Bovine Delta Papillomavirus E5 Oncoprotein Interacts With TRIM25 and Hampers Antiviral Innate Immune Response Mediated by RIG-I-Like Receptors

Abstract

Persistent infection and tumourigenesis by papillomaviruses (PVs) require viral manipulation of various of cellular processes, including those involved in innate immune responses. Herein, we showed that bovine PV (BPV) E5 oncoprotein interacts with a tripartite motif-containing 25 (TRIM25) but not with Riplet in spontaneous BPV infection of urothelial cells of cattle. Statistically significant reduced protein levels of TRIM25, retinoic acid-inducible gene I (RIG-I), and melanoma differentiation-associated gene 5 (MDA5) were detected by Western blot analysis. Real-time quantitative PCR revealed marked transcriptional downregulation of RIG-I and MDA5 in E5-expressing cells compared with healthy urothelial cells. Mitochondrial antiviral signalling (MAVS) protein expression did not vary significantly between diseased and healthy cells. Co-immunoprecipitation studies showed that MAVS interacted with a protein network composed of Sec13, which is a positive regulator of MAVS-mediated RLR antiviral signalling, phosphorylated TANK binding kinase 1 (TBK1), and phosphorylated interferon regulatory factor 3 (IRF3). Immunoblotting revealed significantly low expression levels of Sec13 in BPV-infected cells. Low levels of Sec13 resulted in a weaker host antiviral immune response, as it attenuates MAVS-mediated IRF3 activation. Furthermore, western blot analysis revealed significantly reduced expression levels of pTBK1, which plays an essential role in the activation and phosphorylation of IRF3, a prerequisite for the latter to enter the nucleus to activate type 1 IFN genes. Our results suggested that the innate immune signalling pathway mediated by RIG-I-like receptors (RLRs) was impaired in cells infected with BPVs. Therefore, an effective immune response is not elicited against these viruses, which facilitates persistent viral infection.

Keywords: bovine papilloma virus E5 oncoprotein; melanoma differentiation-associated gene 5 (MDA5); mitochondrial antiviral-signalling (MAVS) protein; retinoic acid-inducible gene I (RIG-I); tripartite motif containing 25 (TRIM25).

Figure 1

(A) Copy number of BPV E5 mRNA by one-step RT-ddPCR analysis detected in 15 bladder tumours. Each plot represents an independent experiment. Statistical analysis revealed no significant differences in copy number of BPV E5 mRNA of each analysed bladder tumours. (B, C) showed immunoprecipitation assay using anti-TRIM25 and anti-Riplet antibodies, respectively, in healthy and infected bladder samples. Western blot analysis revealed that TRIM25 only interacted with E5 protein. Panels (B, C) show representative data from three independent experiments.

Figure 2

(A) Western blot analysis of total Riplet protein in infected, healthy and non-infected but inflammatory bladder samples. (B) Densitometric analysis of total Riplet protein relative to the β-actin protein level. Plots represent value found in each sample performed in duplicate. (C) Western blot analysis of total TRIM25 protein in infected, healthy and non-infected but inflammatory bladder samples. (D) Densitometric analysis of total TRIM25 protein relative to the β-actin protein level (*** p ≤ 0.001). Plots represent value found in each sample performed in duplicate. (E) Real time RT-PCR. TRIM25 mRNA levels in infected, healthy and non-infected but inflammatory bladder samples. Plots represent value found in each sample performed in triplicate. Statistical analysis revealed no significant difference between value of each pathological bovine bladder compared to the value of both healthy and uninfected bladders.

Figure 3

(A) Western blot analysis of RIG-1 and MDA5 in infected, healthy and non-infected but inflammatory bladder samples. (B) Densitometric analysis was performed by comparing the protein expression levels of total RIG-1 and MDA5 with those of β-actin. RIG-1 and MDA5 protein levels were significantly reduced in the infected mucosa samples compared with both healthy and uninfected samples. (***p ≤ 0.001; ****p ≤ 0.0001). Plots represent value found in each sample performed in duplicate. (C, D) Real-time RT-PCR analysis of RIG-I and MDA5 mRNA levels, respectively, in infected, healthy and non-infected but inflammatory bladder samples. RIG-I and MDA5 mRNA expressions were significantly reduced in diseased bladder samples compared with both healthy and uninfected bladder (*p ≤ 0.05). Plots represent value found in each sample performed in triplicate.

Figure 4

Immunoprecipitation using an anti-MAVS antibody in healthy and diseased bladder samples. Western blot analysis revealed that MAVS interacted with RIG-I, MDA5, TRIM25, phosphorylated TBK1 (pTBK1), phosphorylated IRF3 (pIRF3) and Sec13. Immunoprecipitation panel showed representative data from three independent experiments.

Figure 5

(A) Western blot analysis of total Sec13 protein in infected, healthy and non-infected but inflammatory bladder samples. (B) Densitometric analysis of total Sec13 protein relative to the β-actin protein level (**p ≤ 0.01). Plots represent value found in each sample performed in duplicate. (C) Western blot analysis of IRF3 and TBK1 proteins in infected, healthy and non-infected but inflammatory bladder samples. (D) Densitometric analysis of both proteins was performed relative with the β-actin protein levels (***p ≤ 0.001; ****p ≤ 0.0001). IRF3 band used for densitometry was indicated by arrow. Plots represent value found in each sample performed in duplicate. (E) Western blot analysis of phosphorylated TBK1 (pTBK1) in infected, healthy and non-infected but inflammatory bladder samples. (F) Densitometric analysis of pTBK1 protein relative to β-actin protein levels (****p ≤ 0.0001). Plots represent value found in each sample performed in duplicate.

Figure 6

Real-time RT-PCR analysis of (A) IKKα, (B) IKKβ, (C) IKKγ, (D) IFN-β mRNA levels in infected, healthy and non-infected but inflammatory bladder samples. IKKα, IKKβ, IKKγ and IFN-β mRNA levels were significantly reduced in diseased bladder samples compared with both healthy and uninfected bladder samples (****p ≤ 0.0001). Plots represent value found in each sample performed in triplicate.

Figure 7

Virus-mediated RIG-I and MDA5 signalling pathway resulting in IFN production. BPV E5 oncoprotein perturbs this pathway leading to a downregulation of IFN production thus contributing to cause a BPV persistent infection. Red arrows indicate downregulation of the proteins by BPV E5.