The triplet repeats of the Sin Nombre hantavirus 5' untranslated region are sufficient in cis for nucleocapsid-mediated translation initiation

Abstract

Hantavirus nucleocapsid protein (N) can replace the cellular cap-binding complex, eukaryotic initiation factor 4F (eIF4F), to mediate translation initiation. Although N can augment translation initiation of nonviral mRNA, initiation of viral mRNA by N is superior. All members of the Bunyaviridae family, including the species of the hantavirus genus, express either three or four primary mRNAs from their tripartite negative-sense genomes. The 5' ends of the mRNAs contain nonviral heterologous oligonucleotides that originate from endonucleolytic cleavage of cellular mRNA during the process of cap snatching. In the hantaviruses these caps terminate with a 3' G residue followed by nucleotides arising from the viral template. Further, the 5' untranslated region (UTR) of viral mRNA uniformly contains, near the 5' end, either two or three copies of the triplet repeat sequence, UAGUAG or UAGUAGUAG. Through analysis of a panel of mutants with mutations in the viral UTR, we found that the sequence GUAGUAG is sufficient for preferential N-mediated translation initiation and for high-affinity binding of N to the UTR. This heptanucleotide sequence is present in viral mRNA containing either two or three copies of the triplet repeat.

FIG. 1.

Organization of hantaviral mRNAs. The S, M, and L genome segments of SNV are displayed in 3′-to-5′ orientation. The corresponding mRNA is shown above the vRNA from which it is derived. Green circles represent cellular 5′ oligomeric caps acquired from cap snatching. The 5′ UTR sequence of each mRNA is also shown. For the sake of simplicity, three copies of the triplet repeat are shown, although some mRNAs contain two copies of the repeat.

FIG. 2.

N enhances the translational expression of RNAs with the 5′ UTR from S, M, and L mRNAs. (A) Sequences of viral and nonviral reporters. (B) Translational expression of capped reporter RNAs S n and NV gfp (left) and from S gfp and NV n (right) with increasing concentrations of N. (C) Expression from reporters L gfp (left) and NV n (right) with increasing concentrations of N. Error bars indicate standard deviations.

FIG. 3.

Time course of translation of RNA reporters with viral and nonviral UTRs in the absence or presence of N. (A) Translation of capped S n and NV gfp in the absence (left) or presence (right) of 250 nM N as a function of time. (B) Translational expression of S gfp and NV n in the absence (left) or presence (right) of N. See the Fig. 2A for descriptions of S n, NV gfp, S gfp, and NV n. Error bars indicate standard deviations.

FIG. 4.

GFP reporter RNAs containing derivatives of the 5′ UTR from S segment mRNA. Nucleotides that would be derived from the S segment template are shown in black, and the triplet repeats and the start codon are underlined. As described in the text, viral mRNA contains a nonviral oligonucleotide that arises from cap snatching. RNAs 2 through 11 all contain such a nonviral sequence (shown in gray). Triangles highlight nucleotides or regions that are deleted relative to the full-length viral UTR.

FIG. 5.

Competition-translation analysis of reporter RNAs with deletions or rearrangements in the 5′ UTR. Each of the GFP reporter RNAs shown in Fig. 4 was used in competitive translation reactions with a reporter RNA containing a nonviral 5′ UTR (NV n) (Fig. 2) and increasing concentrations of SNV N. For clarity, the data for pairs of the viral UTR reporters are displayed on five parallel graphs. The expression data for each of the plotted competitor NV n RNA used for each pair were compiled together since their values were similar. Error bars indicate standard deviations.

FIG. 6.

Binding affinity and binding stoichiometry between 5′ UTR oligoribonucleotides and N. (A and D) Representative RNA binding curves with capped RNAs corresponding to mRNA from the S segment 5′ UTR and to a nonviral RNA with increasing concentrations of N, respectively. The determined dissociation constants based on these and additional experiments are presented in Table 1. Stoichiometry was calculated using two complementary methods. (B and E) Representative least-square fit plots derived for the interaction between the UTR from the S segment and nonviral UTR, respectively. The binding stoichiometry derived from the intersection of the two pertinent lines is depicted. These calculations were verified using continuous-variation plots, which were generated based on the change in fluorescence emission with various concentrations RNA and N. (C and F) Plots for the UTR from S segment mRNA and for the nonviral UTR, respectively. Each of the UTRs in this study was similarly analyzed by both methods to determine stoichiometry. These data are summarized in Table 1.