ARC-1, a sequence element complementary to an internal 18S rRNA segment, enhances translation efficiency in plants when present in the leader or intercistronic region of mRNAs

Abstract

The sequences of different plant viral leaders with known translation enhancer ability show partial complementarity to the central region of 18S rRNA. Such complementarity might serve as a means to attract 40S ribosomal subunits and explain in part the translation-enhancing property of these sequences. To verify this notion, we designed beta-glucuronidase (GUS) mRNAs differing only in the nature of 10 nt inserts in the center of their 41 base leaders. These were complementary to consecutive domains of plant 18S rRNA. Sucrose gradient analysis revealed that leaders with inserts complementary to regions 1105-1114 and 1115-1124 ('ARC-1') of plant 18S rRNA bound most efficiently to the 40S ribosomal subunit after dissociation from 80S ribosomes under conditions of high ionic strength, a treatment known to remove translation initiation factors. Using wheat germ cell-free extracts, we could demonstrate that mRNAs with these leaders were translated more than three times more efficiently than a control lacking such a complementarity. Three linked copies of the insert enhanced translation of reporter mRNA to levels comparable with those directed by the natural translation enhancing leaders of tobacco mosaic virus and potato virus Y RNAs. Moreover, inserting the same leaders as intercistronic sequences in dicistronic mRNAs substantially increased translation of the second cistron, thereby revealing internal ribosome entry site activity. Thus, for plant systems, the complementary interaction between mRNA leader and the central region of 18S rRNA allows cap-independent binding of mRNA to the 43S pre-initiation complex without assistance of translation initiation factors.

Figure 1

Secondary structure model of rice (Oryza sativa) 18S rRNA (A) with an enlargement of the central domain (B). Numbering of nucleotides in rice 18S rRNA is according to the sequence with accession no. X00755. (C) Primary structure of rice 18S rRNA segment from nucleotides 1073 to 1134 (in antisense orientation) divided into six consecutive 10 base subsegments. (D) Schematic structure of mRNAs containing in their leaders 10 base inserts (bold and underlined) complementary to adjacent consecutive segments of 18S rRNA from nucleotides 1073 to 1134. Arrows indicate positions of restriction enzymes used for insertion (HindIII) and for linearization (NcoI or EcoRI). 3′-TMV corresponds to the TMV 3′-UTR present in all mRNAs tested. GUS is the β-glucuronidase coding region, with the AUG as shown.

Figure 2

Electrophoresis of products of RT–PCR. WG 40S ribosomal subunits were used as a template. Lane 1, RT–PCR from primers AS(1115–1124) and (1043–1052) without reverse transcriptase; lane 2, RT–PCR from primers AS(1115–1124) and (1043–1052); lane 3, RT–PCR from primers AS(1115–1124) and (1009–1018); lane 4, RT–PCR from primers AS(1115–1124) and (923–932).

Figure 3

Ability of different radioactively labeled leaders to bind to 40S ribosomal subunits dissociated from 80S ribosomes under high ionic strength conditions. A, plasmid polylinker-derived leader; B, ARC-1; C, 2xARC-1; D, PVY viral genomic RNA leader; E, 3xARC-1. Preparation of 40S subunits and sucrose gradient analysis was performed as described in Materials and Methods. The arrow indicates the position of the 40S subunit.

Figure 4

Autoradiogram of an SDS–polyacrylamide gel after electrophoresis of translational products synthesized in a WG cell-free system incubated for 1 h at 26°C (odd lanes) or 37°C (even lanes). Translated mRNAs: lanes 1 and 2, pl-GUS; lanes 3 and 4, [1073–1082]-GUS; lanes 5 and 6, ARC-1-GUS; lanes 7 and 8, 2xARC-1-GUS.

Figure 5

Northern blot hybridization of total RNA purified from O.violaceus protoplasts at different time intervals after transfection with dicistronic (lanes 1–4) and monocistronic (lanes 5–7) constructs and probed with a ³²P-labeled RNA complementary to the CAT gene. Total RNA was extracted at 2 (lanes 1 and 2) and 4 h (lanes 3–7) post-transfection. Constructs used for transfection: lanes 1 and 3, dicistronic ‘pl-GUS-4x[1083–1092]-CAT’; lanes 2 and 4, dicistronic ‘pl-GUS-4xARC-1-CAT’; lane 5, monocistronic ‘Y-CAT’; lane 6, monocistronic ‘4x[1083–1092]-CAT’; lane 7, monocistronic ‘4xARC-1-CAT’.