The trimeric translation initiation factor a/eIF2 of the crenarchaeon Sulfolobus solfataricus is pivotal for binding of initiator tRNA to the ribosome. Here, we present in vitro and in vivo evidence that the a/eIF2 ␥-subunit exhibits an additional function with resemblance to the eukaryotic cap-complex. It binds to the 5-triphosphate end of mRNA and protects the 5 part from degradation. This unprecedented capacity of the archaeal initiation factor further indicates that 5 3 3 directional mRNA decay is a pathway common to all domains of life.
In Escherichia coli, the decay of most RNA transcripts appears to be initiated by 5Ј pyrophosphate removal (1), followed by endonucleolytic cleavages that are generated by RNase E (2, 3). The intermediate cleavage products are further degraded by the 3Ј 3 5Ј exonucleases polynucleotide phosphorylase (PNPase), RNase II, and oligoribonuclease, converting the decay intermediates into poly-and mononucleotides (4). At variance with E. coli, Bacillus subtilis possesses a 5Ј 3 3Ј exonuclease activity, which could explain why RNAs in this organism are stabilized for great distances downstream of stable secondary structures or bound ribosomes (5). In eukaryotes, mRNA decay is mainly catalyzed by exonucleases (6). Eukaryotic mRNAs generally have a 7-methylguanosine cap at their 5Ј end and a poly(A) tail at their 3Ј end. Removal of these terminal modifications is considered rate-limiting for mRNA decay (7). Translation initiation factor eIF4E binds to the 7-methylguanosine cap and thereby protects the cap structure from the decapping enzyme and consequently the mRNA from 5Ј 3 3Ј exonucleolytic decay (7). Different RNases with either endo-or exonuclease activity (8-11) have been inferred or described in Archaea. In S. solfataricus a 3Ј 3 5Ј directional decay by a multisubunit exosome complex has been demonstrated in vitro (10). In addition, the Sulfolobus solfataricus exosome is able to polyadenylate the 3Ј end of RNAs in the presence of ADP (12). With the exception of the exosome, no other endo-or exonuclease activities have been described in S. solfataricus. At variance with a previous study wherein the longevity of selected S. solfataricus mRNAs was found to be rather high (13), a recent microarraybased analysis (14) indicated a rather short mRNA half-life, comparable with that of bacterial mRNAs.Two different mechanisms for translational initiation seem to exist in Sulfolobus (15). One is based on a canonical SD/anti-SD interaction and operates on internal cistrons of polycistronic mRNAs. In contrast, monocistronic mRNAs and proximal genes of polycistronic mRNAs are frequently devoid of a 5Ј untranslated region. Decoding of these leaderless mRNAs requires, analogously to Bacteria (16), pairing of the start codon with initiator-tRNA (15). The complexity of archaeal translational initiation seems to be underscored by the presence of a largerthan-bacterial set of factors because Archaea encode Ϸ10 orthologs of eukaryal and bacterial initiation factors (17). Like its eukaryotic counterp...