Equine lentivirus Gag protein degrades mitochondrial antiviral signaling protein via the E3 ubiquitin ligase Smurf1

Abstract

Equine infectious anemia virus (EIAV) and HIV-1 are both members of the Lentivirus genus and are similar in virological characters. EIAV is of great concern in the equine industry. Lentiviruses establish a complex interaction with the host cell to counteract the antiviral responses. There are various pattern recognition receptors in the host, for instance, the cytosolic RNA helicases interact with viral RNA to activate the mitochondrial antiviral signaling protein (MAVS) and subsequent interferon (IFN) response. However, viruses also exploit multiple strategies to resist host immunity by targeting MAVS, but the mechanism by which lentiviruses are able to target MAVS has remained unclear. In this study, we found that EIAV infection induced MAVS degradation, and that EIAV Gag protein recruited the E3 ubiquitin ligase Smurf1 to polyubiquitinate and degrade MAVS. The CARD domain of MAVS and the WW domain of Smurf1 are responsible for the interaction with Gag. EIAV Gag is a precursor polyprotein of the membrane-interacting matrix p15, the capsid p26, and the RNA-binding nucleocapsid proteins p11 and p9. Therefore, we analyzed which protein domain of Gag could interact with MAVS and Smurf1. We found that p15 and p26, but not p11 or p9, target MAVS for degradation. Moreover, we identified the key amino acid residues that support the interactions between p15 or p26 and MAVS or Smurf1. The present study describes a novel role of the EIAV structural protein Gag in targeting MAVS to counteract innate immunity, and reveals the mechanism by which the equine lentivirus can antagonize against MAVS.IMPORTANCEHost anti-RNA virus innate immunity relies mainly on the recognition by retinoic acid-inducible gene I (RIG-I) and melanoma differentiation-associated protein 5 (MDA5), and subsequently initiates downstream signaling through interaction with mitochondrial antiviral signaling protein (MAVS). However, viruses have developed various strategies to counteract MAVS-mediated signaling, although the method of antagonism of MAVS by lentiviruses is still unknown. In this article, we demonstrate that the precursor (Pr55gag) polyprotein of EIAV and its protein domains p15 and p26 target MAVS for ubiquitin-mediated degradation through E3 ubiquitin ligase Smurf1. MAVS degradation leads to the inhibition of the downstream IFN-β pathway. This is the first time that lentiviral structural protein has been found to have antagonistic effects on MAVS pathway. Overall, our study reveals a novel mechanism by which equine lentiviruses can evade host innate immunity, and provides insight into potential therapeutic strategies for the control of lentivirus infection.

Keywords: EIAV; HIV; MAVS; RLR; lentivirus.

Fig 1

EIAV infection downregulated MAVS expression to inhibit activation of IFN-β. (A) The IFN-β mRNA levels in eMDMs infected with EIAV. The fold change in levels of IFN-β mRNA in eMDMs infected with 2 × 10⁴ TCID50 EIAV was measured using qPCR at 0, 12, 24, 36, 48, 72, and 96 hpi. The data represent the means ± SEM from three independent experiments. (B) The expression level of MAVS after siRNA knockdown. eMDMs were treated with NC siRNA (siNC) or MAVS siRNA (siMAVS) for 36 h. The cellular lysates were analyzed by immunoblot analysis with anti-MAVS or anti-β-actin. The results of the intensity band ratio of MAVS to β-actin are shown by histogram. The data represent the means ± SEM from three independent experiments. (C) The IFN-β mRNA levels in eMDMs were mock treated or infected with EIAV after NC or MAVS siRNA knockdown. The fold change in levels of IFN-β mRNA in NC and MAVS siRNA knockdown eMDMs infected with 2 × 10⁴ TCID50 EIAV was measured using qPCR at 48 and 72 hpi. The data represent the means ± SEM from three independent experiments. (D) The mRNA levels of IFN-β in eMDMs (treated by poly I:C) with or without EIAV infection after NC or MAVS siRNA knockdown. The NC and MAVS siRNA knockdown eMDMs were treated with poly I:C (20 µg/mL), then mock infected or infected with 2 × 10⁴ TCID50 EIAV. The fold change in levels of IFN-β mRNA in cell lysate was measured using qPCR at 72 hpi. The data represent the means ± SEM from three independent experiments. (E) NBL-6 cells were mock treated or infected with EIAV_CMV3-8 at a TCID50 of 1 × 10³, 5 × 10³, 1 × 10⁴, 2 × 10⁴, or 5 × 10⁴ (wedge). Cells were harvested at 72 hpi and were assessed using immunoblot analysis with anti-MAVS, anti-Gag, or anti-β-actin antibody. (F) HEK293T cells were mock treated or transfected with pCMV3-8 plasmid at a dose of 0.5, 1, 1.5, 2, or 2.5 µg (wedge), co-transfected with 1 µg MAVS. Cells were harvested at 48 hpt and were assessed using immunoblot analysis with anti-Flag, anti-Gag, or anti-β-actin antibody. For panels E and F, the results of the densitometry analysis to quantify the ratio of MAVS to β-actin are shown at the bottom (lane 1 set as 1). The experiment was performed three times. (G) HEK293T cells were co-transfected with 1 µg pCMV3-8 and 1 µg MAVS plasmids. Cells were harvested at 0, 6, 12, 24, 36, and 48 hpt and were assessed using immunoblot analysis with anti-Flag, anti-Gag, anti-p26, or anti-β-actin antibody. The results of the densitometry analysis to quantify the ratio of MAVS to β-actin are shown at the bottom (lane 5 set as 1). This experiment was performed three times. (H) HEK293T cells were co-transfected with pGL3-IFN-Luc, PRL-TK, and either an empty vector or pCMV3-8. Simultaneously, cells were mock treated or transfected with 1 µg MAVS-Flag plasmid. Cells were harvested at 24 hpt and were assessed for luciferase activity. The results are presented as relative luciferase activity. Expression levels of expressed proteins were analyzed by immunoblot analysis of the lysates with anti-Flag, anti-Gag, or anti-β-actin antibody. The results of the densitometry analysis to quantify the ratio of MAVS to β-actin are shown at the bottom (lane 3 set as 1). For panels A, B, C, D, and H, significant differences between the different groups were determined using Student’s t-tests. *, P < 0.05; **, P < 0.01; ***, P < 0.001. Error bars represent the standard error over three independent experiments.

Fig 2

EIAV Gag protein triggers MAVS degradation. (A) HEK293T cells were co-transfected with either an empty vector (negative control, NC) or HA-tagged S2, Tat, Rev, Gag, or Env and MAVS-Flag. Cells were harvested at 24 hpt and were assessed using immunoblot analysis with anti-HA, anti-Flag, or anti-β-actin antibody. (B) HEK293T cells were co-transfected with either an empty vector or MAVS-Flag and Gag (1 or 2 µg; wedge). Cells were harvested at 24 hpt and were assessed using immunoblot analysis with anti-Flag, anti-Gag, or anti-β-actin antibody. For panels A and B, the results of the densitometry analysis to quantify the ratio of MAVS to β-actin are shown at the bottom (lane 1 set as 1). The experiment was performed three times. (C) HEK293T cells were co-transfected with pGL3-IFN-Luc, pRL-TK, and either an empty vector or Gag plasmid. Simultaneously, cells were mock treated or transfected with 1 µg MAVS-Flag plasmid. Cells were harvested at 24 hpt and were assessed for luciferase activity. The results are presented as relative luciferase activity. Expression levels of expressed proteins were analyzed by immunoblot analysis of the lysates with anti-Flag, anti-Gag, or anti-β-actin antibody. The results of the densitometry analysis to quantify the ratio of MAVS to β-actin are shown at the bottom (lane 3 set as 1). (D) HEK293T cells were mock treated or transfected with Gag. Cells were harvested at 0, 6, 12, and 24 hpt and were assessed using qPCR with MAVS mRNA. For panels C and D, significant differences between the different groups were determined using Student’s t-tests. NS, not significant, P > 0.05; *P < 0.05; **P < 0.01; ***P < 0.001. Error bars represent the standard error over three independent experiments.

Fig 3

EIAV Gag protein interacts with MAVS in cytoplasm. (A) HEK293T cells were co-transfected with MAVS-HA and either an empty vector or Gag-Flag. At 24 hpt, cells were harvested, immunoprecipitated with anti-Flag antibody, and further detected using immunoblot analysis with anti-HA or anti-Flag antibody. (B) HEK293T cells were co-transfected with Gag-HA and either an empty vector or MAVS-Flag. At 24 hpt, cells were harvested, immunoprecipitated with anti-Flag antibody, and further detected using immunoblot analysis with anti-HA or anti-Flag antibody. For panels A and B, expression levels of the proteins were analyzed by immunoblot analysis of the lysates with anti-HA, anti-Flag, or anti-β-actin antibody. The experiment was performed three times. (C) HEK293T cells were co-transfected with Gag-HA and MAVS-Flag. The cells were fixed at 12, 24, and 36 hpt and were stained with anti-HA or anti-Flag antibody to detect Gag-MAVS (Alexa Fluor 555 [AF555] and AF488 readout) using immunofluorescence assay; DAPI staining (blue) was performed to visualize nuclei (scale bar, 20 µm; 5 µm). Shown is an intensity profile of the linear region of interest (ROI) across the HEK293T cell co-stained with Gag and MAVS. Ten visual fields for each group were examined. (D) eMDMs were mock treated or infected with EIAV at 2 × 10⁴ TCID50. Cells were harvested at 24, 48, and 72 hpi and were detected using immunofluorescence assay with anti-MAVS or anti-Gag antibody (Alexa Fluor 555 [AF555] and AF488 readout) (scale bar, 10 µm; 5 µm). Shown is an intensity profile of the linear ROI across the eMDMs co-stained with Gag and MAVS. (E) HEK293T cells were co-transfected with Gag-Flag and either an empty vector or MAVS-HA, or MAVS-HA and empty vector. PLA was used to quantify the overlap between MAVS and Gag. The red dots represent the interactions between Gag and MAVS in situ (at distances of <40 nm); DAPI staining (blue) was performed to visualize nuclei (scale bar, 10 µm). The number of red dots (blobs/nucleus) is presented by the histogram. The data represent the means ± SEM from 30 cells in three independent experiments.

Fig 4

EIAV Gag protein triggers MAVS degradation through Smurf1-mediated ubiquitin-proteasome pathway. (A) HEK293T cells were co-transfected with Gag-HA or either an empty vector and MAVS-Flag and were maintained in the presence or absence of the proteasome inhibitor MG132 (20 µM, 6 h prior to immunoblot analysis), the lysosome inhibitor CQ (50 µM), or the autophagy inhibitor wortmannin (100 nM) for 12 h. Cells were harvested and assessed using immunoblot analysis with anti-Flag, anti-HA, or anti-β-actin antibody. The results of the densitometry analysis to quantify the ratio of MAVS to β-actin are shown at the bottom (lane 1 set as 1). (B) HEK293T cells were co-transfected with MAVS-Flag, ubiquitin-HA, ubiquitin-HA (K48), or ubiquitin-HA (K63) and either an empty vector or Gag-Myc, and were maintained in the presence of the proteasome inhibitor MG132 (20 µM, 6 h prior to immunoprecipitation). At 24 hpt, cells were harvested, immunoprecipitated with anti-Flag antibody, and further assessed using immunoblot analysis with anti-Flag and anti-HA antibody. Expression levels of the proteins were analyzed using immunoblot analysis of the lysates with anti-Myc, anti-Flag, anti-HA, or anti-β-actin antibody. (C) HEK293T cells were transfected with either a scrambled siRNA (siNC) or a specific siRNA targeting AIP4, RNF125, MARCH5, MUL1, RNF5, or Smurf1. At 48 h after transfection, cells were co-transfected with MAVS-Flag and Gag-HA plasmids. Cells were harvested at 18 hpt and were assessed using immunoblot analysis with anti-Flag, anti-HA, or anti-β-actin antibody. The results of the densitometry analysis to quantify the ratio of MAVS to β-actin are shown at the bottom (lane 7 set as 1). (D) HEK293T cells were co-transfected with Gag-Flag and HA-tagged RNF125, MUL1, RNF5, Smurf1, MARCH5, AIP4 or the empty vector pCAGGS. (E) HEK293T cells were co-transfected with Gag-HA and either an empty vector or Smurf1-Flag. For panels D and E, at 24 hpt, cells were harvested, immunoprecipitated with anti-Flag antibody, and further assessed using immunoblot analysis with anti-HA or anti-Flag antibody. Expression levels of the proteins were analyzed using immunoblot analysis of the lysates with anti-HA, anti-Flag, or anti-β-actin antibody. (F) HEK293T cells were co-transfected with MAVS-Flag and either an empty vector or Smurf1-HA or Gag-Myc. Expression levels of the proteins were analyzed using immunoblot analysis of the lysates with anti-HA, anti-Flag, anti-Myc, or anti-β-actin antibody. The results of the densitometry analysis to quantify the ratio of MAVS to β-actin are shown at the bottom (lane 1 set as 1). All experiments were performed three times.

Fig 5

EIAV Gag protein enhances the interaction between MAVS and Smurf1. (A) HEK293T cells were co-transfected with Smurf1-HA and either an empty vector or MAVS-Flag. (B) HEK293T cells were co-transfected with MAVS-HA and either an empty vector or Smurf1-Flag. (C) HEK293T cells were co-transfected with Smurf1-HA and either an empty vector or MAVS-Flag or Gag-Myc. For panels A, B, and C, at 24 hpt, cells were harvested, immunoprecipitated with anti-Flag antibody, and further assessed using immunoblot analysis with anti-HA or anti-Flag antibody. Expression levels of the proteins were analyzed using immunoblot analysis of the lysates with anti-HA, anti-Flag, anti-Myc, or anti-β-actin antibody. All experiments were performed three times. (D) Detection of MAVS and Smurf1 interaction using BiFC assay. VN-MAVS and Smurf1-VC were expressed together with or without Gag in HEK293T cells, and Gag protein was stained with mouse anti-Gag antibody followed by Alexa Fluor 647-conjuated rabbit anti-mouse antibody. BiFC green fluorescent signals together with the expression of Gag were visualized using confocal microscopy. DAPI staining (blue) was performed to visualize nuclei (scale bar, 20 µm; 2 µm). The BiFC fluorescence intensity was used to quantify the overlap between MAVS and Smurf1. The fluorescence intensity is presented in the histogram. The data represent the means ± SEM from 30 cells in three independent experiments. Significant differences between the different groups were determined using Student’s t-tests. NS, not significant, P > 0.05; *P < 0.05; **P < 0.01; ***P < 0.001.

Fig 6

Domain mapping of the Gag-MAVS and Gag-Smurf1 interactions. (A) Schematic diagram of wild-type MAVS and structures of respective mutants. (B) HEK293T cells were co-transfected with Gag-Flag and either an empty vector or truncated MAVS-HA (aa 1 to 180, aa 1 to 341, aa 1 to 503, aa 180 to 341, and aa 341 to 530). At 24 hpt, cells were harvested, immunoprecipitated with anti-Flag antibody, and further assessed using immunoblot analysis with anti-HA or anti-Flag antibody, and expression levels of the proteins were analyzed using immunoblot analysis of the lysates with anti-HA, anti-Flag, or anti-β-actin antibody. This experiment was performed three times. (C) Schematic diagram of wild-type Smurf1 and structures of respective mutants. (D) HEK293T cells were co-transfected with Gag-Flag and either an empty vector or wild-type Smurf1 or truncated Smurf1-HA (aa 1 to 120, aa 121 to 419, and aa 420 to 757). At 24 hpt, cells were harvested, immunoprecipitated with anti-Flag antibody, and further assessed using immunoblot analysis with anti-HA or anti-Flag antibody. Expression levels of the proteins were analyzed by immunoblot analysis of the lysates with anti-HA, anti-Flag, or anti-β-actin antibody. This experiment was performed three times.

Fig 7

The domains p15 and p26, but not p11 and p9, of Gag could mediate MAVS degradation. (A) Wild-type Gag and structures of respective mutants. (B) HEK293T cells were co-transfected with MAVS-Flag and either an empty vector or Gag-HA (aa 1 to 486) or Δp9 (aa 1 to 435), p15-p26 (aa 1 to 354), p26 (aa 125 to 354), p15 (aa 1 to 124), and p11 (aa 360 to 435). At 24 hpt, cells were harvested and assessed using immunoblot analysis with anti-Flag, anti-HA, or anti-β-actin antibody. The results of the densitometry analysis to quantify the ratio of MAVS to β-actin are shown at the bottom (lane 1 set as 1). The experiment was performed three times. (C) HEK293T cells were co-transfected with MAVS-Flag and either an empty vector or Gag-HA (aa 1 to 486) or Δp9 (aa 1 to 435), p15-p26 (aa 1 to 354), p26 (aa 125 to 354), p15 (aa 1 to 124), and p11 (aa 360 to 435). At 24 hpt, cells were harvested, immunoprecipitated with anti-Flag antibody, and further assessed using immunoblot analysis with anti-HA or anti-Flag antibody. (D) HEK293T cells were co-transfected with Smurf1-Flag and either an empty vector or Gag-HA (aa 1 to 486) or Δp9 (aa 1 to 435), p15-p26 (aa 1 to 354), p26 (aa 125 to 354), p15 (aa 1 to 124), and p11 (aa 360 to 435). At 24 hpt, cells were harvested, immunoprecipitated with anti-Flag antibody, and further assessed using immunoblot analysis with anti-HA or anti-Flag antibody. For panels C and D, expression levels of the proteins were analyzed by immunoblot analysis of the lysates with anti-HA, anti-Flag, or anti-β-actin antibody. The experiments were performed three times.

Fig 8

Analysis of key amino acid sites in the MAVS-p15-Smurf1 interactions. (A) Structure-based prediction of the interface between p15 (pink) and MAVS (green). Crystal structure of the p15-MAVS conjugate (PDB 1HEK-4O9L) shows non-covalent interactions between p15 and MAVS (possible hydrogen bonds are indicated by yellow dashed lines, with distances in Å). D3 and E109 of p15 facilitate the formation of covalent bonds. (B) Structure-based prediction of the interface between p15 (blue) and Smurf1 (green). Crystal structure of the p15-Smurf1 conjugate (PDB 1HEK-F6R1Z0) shows non-covalent interactions between p15 and Smurf1 (possible hydrogen bonds are indicated by yellow dashed lines, with distances in Å). D55, Q70, E76, and R105 of p15 facilitate the formation of covalent bonds. (C) HEK293T cells were co-transfected with MAVS-Flag and either an empty vector or p15-HA or the p15(D3A)-HA (from Asp to Ala) and p15(E109A)-HA (from Glu to Ala). At 24 hpt, cells were harvested, immunoprecipitated with anti-Flag antibody, and further assessed using immunoblot analysis with anti-HA or anti-Flag antibody. (D) HEK293T cells were co-transfected with Smurf1-Flag and either an empty vector or p15-HA or the p15(D55A)-HA (from Asp to Ala), p15(Q70A)-HA (from Gln to Ala), p15(E76A)-HA (from Glu to Ala), and p15(K105A)-HA (from Lys to Ala). At 24 hpt, cells were harvested, immunoprecipitated with anti-Flag antibody, and further assessed using immunoblot analysis with anti-HA or anti-Flag antibody. For panels C and D, expression levels of the proteins were analyzed using immunoblot analysis of the lysates with anti-HA, anti-Flag, or anti-β-actin antibody. The experiments were performed three times.

Fig 9

Analysis of key amino acid sites in the MAVS-p26-Smurf1 interactions. (A) Structure-based prediction of the interface between p26 (blue) and MAVS (green). Crystal structure of the p26-MAVS conjugate (PDB 1EIA-4O9L) shows non-covalent interactions between p26 and MAVS (possible hydrogen bonds are indicated by yellow dashed lines, with distances in Å). T22, N25, T26, L40, Q192, and R203 of p26 facilitate the formation of covalent bonds. (B) Structure-based prediction of the interface between p26 (blue) and Smurf1 (green). Crystal structure of the p26-Smurf1 conjugate (PDB 1EIA-F6R1Z0) shows non-covalent interactions between p26 and Smurf1 (possible hydrogen bonds are indicated by yellow dashed lines, with distances in Å). T16, P61, K67, N83, E115, K149, and E180 of p26 facilitate the formation of covalent bonds. (C) HEK293T cells were co-transfected with MAVS-Flag and either an empty vector or p26-HA or the p26(T22A)-HA (from Thr to Ala), p26(N25A)-HA (from Asn to Ala), p26(T26A)-HA (from Thr to Ala), p26(L40A)-HA (from Leu to Ala), p26(Q192A)-HA (from Gln to Ala), and p26(R203A)-HA (from Arg to Ala). At 24 hpt, cells were harvested, immunoprecipitated with anti-Flag antibody, and further assessed using immunoblot analysis with anti-HA or anti-Flag antibody. (D) HEK293T cells were co-transfected with Smurf1-Flag and either an empty vector or p26-HA or the p26(T16A)-HA (from Thr to Ala), p26(P61A)-HA (from Pro to Ala), p26(K67A)-HA (from Lys to Ala), p26(N83A)-HA (from Asn to Ala), p26(E115A)-HA (from Glu to Ala), p26(K149A)-HA (from Lys to Ala), and p26(E180A)-HA (from Glu to Ala). At 24 hpt, cells were harvested, immunoprecipitated with anti-Flag antibody, and further assessed using immunoblot analysis with anti-HA or anti-Flag antibody. For panels C and D, expression levels of the proteins were analyzed using immunoblot analysis of the lysates with anti-HA, anti-Flag, or anti-β-actin antibody. The experiments were performed three times.

Fig 10

Schematic diagrams illustrate how EIAV negatively regulates RLR pathway activation by degrading MAVS. Virus infection results in activation of the RLR pathway and production of type I IFN. To escape such innate immunity of the host, in the late period of infection, EIAV core structure protein Gag binds to MAVS to recruit E3 ubiquitin ligase Smurf1, which can catalyze their K48-linked polyubiquitination for proteasome-dependent degradation. All above processes lead to the inhibition of the RLR signaling pathway and downregulation of IFN-β.