Eukaryotic mRNA 3′ processing: a common means to different ends

In principle, the formation of the 3′ end of a eukaryotic mRNA is a simple process. At a minimum, it involves the hydrolysis of a single phosphodiester bond in the nascent transcript. For one unique class of mRNAs, transcripts of the metazoan replication-dependent histone genes, this is indeed the only requirement. All other eukaryotic mRNAs require both cleavage and subsequent poly(A) addition to the newly generated 3′ hydroxyl. Despite this apparent simplicity, and more than two decades of work, basic questions concerning the mechanisms of 3′-end formation for both classes of transcripts persist. The accumulated evidence indicates that the mechanisms by which the ends of histone and polyadenylated mRNA are formed are fundamentally distinct, both in the factors responsible for processing and the RNA sequence elements that direct their action (for review, see Marzluff 2005; Zhao et al. 1999a). The 3′ processing of polyadenylated mRNAs requires a complex of over a dozen proteins, nearly all of which are conserved from yeast to humans. In contrast, the formation of the 3′ ends of the metazoan replication-dependent histone transcripts not only requires a unique set of proteins, but an essential snRNA component as well. In both cases, the identity of the endonuclease has remained largely a matter of speculation.

Two reports—one from Kolev and Steitz (2005) in this issue of Genes & Development, and the second from Dominski et al. (2005a) have revealed that the mechanisms of poly(A) site processing and histone 3′ processing are not quite as unique as they appear. Surprisingly, the two processing machines share a common core of proteins that almost certainly includes the endonuclease. The most exciting aspect of these reports, however, is that they provide a fresh insight into the evolution of what had seemed to be needlessly redundant mechanisms for generating mRNA 3′ …