Noroviruses Co-opt the Function of Host Proteins VAPA and VAPB for Replication via a Phenylalanine-Phenylalanine-Acidic-Tract-Motif Mimic in Nonstructural Viral Protein NS1/2

Abstract

The Norovirus genus contains important human pathogens, but the role of host pathways in norovirus replication is largely unknown. Murine noroviruses provide the opportunity to study norovirus replication in cell culture and in small animals. The human norovirus nonstructural protein NS1/2 interacts with the host protein VAMP-associated protein A (VAPA), but the significance of the NS1/2-VAPA interaction is unexplored. Here we report decreased murine norovirus replication in VAPA- and VAPB-deficient cells. We characterized the role of VAPA in detail. VAPA was required for the efficiency of a step(s) in the viral replication cycle after entry of viral RNA into the cytoplasm but before the synthesis of viral minus-sense RNA. The interaction of VAPA with viral NS1/2 proteins is conserved between murine and human noroviruses. Murine norovirus NS1/2 directly bound the major sperm protein (MSP) domain of VAPA through its NS1 domain. Mutations within NS1 that disrupted interaction with VAPA inhibited viral replication. Structural analysis revealed that the viral NS1 domain contains a mimic of the phenylalanine-phenylalanine-acidic-tract (FFAT) motif that enables host proteins to bind to the VAPA MSP domain. The NS1/2-FFAT mimic region interacted with the VAPA-MSP domain in a manner similar to that seen with bona fide host FFAT motifs. Amino acids in the FFAT mimic region of the NS1 domain that are important for viral replication are highly conserved across murine norovirus strains. Thus, VAPA interaction with a norovirus protein that functionally mimics host FFAT motifs is important for murine norovirus replication.IMPORTANCE Human noroviruses are a leading cause of gastroenteritis worldwide, but host factors involved in norovirus replication are incompletely understood. Murine noroviruses have been studied to define mechanisms of norovirus replication. Here we defined the importance of the interaction between the hitherto poorly studied NS1/2 norovirus protein and the VAPA host protein. The NS1/2-VAPA interaction is conserved between murine and human noroviruses and was important for early steps in murine norovirus replication. Using structure-function analysis, we found that NS1/2 contains a short sequence that molecularly mimics the FFAT motif that is found in multiple host proteins that bind VAPA. This represents to our knowledge the first example of functionally important mimicry of a host FFAT motif by a microbial protein.

Keywords: noroviruses; plus-strand RNA virus; protein structure-function; reverse genetic analysis; viral replication; virus-host interactions.

FIG 1

Murine norovirus replication in Vapa^−/− cells is diminished. (A) VAPA Western blot of Vapa^−/− cell lines. (B) Representative infection frequency of MNoV-CW3 in Vapa^−/− cells, measured by FACS analysis of intracellular NS1/2 (18 h postinfection; MOI of 5.0). (C) Same as panel B. Data represent results of combined experiments (repeated-measure two-way ANOVA, Dunnett posttest; n = 3). (D) MNoV strain CW3 growth in Vapa^−/− and Vapa^+/+ cell lines (MOI, 0.05 [left] or 5.0 [right] PFU/cell). Data represent results of repeated-measure one-way ANOVA and the Dunnett posttest (n = 6). (E) Western blot of Vapa^+/+ or Vapa^−/− cell lines lentivirally transduced with FLAG-GFP or FLAG-Vapa. (F) Infection frequency in Vapa- or GFP-transduced cells determined as described for panel B (two-way ANOVA, Sidak posttest; n = 9). (G) CW3 growth in Vapa- or GFP-transduced cells. Data represent results of repeated-measure two-way ANOVA and the Dunnett posttest (n = 5). For G the asterisks refer to a comparison to the time-matched +/+ GFP control.

FIG 2

Murine norovirus replication in RAW 264.7-Vapa^−/− cells is impaired early in the viral life cycle. (A) Western blot of NS1/2 in Vapa^+/+ and Vapa^−/− (3A11) cell lines (MOI of 5). (Right panel) Combined densitometry data from multiple experiments performed on film exposures for each time point within the linear range of assay (n = 2 to 4) (unpaired t test, means compared to H_o = 100). (B) NS1/2 Western blot after electroporation of viral RNA (vRNA) into Vapa^+/+ and Vapa^−/− 3A11 cells (representative, n = 3 to 5). (Middle panel) Vapa^+/+ and Vapa^−/− cells were transfected equivalently with pMAX-GFP. (Right panel) Combined densitometry data determined as described for panel A (n = 3 to 5). (C) Viral-strand-specific quantitative PCR for negative strand (left) and positive strand (right) over time in infected Vapa^+/+ and Vapa^−/− 3A11 cells (MOI of 5; n = 3; two-way ANOVA).

FIG 3

NS1/2 interactions with VAPA are conserved between norovirus strains and occur during infection. (A) BV2 cells were infected with NS1/2-FLAG or NS4-FLAG MNV for 8 h (MOI of 10 TCID₅₀/cell). FLAG pulldown was performed on lysates, and immunoblotting was performed with the specified antibodies. M, molecular marker. (B) M2H interaction of NS1/2^GI, NS1/2^MNoV (CR6 and CW3), OSBP, and VAPA with VAPA or VAPB (bottom) (one-way ANOVA, Dunnett posttest; fold change data are shown on the right; n = 3). fluc, firefly luciferase; Rluc, Renilla luciferase.

FIG 4

NS1/2 binds FFAT-interacting residues in MSP domain of VAPA. (A) M2H interaction of NS1/2^MNoV with VAPA mutants. (B) Chemical shift perturbations of amide resonances upon unlabeled-NS1^CW3 titration into ¹⁵N-labeled VAPA MSP. The horizontal broken line represents the threshold. (C) M2H analysis of additional single-residue mutant VAPA. Designations of residues interacting with FFAT are underlined (one-way ANOVA, Dunnett posttest; fold change data are shown at the top; n = 3). (D) Murine VAPA MSP domain (PDB 2CRI). Pink highlighting indicates residues that disrupted the NS1/2-VAPA interaction in M2H analysis when mutated; mutations in cyan residues did not disrupt interaction. (E) Multiple alignment of VAPA and VAPB MSP domains from human (Hs) and mouse (Mm). Residues indicated with a black character differ from consensus data. Red asterisks mark residues necessary for interaction in M2H analysis, and triangles mark residues that shifted in NMR during NS1/2 titration.

FIG 5

A poorly conserved NS1 domain within NS1/2^MNoV interacts with VAPA. (A) Alignment of NS1/2 from representative strains from each norovirus genogroup. %ID, percent identity. (B) M2H analysis of full-length or domain truncations of NS1/2^MNoV (CR6) with VAPA (one-way ANOVA and Dunnett posttest; fold change data are shown at the top; n = 3).

FIG 6

The N-terminal segment of NS1-MNoV interacts with VAPA. (A) Chemical shift perturbations of amide resonances upon titration of unlabeled VAPA into ¹⁵N-labeled NS1-CR6 and CR6^M48G. The horizontal broken line represents the threshold. Purple residues are indicated as described for panel B. (B) Sequence logo of FFAT-like amino acid sequence of NS1/2 derived from BLAST alignment (Fig. S4C). The font size for each amino acid is proportional to percent conservation at each position. Residues exhibiting greater variability across MNoV strains are highlighted with arrows (colored purple here). (C) M2H interaction with NS1/2 substitutions (NS1/2, bait, VAPA, prey). Residues 69, 121, and 131 are not predicted to interact with VAPA. Purple residues are indicated as described for panel B (one-way ANOVA, Dunnett posttest; fold change data are shown at the top; n = 3).

FIG 7

NS1/2 interaction with VAPA enhances recovery of murine norovirus from infectious clones. (A) Recovery titers of mutants of MNoV strain pCR6. Data represent passage 1 titers (n = 7 to 20). (B) Summary of interaction of NS1/2 mutants with VAPA in M2H analysis, and recovery of virus from infectious clones for CW3 and CR6 NS1/2 mutants. (C) Molecular surface-and-ribbon diagram of solution structure of NS1-MNoV (PDB 2MCD [10]) with viable (black) and nonviable (red) mutants.