The three states of the RNA polymerase II elongation complex. The dashed oval represents the empty nucleotide addition site.
RNA polymerases catalyze rapid RNA chain growth. RNA polymerase II adds between 20 and 70 nucleotides per second to RNA and does this by moving forwards and backwards over at template at each step. Transcribing RNA polymerases swap between three states: pre-translocated, reverse-translocated and post-translocated. Structures of the pre- and post-translocated states are available, but that of the intermediate step – reverse-translocated or 'backtracked' – has only just been solved and reveals why it is so good at proofreading RNA during transcription.
During pre-translocation, the nucleotide that has just been added is still within the nucleotide addition site of RNA polymerase II. During post-translocation, RNA polymerase II has a free nucleotide addition site that is available to new nucleotides. In addition to these two states, there is the backtracked state, when RNA polymerase II moves backwards over the template, and the 3′ end of the new RNA is extruded. The advantage of backtracking is that it allows the RNA transcript to be checked and is the dominant state when the template is damaged.
Wang et al. 1 used two approaches to produce the backtracked complex. They used RNA–DNA hybrids with mismatched nucleotides at the 3′ ends of the RNA and bound it to polymerase II, thus directly creating the backtracked state. They also used DNA–RNA hybrids with damaged DNA, presuming that the enzyme encountering the damage would retreat to the backtracked state. They found that the structures of both complexes were very similar.
One of the structures solved by Wang et al. is of a hybrid containing one mismatched residue at the 3′ end of the RNA. The authors found that the last correctly matched residue was positioned within the nucleotide addition site and that the mismatched residue is located two residues downstream at a site the authors call 'P', for proofreading. The mismatched residue's interaction with RNA polymerase II distorts the RNA–DNA helix, perhaps producing a structure that is poised for intrinsic cleavage.
With two mismatched residues, the last matched residue and the first mismatched residues are in a location similar to that seen with one mismatched residue. But the second mismatched residues could not be seen in the structure, presumably because they are highly mobile.
Overall, their results lead to two conclusions. The first is that RNA polymerase II backtracked by one residue is stable. This supports the idea that there is an equilibrium between forward and backward motion during transcription. It also confirms that backtracking by one residue is favorable whereas going back over several residues is not.
The second conclusion is that the one-residue backtracked state is readily cleaved in the presence of the elongation factor IIS (TFIIS) and that a dinucleotide is released. This adds weight to the theory that cleavage occurs in the RNA polymerase II active site and that it is important for removal of misincorporated nucleotides.
RNA polymerase II moves forwards and backwards on the template until a mismatch of RNA–DNA causes the helix to distort and shift the polymerase into the backtracked state. If it remains in this state for a long time, cleavage ensues. This one-residue backtracked state is a key contributor to proofreading by RNA polymerase II.