actor 'TATA-box binding protein' (TBP) regardless of whether the promoters have a TATA box or not. In addition, modules and domains present in general transcription factors (GTFs) specific for RNAPII are found in subunits of RNAPI and RNAPIII and/or RNAPI and RNAPIII specific factors. The initiation, elongation and termination of transcription are steps shared by all systems, albeit with differing details and/or degrees of complexity. Beyond the shared or related elements, specific factors confer unique topological and functional features to the particular initiation complex associated with the three polymerases and account for the differences in promoter recognition and the class or category of transcribed genes. Pol I transcribes ribosomal RNA, all protein-coding and many non-coding genes are transcribed by pol II, while transfer RNA, the small 5S ribosomal RNA and a subset of small non-coding genes are the realm of pol III-mediated transcription. Transcription by RNA pol II is described here.

RNA polymerase II transcription pathway
Transcription Initiation Pol II transcribes all protein-coding genes and many non-coding genes classes. Assembly of the pre-initiation complex (PIC) sets the stage for the recruitment of pol II, transcription start site selection and initiation of transcription. PIC involves the sequential assembly of TBP-containing TFIID complex, TFIIA and TFIIB, Pol II and TFIIF, TFIIE and TFIIH. TFIID complex is composed of TATA-box binding protein (TBP) and TBP associated factors (TAFs); it constitutes a scaffold for the assembly of other transcription factors. Topologically, it has a horse-shoe structure consisting of three lobes with TFIIA and TFIIB bound on opposite sides of the central cavity. TBP bends DNA by ~90 degrees which wraps DNA around the polymerase and probably accounts for its universal requirement. ~80% of the protein-coding genes do not have a TATA box and the transcription start site (TSS) may be spread over tens of base pairs. Promoter elements include the Initiator (Inr) surrounding the start site on both TATA and TATA-less promoters, the downstream promoter element (DPE) and downstream core element (DCE) on TATA-less and TATA-containing promoters, respectively. The presence of numerous TBP-associated factors (TAFs) allows for recognition of other regulatory elements. TAFs interact with GTFs, with activators, or transcription and signaling factors. They can form dimers and heterodimers, the latter among histone domain-containing TAFs. TFIIA binds the TBP-DNA complex, stabilizes it and blocks certain regulators. TFIIB bridges TBP and pol II; all its four functional domains interact with pol II. Pol II, in complex with TFIIF, assembles at the promoter, followed by TFIIE which recruits the TFIIH complex. TFIIH consists of a core with 3'-5' and 5'-3' helicase activities linked to a kinase module; ERCC5 known as XPG with 3'-endonuclase activity is also considered part of the complex. TFIIH plays essential roles in both pol II transcription and in DNA nucleotide excision repair (NER) pathways. In pol II transcription, it promotes ATP-dependent unwinding of ~10 base pairs of promoter DNA and phosphorylation of Ser5 within the carboxyl-terminal domain (CTD) of pol II. Pol II is unique among the three polymerases in its requirement for ATP-dependent promoter melting and the presence of CTD in its largest subunit [see CTD entry further down]. Unwinding of ~15 bases within the promoter forms the 'transcription bubble' and marks the transition of PIC from a closed to an open conformation (OC) state. Once OC is formed, TSS selection occurs followed by RNA synthesis. The Mediator complex, a coactivator required for transcription of most protein-coding genes that bridges regulatory and promoter complexes, likely stabilizes PIC via interactions with many of its components. It interacts with CTD and the rest of pol II subunits; in both yeast and human, pol II contacts the same region of Mediator. SAGA, a multisubunit coactivator, better characterized in yeast, may have functions redundant with TFIID, particularly at TATA promoters [see Mediator and SAGA in Complexes entry, further down].

Transcription Elongation
Transcription elongation is a reiterative process whereby pre-mRNA is 5' extended by 1 nucleotide (nt) and the polymerase translocates 1 nt along DNA. It involves a condensation reaction, thought to occur via a general two-metal-ion mechanism. Nascent RNA is annealed to the template strand forming an RNA-DNA hybrid of ~8-9 nts whose size is maintained throughout elongation; as RNA is 3' extended by 1 nt, the 5' end of the hybrid peels off the template and is threaded to the exit channel. Transition from promoter escape to full productive elongation is often accompanied by an intermediate promoter-proximal pausing. Once the initial transcript reaches 20-60 nucleotides, pol II pauses between the promoter and the 1+ nt, backtracks and is also prone to arrest. Paused pol II is recognized by pausing factors such as DSIF (DRB sensitivity-inducing factor) and NELF (negative elongation factor) complex. Arrested or paused pol II is rescued by TFIIS; TFIIF which acts as an elongation factor, may synergize with TFIIS. DSIF and NELF interact with RNA and pol II, Once elongation starts, some GTFs, TFIIB in particular, dissociate; others, such as TFIIF, remain associated with pol II. Productive elongation is promoted by the 5' capping of nascent RNA and phosphorylation of pol II. Capping consists of three enzymatic reactions of which the triphosphatase and guanylyltransferase reside on the same polypeptide in higher eukaryotes, the Rngtt gene product; a guanine-7-methyltransferase (Rnmt) adds the methyl group. Rngtt interacts with DSIF and the methyl transferase complex. Nuclear cap-binding complex (CBC), binds 5'-capped RNA, stimulates elongation and is involved in a range of other events, including splicing where it may act as a recruitment platform. Elongation-related entities are shown. Phosphorylation at Ser2 of CTD is carried out by P-TEFb (positive elongation factor b), whose kinase activity also targets DSIF and NELF. P-TEFb is a heterodimer composed of cyclin-dependent kinase 9 (CDK9) and cyclin T or, in humans, cyclin K. Several factors contribute to the recruitment of the P-TEFb complex, including CBC or the Med26 subunit of the Mediator complex. In the latter case, P-TEFb is a component of SECs (Super Elongation Complexes). Paf1 and FACT complexes among others, promote pol II transcription through nucleosomes and the gene body. Paf1 recruits SEC, FACT and histone- modifying enzymes. Chromatin and co-activator components are readers of modified CTD. Splicing occurs co-transcriptionally and the coupling between the splicing machinery and the elongating Pol II is likely mediated by the CTD domain [see Mediator and SEC in Complexes entry, and CTD entry, further down].

Transcription Termination
The termination step of pol II is mechanistically, probably the least understood one. Transcription termination occurs when the polymerase stops synthesizing RNA and both the enzyme and the nascent RNA are released, and it is coupled with pre-mRNA 3' end processing which consists of 3'-end cleavage and subsequent polyadenylation. 3'-end processing is at the crossroads of gene expression connecting transcription, splicing, mRNA biogenesis and mRNA export, and translation. Cleavage entities and events are presented/shown. The highly conserved A(A/U)UAAA element is a primary poly(A) signal located ~10-30 nts upstream of the cleavage site, followed by the U/GU-rich region ~30 nts downstream; the upstream UGUA rich motifs also play a role. The processing machinery consists of several complexes that include the cleavage and polyadenylation specificity factor (CPSF), the cleavage stimulation factor (CstF), the cleavage factors I and II (CF Im and CF IIm), and the poly(A) polymerases (PAPs) and associated proteins. CPSF recognizes the primary poly(A) element, recruits other complexes and catalyzes the cleavage reaction. CstF binds the downstream U/GU element and requires cooperative binding of CPSF to the upstream poly(A) signal. In addition to CPSF, CstF interacts with pol II CTD and CF IIm. CF Im binds the upstream UGUA motif and plays a role in the cleavage step. CF IIm interacts with pol II CTD and CstF, its exact function is not well known. The uncapped 5' downstream sequence generated by cleavage is a substrate for the exoribonuclease Xrn2, which in complex with Dxo, known as Dom3z, carries out its 5'-3' digestion. This activity is proposed to promote termination and is described as the 'torpedo' model. Alternatively, pol II-bound CF IIm may also promote termination in the 'allosteric' or 'anti-terminator' model. Released pol II needs to revert to its hypo-phosphorylated form in order to initiate another round of transcription. Terminal and promoter regions can interact via gene looping, enabling contacts between GTFs and 3'-end processing factors and leading to rapid pol II recycling and re-initiation of transcription.

Pol II carboxyl-terminal domain (CTD)
The CTD domain of Polr2a, the largest pol II subunit, contains multiple tandem repeats of a heptapeptide, with the consensus YSPTSPS sequence. The number of repeats appears to correlate with the complexity of the system - there are 26 in yeast but 52 in mammals, of which 21 are consensus and 31 are non-consensus. Via its many interactions with modification readers, CTD is involved in all steps of pol II transcription and associated pre-mRNA processing, splicing and chromatin modulation. The serine, threonine and tyrosine residues can all be phosphorylated in vivo and may undergo other post-translational modifications; non-consensus repeats can also be modified (for instance arginine methylation). The phosphorylation of serines is the best characterized modification. Ser5-(P)hosphorylation is primarily associated with the initiation step, Ser2-P with elongation while Ser7-P appears to be throughout the transcription cycle. The enzymes involved include: Cdk7 and Cdk8 kinases for Ser5, Cdk9 and Cdk12 kinases for Ser2 and Cdk7 and Cdk9 kinases for Ser7. Cdk7 is a subunit of TFIIH; Cdk8 is a subunit of the kinase module of Mediator; Cdk9 is a subunit of P-TEFb. Ser5-P is a substrate for Rpap2, Ctdsp1, Cdc14 and Ssu72 phosphatases, Ser2-P is a substrate for Ctdp1 and Cdc14 phosphatases and Ser7-P is a substrate for Ssu72. The cis-trans isomerization of the two prolines has also been reported and is dependent upon Pin1 isomerase. Many protein domains have been identified as readers of the CTD code, particularly in yeast; the precise function of these interactions remains to be characterized. The existence of multiple modification marks within a repeat and of many repeats within CTD, render the number of possible combinations extraordinarily large. This, combined with the dynamic, yet highly orchestrated nature of changes, makes unraveling the complexity of the CTD code an indomitable task.

Complexes: Mediator, SAGA and SEC
The Mediator is a large ~30 subunit complex largely conserved throughout evolution. It consists of several modules described as the 'head', 'middle' and 'tail' and an additional kinase module; in some isoform complexes, the kinase module and/or certain subunits are not present. Mediator interacts with elements of the core pol II machinery or modifies them via the Cdk8 activity of the kinase module. It interacts with many transcriptional regulators via an activator-binding domain (ABD) and the transactivation domains (TADs) in regulators; in addition, it interacts with epigenetic regulators, long non-coding RNA, other factors and/or large complexes. It is associated with all phases of pol II transcription as well as with DNA loop formation, chromatin modulation, RNA processing. Subunits can undergo post-translational modifications that further regulate their function and some may function as individual entities. The structures of many subunits have been solved; a cryo-EM structure of the yeast complex is available but solving a more refined structure is challenging both by virtue of its size and by its high degree of flexibility, conferred by many disordered regions. Additional subunits, present in vertebrates, are duplicates of Cdk8. Their function has not been well studied, existing data suggest that their roles are not redundant.
The SAGA coactivator appears to have functions redundant to TFIID in PIC assembly, particularly at TATA-containing promoters; TFIID appears to be associated with the TATA-less and the more abundant of promoter types. The proteins that make up the complexes have been largely conserved from yeast to man and they include a subset of TAFs also present in TFIID. SAGA modules carry out distinct functions such as ubiquitination, histone acetyltransferase, transcription activation and TBP binding; non-transcriptional functions include mRNA transport, maintenance of DNA integrity. Mammalian systems exhibit four SAGA-like subcomplexes: SAGA itself or ATAC, both depending on the presence of either Kat2a (Gcn5) or the related Kat2b (Pcaf) histone lysine acetyltransferases and each present in two versions. The two depend on the presence of Tada2b (Ada2b), exclusively in SAGA or Tada2a (Ada2a), exclusively in ATAC type complexes. The yeast Ada2 in higher metazoans has been duplicated into TADA2A and TADA2B.
The Super Elongation Complex (SEC) is recruited by coactivators such as the Mediator complex and in this context, delivers the P-TEFb complex to the elongation complex. In mammals, several SEC complexes are present resulting from combinations of AF4/FMR2 (Aff1 or 4), ELL (elongation factor - Ell1, 2 or 3), EAF (ELL-associated factor - Eaf1 or 2) family members, and Mllt1, known as ENL (eleven-nineteen leukemia) or Mllt3, known as AF9. Aff4 acts as a scaffold for assembling SEC via interactions with Mllt1, Mllt3, Ell and P-TEFb. To see the ontology report for annotations, GViewer and download, click here

...(less)