Polyprotein processing and intermolecular interactions within the viral replication complex spatially and temporally control norovirus protease activity

Abstract

Norovirus infections are a major cause of acute viral gastroenteritis and a significant burden on global human health. A vital process for norovirus replication is the processing of the nonstructural polyprotein by a viral protease into the viral components required to form the viral replication complex. This cleavage occurs at different rates, resulting in the accumulation of stable precursor forms. Here, we characterized how precursor forms of the norovirus protease accumulate during infection. Using stable forms of the protease precursors, we demonstrated that all of them are proteolytically active in vitro, but that when expressed in cells, their activities are determined by both substrate and protease localization. Although all precursors could cleave a replication complex-associated substrate, only a subset of precursors lacking the NS4 protein were capable of efficiently cleaving a cytoplasmic substrate. By mapping the full range of protein-protein interactions among murine and human norovirus proteins with the LUMIER assay, we uncovered conserved interactions between replication complex members that modify the localization of a protease precursor subset. Finally, we demonstrate that fusion to the membrane-bound replication complex components permits efficient cleavage of a fused substrate when active polyprotein-derived protease is provided in trans These findings offer a model for how norovirus can regulate the timing of substrate cleavage throughout the replication cycle. Because the norovirus protease represents a key target in antiviral therapies, an improved understanding of its function and regulation, as well as identification of interactions among the other nonstructural proteins, offers new avenues for antiviral drug design.

Keywords: Calicivirus; Norovirus; Precursor; RNA virus; fluorescence resonance energy transfer (FRET); plus-stranded RNA virus; protease; viral replication.

Figure 1.

Norovirus protease produced during authentic viral replication inefficiently cleaves a cytoplasmic-localized protease substrate. A, genome schematic of the murine norovirus genome highlighting the major open reading frames and the components of the polyprotein. B, schematic of the CFP-YFP FRET sensor highlighting the NS1/2-NS3 boundary cleavage site. C, Western blot analysis of HEK-293T cells co-transfected with the FRET sensor and either WT or catalytically inactive protease (WT and H30A, respectively), or BV-2 cells electroporated with in vitro transcribed RNA encoding the FRET sensor and subsequently infected at m.o.i. 10 TCID₅₀/cell with MNV for 9 h. Higher molecular weight NS6 containing isoforms produced during viral infection are indicated with asterisks. Note that the left- and right-hand regions of the NS6 Western blotting were imaged at different intensities to compensate for the lower NS6 signal intensity in infected cells. D, confocal microscopy of HeLa-CD300lf cells transfected with the FRET sensor and either co-transfected with an NS6 expression construct (NS6) and harvested at 18 h post-transfection, or infected with MNV (+MNV) at a m.o.i. of 10 TCID₅₀/cell at 6 h post-transfection, and harvested at 12 h post-infection. The FRET sensor was visualized in the CFP and YFP channels, the protease with antisera against NS6, and nuclei stained with DAPI. All experiments were repeated two times, with the exception of C, which was repeated three times. In each case one representative dataset is shown.

Figure 2.

Targeting protease substrates to the norovirus replication complex enhances cleavage by NS6. A, Western blotting of cells BSR-T7 cells transfected with either proteolytically active (MNV-FLC) or inactive (MNV-H30A) full-length infectious MNV clone, or polyprotein fused to the FRET sensor (ORF1-FRET-WT, ORF1-FRET-H30A). Samples were lysed at 18 h post-transfection. B, confocal microscopy of BSR-T7 cells fixed at 18 h post-transfection with ORF1-FRET-WT or ORF1-FRET-H30A constructs. The FRET substrate was visualized in the CFP and YFP channels, and nuclei stained with DAPI. All experiments were repeated two times, with one representative dataset shown.

Figure 3.

Accumulation of polyprotein precursors during MNV infection. BV-2 cells were infected at a m.o.i. of 10 TCID₅₀/cell, and harvested at the indicated times post-infection. Western blot analysis with antisera against (A) NS6 or (B) NS5/VPg revealed the high levels of precursor proteins present at early stages of infection. To ensure efficient transfer and resolution of the various precursors, samples were separated on both 7.5 (upper) and 17.5% (lower) SDS-PAGE gels prior to transfer and Western blotting. C, infectious virus produced across the time course was determined by TCID₅₀ following freeze-thaw to release intracellular virus. D, to investigate precursor localization, BV-2 cells were infected at m.o.i. 10 and harvested at 9 h post-infection where they were subject to fractionation into soluble cytoplasmic, and cytoplasmic membrane-associated fractions. Lysates were blotted for viral and cellular markers as indicated. All experiments with the exception of C were repeated two times, with one representative dataset shown. The virus titer experiments in C were performed in triplicate and error bars represent standard deviation. Where unambiguous identification of the specific precursor was not possible due to co-migration and antibody reactivity, both possible precursors are named and this is indicated with '*'.

Figure 4.

Activity of NS6-containing stable precursors. FLAG-tagged forms of the various protease-containing precursors were generated by mutagenesis of the P1 residue at each cleavage site to an alanine. A H30A mutation within the protease abolished proteolytic activity. A, a schematic illustrating the 10 possible NS6-containing precursors generated by NS6 cleavage of the polyprotein. Protease-containing precursors may all be observed following: B, in vitro translation of the precursors incorporating [³⁵S]Met labeling followed by phosphorimaging, or C, expression in BSR-T7 cells visualized by Western blotting. Incubation of unlabeled in vitro translated precursors with [³⁵S]methionine-labeled substrate demonstrates cleavage of the (D) cellular substrate PABP. E, Western blotting of BSR-T7 cells harvested at 18 h post-transfection with the precursors and FLAG-PABP reveals only a subset of precursors cleave PABP in cells. FLAG-tagged precursors are also visible on this image and are indicated with an asterisk (*). F, Western blotting of BSR-T7 cells co-expressing ORF1-FRET-H30A as substrate for trans-cleavage by co-expressed protease precursors shows cleavage of this substrate by all precursors at 18 h post-transfection. For E, the amount of cleavage product relative to that observed with the NS6–7 precursor and F, loss of the full-length substrate relative to that observed H30A mutant was determined by densitometry. The levels of cleavage are indicated below. 0–10% cleavage = −, 11–50% cleavage = *, 51–90% cleavage = **, 91–100% cleavage = ***. All experiments were repeated two times, with one representative dataset shown.

Figure 5.

Confocal microscopy of protease precursors in mock or MNV-infected cells. HeLa-CD300lf cells were transfected with the various protease precursor constructs and at 6 h post-transfection, mock- or infected with MNV at m.o.i. 10. The cells were fixed and processed for microscopy at 12 h post-infection. Nuclei were visualized with DAPI, the precursors with anti-FLAG antisera, and infected cells with anti-dsRNA antisera. All experiments were repeated two times, and 3 fields of view imaged for cells transfected with each protease precursor. One representative image for each condition is shown.

Figure 6.

Components of the norovirus replication complex exhibit homo- and heteromeric interactions. HEK-293T cells expressing protein A and Renilla luciferase fusions of the various MNV proteins were used for LUMIER analysis to identify protein–protein interactions. The numbers are robust Z-scores. Positive protein–protein interactions are colored by the strength of interaction with weak interactions showing in pale yellow, with the strongest interactions in purple. Protein expression was confirmed by Western blotting (Fig. S4). LUMIER experiments were performed in quadruplicate.

Figure 7.

Mutagenesis of polyprotein-FRET fusions reveals the importance of individual polyprotein components for replication complex targeting and substrate cleavage. A, schematic illustrating the various N-, C-, and internal deletion mutants generated from ORF1-FRET-H30A. These mutants were transfected into BSR-T7 cells to function as trans-cleavage substrates. WT or proteolytically inactive full-length clone (MNV-FLC, MNV-H30A) was provided in trans to determine cleavage efficiency. Samples were harvested at 18 h post-transfection. N-terminal (B), C-terminal (C), and internal (D) deletions are shown. Substrate cleavage was assessed using anti-GFP antisera. ORF1-FRET-WT was used as a positive control (+). For clarity the position of either the full-length (mock cells) or fully-processed cleavage product (MNV-infected cells) is highlighted with a red asterisk (*). The proportion of the full-length polyprotein-FRET fusion showing cleavage was determined by densitometry and calculated from the relative abundance of the individual full-length FRET fusion in the presence of MNV-FLC compared with its MNV-H30A control. 0–10% cleavage = −, 11–50% cleavage = *, 51–90% cleavage = **, 91–100% cleavage = ***. All experiments were repeated two times, the second replicate is shown in Fig. S7.