Exonucleases and endonucleases involved in polyadenylation‐ assisted RNA decay

Poly(A)‐assisted RNA decay occurs in prokaryotes/organelles (left) and the nuclei and cytoplasm of eukaryotic cells (right). Generally, the process can be divided into three steps, as presented in the figure: (1) endonucleolytic cleavage (not always obligatory), (2) adenylation, and (3) 3′ → 5′ exonucleolytic digestion. In Escherichia coli, removal of PPi from the 5′ end by RppH usually precedes endo‐cleavage by RNase E. RNase J has been implicated in a similar, RNase E‐like function in Bacillus subtilis. In the nuclei of eukaryotic cells, endo‐cleavage by the PIN domain of Rrp44 can prepare the substrate for the second step—adenylation by Trf4/5 of the TRAMP complex (Air1/2, Trf4/5, Mtr4). In E. coli, polyadenylation can be performed by Ntr‐PAP, producing homopolymeric tails or by PNPase, producing heteropolymeric poly(A)‐rich tails, which are assumed to fulfill a similar function. (In hyperthermophilic and several methanogenic archaea, heteropolymeric tails are believed to be synthesized by the archaeal exosome.) The 3′ → 5′ exonucleolytic degradation step is carried out by PNPase and RNase II/R in bacteria and organelles. The 3′ → 5′ stage that occurs in the nuclei of eukaryotic cells is fulfilled by the exosome (particularly, the two catalytic subunits associated with it, Rrp44 and Rrp6). Note that 5′ → 3′ degradation events are not included in this depiction.

Suggested scenario for the evolution of polyadenylation. When comparing the various roles and nucleotide compositions of poly(A) tails in different systems, as well as the structures and activities of the polyadenylating and degrading exoribonucleases, the following scenario can be offered: A single‐domain PH‐like ancestor enzyme developed into the PNPase and archaeal exosome complexes which produced heteropolymeric extensions—poly(A)‐rich in nature, because despite the non‐specific polymerization properties of these enzymes, adenosine triphosphate (ATP) was the cell's ‘energy currency’ and therefore, present at a relatively high concentration. The function of these transient (unstable) tails was to assist in RNA decay, which the same complexes carried out. Later, enzymes specific to ATP emerged and the unstable homopolymeric poly(A) tails that they produced assisted in RNA decay as well. At the next stage, cellular mechanisms adopted new roles for the homopolymeric tails which required their stability. In some cases the stable tails influenced translation initiation and longevity of the tailed transcript (nucleus) and in others (animal mitochondrion) stable poly(A) tails helped to complete partially encoded stop codons. Today, tails can be categorized as hetero‐ or homopolymeric, wherein the former is unstable and promotes RNA decay (although additional function cannot be ruled out). The latter can be both unstable (transient) and stable, depending on the system/process and in some cases, stable poly(A) tails can encourage RNA stability. In summary, the reason that nucleus‐encoded mRNAs bear homopolymeric stable adenosine tails and that adenosines dominate transient poly(A) extensions is linked to the reason that ‘A’ was chosen as life's ‘energy coin’.

RNase E/G proteins. RNase E, first discovered in E. coli, has homologs in many other biological systems, displayed here with different grayscale patterns representing the domain types. As evident, amino acid alignment reveals a high level of homology shared between the RNase E/G‐like enzymes. The E. coli RNase E contains a C‐terminal domain which is not present in RNase G or the other RNase E homologs. This domain is the scaffold for the degradosome complex, which includes an RNA helicase, enolase, and part of the PNPase population. In plants, RNase E bears an N‐terminal chloroplast target peptide and a stretch of ∼120 amino acids within the S1 domain, which is not present in any bacterial sequence.

RNase PH, the bacterial and organellar PNPases, and the archaeal and eukaryotic exosome cores share a similar ring‐like structure. Phylogenetic analysis and structural comparison lead to an evolutionary model wherein a single‐domain phosphorolytic primordial enzyme developed into these enzymes and complex subunits. Ultimately, a common ring‐like structure, employing a central channel to lead substrate RNA to the catalytic site within, evolved. The yeast and human exosome cores apparently lost their phosphorolytic capabilities and rely on association with hydrolytically active enzymes, Rrp44 and Rrp6 (not shown), in order to digest RNA. A cap structure (not shown) provides RNA binding and selective properties. (Reprinted with permission from Ref 19. Copyright 2008 Elsevier.).

The ‘zipper’ model of phosphorolytic polymerization and degradation. PNPase and the archaeal exosome are bidirectional; they can synthesize heteropolymeric tails and phosphorolytically degrade RNA from 3′ → 5′, as well. These two activities are modulated by Pi and NDP concentrations, which are influenced by one another and by the enzyme's activity: ↑Pi/↓NDP = degradation. ↑ NDP/↓Pi = polymerization. (Reprinted with permission from Ref 19. Copyright 2008 Elsevier.).