Endonucleolytic cleavage separates the pre-mRNA into an upstream fragment destined to become the mature mRNA, and a downstream fragment that is rapidly degraded. Cleavage depends on two signals in the RNA, a highly conserved hexanucleotide, AAUAAA, 10 to 30 nucleotides upstream of the cleavage site, and a poorly conserved GU- or U-rich downstream element. Additional sequences, often upstream of AAUAAA, can enhance the efficiency of the reaction. Cleavage occurs most often after a CA dinucleotide. A single gene can have more than one 3' processing site.
Cleavage is preceded by the assembly of a large processing complex, the composition of which is poorly defined. ATP, but not its hydrolysis, is required for assembly. Cleavage at the 3'-end of mRNAs depends on a number of protein factors. CPSF, a heterotetramer, binds specifically to the AAUAAA sequence. The heterotrimer CstF binds the downstream element. CF I, which appears to be composed of two subunits, one of several related larger polypeptides and a common smaller one, also binds RNA, but with unknown specificity. RNA recognition by these proteins is cooperative. Cleavage also requires CF II, composed of at least two subunits, and poly(A) polymerase, the enzyme synthesizing the poly(A) tail in the second step of the reaction. The polypeptide catalyzing the hydrolysis of the phosphodiester bond remains to be identified.
Cleavage produces a 3'-OH on the upstream fragment and a 5'-phosphate on the downstream fragment. At some unknown point after cleavage, the downstream RNA fragment, CstF, CF I and CF II are thought to be released, whereas CPSF and poly(A) polymerase remain to carry out polyadenylation.