymerase from the strand. This process is more efficient with the pause site because the polymerase is moving very slowly, while the CoTC site is moving normally (or much faster) (11).