The hepatitis C viral nonstructural protein 5A stabilizes growth-regulatory human transcripts

Abstract

Numerous mammalian proto-oncogene and other growth-regulatory transcripts are upregulated in malignancy due to abnormal mRNA stabilization. In hepatoma cells expressing a hepatitis C virus (HCV) subgenomic replicon, we found that the viral nonstructural protein 5A (NS5A), a protein known to bind to viral RNA, also bound specifically to human cellular transcripts that encode regulators of cell growth and apoptosis, and this binding correlated with transcript stabilization. An important subset of human NS5A-target transcripts contained GU-rich elements, sequences known to destabilize mRNA. We found that NS5A bound to GU-rich elements in vitro and in cells. Mutation of the NS5A zinc finger abrogated its GU-rich element-binding and mRNA stabilizing activities. Overall, we identified a molecular mechanism whereby HCV manipulates host gene expression by stabilizing host transcripts in a manner that would promote growth and prevent death of virus-infected cells, allowing the virus to establish chronic infection and lead to the development of hepatocellular carcinoma.

Figure 1.

GRE-containing NS5A target transcripts encode regulators of apoptosis (A) and cell growth/proliferation (B). Transcripts depicted in green are NS5A target transcripts that contain GREs based on a FCE > 3 as defined in the Materials and Methods. These pathway figures were created using Ingenuity Pathway Assist software (Qiagen Inc).

Figure 2.

NS5A target transcripts are highly enriched for transcript stabilization and up-regulation. (A) The percentage of NS5A target transcripts, non-NS5A target transcripts and all transcripts that were stabilized (blue bars) or stabilized and up-regulated (red bars) is shown. ** represents statistically significant differences (P-value < 10⁻¹⁶, Fisher's exact test with R) and *represents statistically significant differences (P-value < 10⁻¹¹, Fisher's extact test, R) in the percentages comparing NS5A target transcripts to Non-NS5A target transcripts or all transcripts. (B) Top: A motif search was performed to look for conserved consensus sequences in the 3′ UTRs of NS5A target transcripts. The top 12-mer motif is shown. The position in the signal (bases) is depicted on the horizontal axis. The height of each stack of letters on the vertical axis is proportional to the residue frequency in the given position. Bottom: The motif previously found in CELF1 target transcripts that resembles the top 12-mer motif shown above.

Figure 3.

GRE-containing host mRNA transcripts are stabilized in Huh cells stably expressing an HCV subgenomic replicon (Huh-HCV). Huh or Huh-HCV cells were transfected with the BBB-GRE or BBB-ARE beta-globin reporter constructs. Actinomycin D was added to stop transcription and total cellular RNA was isolated after 0, 3 or 6 h. Specific mRNA levels were determined by quantitative real time RT-PCR. Beta-globin transcript levels at each time point were normalized to the transcript levels from a co-transfected GFP reporter. Transcript levels at the 0 time point were set to 100%, and the percent mRNA remaining was plotted as a function of time. The error bars indicate the standard error of the mean (SEM) from three experiments. Transcript half-life and SEM are shown to the right of each graph.

Figure 4.

NS5A binds to GRE-containing transcripts in cells. Huh7-HCV cells were transfected with the BBB, or BBB-GRE reporter plasmids. Cell lysates were immunoprecipitated using specific antibodies against the His-tag (HIS), NS5A or the poly A binding protein (PABP). RNA isolated from the input (I) or the pellet fraction (P) was reverse transcribed and amplified by PCR using beta-globin and HPRT specific primers, and the RNA was separated by electrophoresis. Water (H₂O) was used as a contamination control for the PCR.

Figure 5.

Recombinant NS5A binds to GRE RNA in a manner that is dependent on an intact zinc-binding site and the presence of zinc. (A) The upper panel shows a ribbon diagram of dimeric NS5A domain 1 that was prepared by using reference number 1ZH1 from the Protein Data Bank. One subunit is colored red and the other is colored green. The lower panel zooms in on the zinc-binding site of each subunit. Four conserved cysteines are required for zinc binding. (B) The NS5A domain 1+ polypeptide was titrated into a binding reaction buffer (20 mM HEPES pH 7.5, 5 mM MgCl₂, 10 mM 2-mercaptoethanol, 100 mM NaCl) in the absence (-Zn²⁺) or presence (+Zn²⁺) of 100 μM ZnCl₂ and incubated briefly at 25°C in a final volume of 100 μl. Binding of NS5A was measured by the change in polarization (mP). The change in fluorescence polarization was plotted as a function of NS5A domain 1+ concentration and fit to a hyperbola by using KaleidaGraph (Synergy Software). (C) Experiments were performed as described in panel B in the presence of zinc using the NS5A domain 1+ polypeptide (WT) or the derivative whose zinc-binding site was inactivated by converting the four cysteine residues to serine residues (4C-4S).

Figure 6.

Exogenously expressed NS5A stabilizes GRE-containing reporter genes. (A) HeLa tet-off cells were transfected with BBB, BBB-GRE, BBB-ARE or BBB-mGRE reporter plasmids as well as a plasmid that express NS5A, NS5A-4C-4S, ΔN-NS5A, ΔN-NS5A-4C-4S or a mock control plasmid. A GFP expression plasmid was included in each to control for transfection efficiency. Transcription from the tet-responsive promoter was stopped with 300 ng of doxycycline and RNA harvested after 0, 1.5, 3, 4.5 or 6 h was analyzed by northern blotting using GFP and beta-globin probes. (B) The experiment shown in (A) was performed three times, and the northern blot signals were quantified by a Storm 820 phosphorimager (Amersham Biosciences). For each time point, the intensity of the beta-globin reporter was normalized to the intensity of the GFP band, and the band intensity at the 0 time point was set to 100%. The percent of mRNA remaining was plotted over time. The error bars indicate the standard error of the mean (SEM) from three experiments. The calculated transcript half-life and SEM are shown to the right of each graph.