Efficient translation initiation and optimal stability of most eukaryotic mRNAs depends on the formation of a closed loop structure and the resulting synergistic interplay between the 5′ m 7 G cap and the 3′ poly(A) tail 1,2 . Evidence of eIF4G and Pab1p interaction supports the notion of a closed loop mRNP 3 , but the mechanistic events that lead to its formation and maintenance are still unknown. Here we have used toeprinting and polysome profiling assays to delineate ribosome positioning at initiator AUG codons and ribosome:mRNA association, respectively, and find that two distinct stable (cap analog resistant) closed loop structures are formed during initiation in yeast cell-free extracts. The integrity of both forms requires the mRNA cap and poly(A) tail, as well as eIF4E, eIF4G, Pab1p, and eIF3, and is dependent on the length of both the mRNA and the poly(A) tail. Formation of the first structure requires the 48S ribosomal complex whereas the second requires an 80S ribosome and the termination factors eRF3/Sup35p and eRF1/Sup45p. Surprisingly, the involvement of the termination factors is independent of a termination event.
Keywords
Yeast; mRNA circularization; translation initiation; termination factorsIn vitro translation reactions utilized synthetic mRNAs derived from yeast transcripts, extracts that recapitulate cap/poly(A) tail synergy ( Supplementary Fig. 1a), competitive inhibition of translation initiation by m 7 GpppG (cap analog), and analyses of ribosome positioning or mRNA association by toeprinting and sucrose gradient sedimentation. Addition of the elongation inhibitor, cycloheximide (CHX), to translation reactions programmed by the 2135nt AAA and UAA mRNAs containing long or short ORFs, respectively, (Fig. 1a) allowed detection of CHX-dependent initiator AUG toeprints that reflect 80S ribosomes protecting 16-18 nt 3′ of the AUG 4-6 (Fig. 1b, top and middle panels, lanes 1 and 3). These toeprints were dependent on initiation codon recognition, the presence of yeast extract, and concurrent mRNA translation ( Supplementary Fig. 1b and c). Supporting the latter conclusion, toeprints were almost completely eliminated by 2.7mM cap analog, a concentration that distinguished bona fide toeprints from background bands (Fig. 1b, top and middle panels, lanes 2 and 4). Lower cap analog concentrations also inhibited AUG toeprint accumulation, with 70% and 96% sensitivity obtained at 0.05mM and 0.5mM, respectively (Fig. 1b, lower panel). A shorter mRNA (miniUAA1, 488nt, Fig. 1a) also yielded the AUG toeprint (Fig. 1c,upper panel, lane 1), but this band was resistant to 2.7mM cap analog (lane 2) and only manifested sensitivity at higher concentrations (Fig. 1c, lower panel). Thus, in wild-type extracts, the short capped (see Supplementary Fig. 1d) and polyadenylated miniUAA1 mRNA is ∼160-fold more resistant to cap analog than the longer AAA mRNA. The miniADE2 (485nt) and ADE2 (2070nt) mRNAs, whose respective sizes (but not sequences) are comparable to those of the miniUAA1 and AAA transcripts, also...