Rous Sarcoma Virus Genomic RNA Dimerization Capability In Vitro Is Not a Prerequisite for Viral Infectivity

Abstract

Retroviruses package their full-length, dimeric genomic RNA (gRNA) via specific interactions between the Gag polyprotein and a "Ψ" packaging signal located in the gRNA 5'-UTR. Rous sarcoma virus (RSV) gRNA has a contiguous, well-defined Ψ element, that directs the packaging of heterologous RNAs efficiently. The simplicity of RSV Ψ makes it an informative model to examine the mechanism of retroviral gRNA packaging, which is incompletely understood. Little is known about the structure of dimerization initiation sites or specific Gag interaction sites of RSV gRNA. Using selective 2'-hydroxyl acylation analyzed by primer extension (SHAPE), we probed the secondary structure of the entire RSV 5'-leader RNA for the first time. We identified a putative bipartite dimerization initiation signal (DIS), and mutation of both sites was required to significantly reduce dimerization in vitro. These mutations failed to reduce viral replication, suggesting that in vitro dimerization results do not strictly correlate with in vivo infectivity, possibly due to additional RNA interactions that maintain the dimers in cells. UV crosslinking-coupled SHAPE (XL-SHAPE) was next used to determine Gag-induced RNA conformational changes, revealing G218 as a critical Gag contact site. Overall, our results suggest that disruption of either of the DIS sequences does not reduce virus replication and reveal specific sites of Gag-RNA interactions.

Keywords: Gag polyprotein; RNA secondary structure; RNA structure-probing; RNA–protein interactions; Rous sarcoma virus (RSV); UV crosslinking-coupled SHAPE (XL-SHAPE); dimerization initiation signal (DIS); packaging signal (Ψ); retroviruses; selective 2′-hydroxyl acylation analyzed by primer extension (SHAPE).

Figure 1

Selective 2′-hydroxyl acylation analyzed by primer extension (SHAPE) reactivity-constrained lowest energy secondary structure of the 636-nt Rous sarcoma virus (RSV) 5′-leader RNA. The secondary structure model was generated by applying averaged normalized SHAPE reactivity from three independent trials as pseudo free-energy constraints in RNAstructure [53]. Nucleotides are colored in accordance to SHAPE reactivity as indicated in the key. Nucleotides that could not be analyzed due to primer annealing (3′ end) or the lack of resolution (5′ end) are shown in grey. MΨ is indicated by a black dashed box. Stem loops in µΨ and loops that were tested in the dimerization assays (DIS-0, DIS-1/L3, DIS-2/SL-A, SL-B and SL-C) are labeled. The AUG start codon is indicated by a purple box.

Figure 2

(A) RNA constructs used in the dimerization studies. Sequences in the regions containing DIS-2/SL-A, DIS-1/L3 and DIS-0 are shown. Nucleotides in the loop regions are underlined. Tetraloop mutations are highlighted in red. (B) Native agarose gel showing the results of dimerization assays of the wild-type (WT) RSV 5′-leader and dimerization initiation signal (DIS) loop mutants. RNAs (0.5 μM) were folded under low salt (L) or under high salt (H), as described in the Materials and Methods. Locations for monomeric (M) and dimeric RNA bands (D) are indicated. This gel shown is representative of three independent replicates. (C) Native agarose gel showing the results of dimerization assays of the WT RSV 5′-leader and more extensive stem loop deletion mutants. The conditions are the same as panel (B). This gel shown is representative of two independent replicates.

Figure 3

(A) Schematic of the constructs used in the infectivity studies. (B) Immunoblots of Gag in 50 µg of lysates collected from QT6 cells transfected with RC.pAT.V8 WT, RC.pAT.V8 ∆DIS-2, RC.pAT.V8 ∆DIS-1, RC.pAT.V8 DIS-DM or RC.V8 Bal31 mCherry. Lysates were collected in RIPA at 3 and 6 dpt. (C) Immunoblots of virions collected from QT6 cells transfected with RC.pAT.V8 WT, RC.pAT.V8 ∆SL-A, RC.pAT.V8 ∆L3, RC.pAT.V8 ∆L3+SL-A and RC.V8 Bal31 mCherry (top panel, 2 dpt), and immunoblots of Gag in 50 µg of lysates collected from QT6 cells infected with viruses produced from the transfected cells (bottom panel, 3, 6, 14 and 17 dpi).

Figure 4

Crosslinking (XL)-SHAPE analysis of RSV Gag∆PR binding to the RSV 5′-leader RNA. (A) Schematic showing individual domains of RSV Gag and RSV Gag∆PR. (B) XL-SHAPE results mapped to the secondary structure of the RSV 5′-leader RNA. Sites with decreased and increased SHAPE reactivity upon protein binding are indicated by green and red arrows, respectively. Identified crosslinking sites are labeled with stars. All identified sites have reactivity changes of ≥0.3 and p < 0.05 based on unpaired, two-tailed student t-tests, compared with the no protein control. The MΨ region (dashed box) and PBS are indicated. Results are based on the average of at least 3 independent experiments.

Figure 5

(A) Secondary structure of RSV MΨ. Six mutant MΨ constructs tested are indicated by red boxes and black arrows. Additional nucleotides that were added to facilitate in vitro transcription are shown in grey. (B,C) Bar graphs of K_d,1M (B) and Z_eff (C) values determined by fluorescence anisotropy (FA) salt titration binding assays using RSV Gag∆PR and RSV 167, MΨ-WT and six MΨ mutants. Values of three to four trials are indicated by black dots and the average is indicated by the height of the bar. p values were determined in comparison with MΨ-WT using unpaired, two-tailed student t-tests. (***) p < 0.001; (**) p < 0.01; (*) p < 0.05; (n.s.) not significant.