Transcription units as RNA processing units.

Several observations made over the past year have forged strong molecular links between transcription by RNA polymerase II (Pol II) and pre-mRNA processing. The key findings support the view that the carboxy-terminal domain (CTD) of the large subunit of RNA polymerase II binds directly to protein factors essential for RNA processing. In some cases, there is a clear suggestion that the factors only bind when the polymerase is in the act of transcriptional elongation. As a result, it is now possible to discuss the transcription unit as an RNA processing unit in which the action of RNA Pol II itself brings essential elements of each of these processing steps to the nascent pre-mRNA substrate. Within this RNA processing unit, capping, splicing, and polyadenylation can also regulate one another. As pre-mRNA is synthesized, it undergoes a number of modifications that determine the protein encoded as well as the stability and translatablity of the mRNA. First, the 58 end of the pre-mRNA is modified by capping, in which the terminal phosphate is cleaved and a GMP nucleotide is linked by a 58–58 triphosphate bridge to produce GpppN. Subsequently, the G is methylated at the N7 position, and this mG constitutes the cap (Shatkin 1976). The mG cap is added shortly after transcriptional initiation within a narrow window of nucleotides (+20 to +40), indicating that the reaction is rapid and efficient (Salditt-Georgieff et al. 1980; Coppola et al. 1983; Jove and Manley 1984; Rasmussen and Lis 1993). The observation that cap-binding complex (CBC) associates with pre-mRNA early in RNA synthesis provides additional evidence that capping can occur during splicing (Visa et al. 1996b). The two subunits of CBC, CBP80 and CBP20, are bound to the cap throughout the RNA’s lifetime in the nucleus and have important roles in the identification of the first exon during pre-mRNA transcription, 38-end formation, and nuclear export (for review, see Lewis and Izuarralde 1997). The CBC dissociates from the cap in the cytoplasm, where the cap is subsequently bound by the translational regulator eIF-4E (Lewis and Izuarralde 1997). Pre-mRNA splicing is a two-step transesterification reaction that removes intron sequences and ligates exon sequences together. This reaction requires the splicing small nuclear ribonucleoprotein (snRNPs) as well as a number of non-snRNP splicing factors that assemble with the pre-mRNA to form a multiprotein complex known as the spliceosome (Moore et al. 1993). Cytological studies revealed that complexes begin to associate with proximal splice sites before distal splice sites are synthesized, and that intron removal often occurs while elongating transcripts are still attached to the DNA (Osheim et al. 1985; Beyer and Osheim 1988). The latter observation provided the first convincing evidence that splicing can occur during transcription. Cotranscriptional splicing has subsequently been confirmed in a variety of systems and has been shown to precede polyadenylation (LeMaire and Thummel 1990; Zachar et al. 1993; Bauren and Wieslander 1994; Zhang et al. 1994; Bauren et al. 1996). Polyadenylation of the 38 end of the RNA is achieved by first cleaving the pre-mRNA downstream of a consensus sequence, AAUAAA, which is the binding site of cleavage and polyadenylation specificity factor (CPSF). Cleavage stimulation factor (CstF) binds a GU-rich sequence downstream of AAUAAA, and cleavage factors I and II are both required for the cleavage event. Subsequently, the new poly(A) tail is synthesized in the nucleus and/or the cytoplasm by poly(A) polymerases (Manley 1995). Because both AAUAAA and GU-rich sequences are also required for termination of transcription by RNA Pol II (Whitelaw and Proudfoot 1986; Logan et al. 1987; Connelly and Manley 1988), polyadenylation cleavage is likely to be integrated with the process of transcription. Neither capping, splicing, nor polyadenylation occurs efficiently when transcripts are synthesized by RNA Pol I or Pol III (Smale and Tjian 1985; Sisodia et al. 1987; Gunnery and Matthews 1995), suggesting that there is something special about RNA Pol II that permits the integration of these processes. Corden (1990) proposed that the unique feature of RNA Pol II would prove to be the CTD, a domain comprised of a heptad repeat element (YSPTSPS in mouse) conserved throughout evolution but varying in the number of repeats from 26 in yeast to 52 in mouse. The CTD is dynamic with respect to its phosphorylation state during transcription. During preinitiation complex formation and initiation, the CTD is hypophosphorylated, and this form of the polymerase is known as Pol IIa (Dahmus 1996). The polymerase often pauses after synthesizing only 20–30 nucleotides, and as Corresponding author. E-MAIL [email protected]; FAX (206) 667-6503.