High-resolution functional profiling of the norovirus genome

Abstract

Human noroviruses (HuNoV) are a major cause of nonbacterial gastroenteritis worldwide, yet details of the life cycle and replication of HuNoV are relatively unknown due to the lack of an efficient cell culture system. Studies with murine norovirus (MNV), which can be propagated in permissive cells, have begun to probe different aspects of the norovirus life cycle; however, our understanding of the specific functions of the viral proteins lags far behind that of other RNA viruses. Genome-wide functional profiling by insertional mutagenesis can reveal protein domains essential for replication and can lead to generation of tagged viruses, which has not yet been achieved for noroviruses. Here, transposon-mediated insertional mutagenesis was used to create 5 libraries of mutagenized MNV infectious clones, each containing a 15-nucleotide sequence randomly inserted within a defined region of the genome. Infectious virus was recovered from each library and was subsequently passaged in cell culture to determine the effect of each insertion by insertion-specific fluorescent PCR profiling. Genome-wide profiling of over 2,000 insertions revealed essential protein domains and confirmed known functional motifs. As validation, several insertion sites were introduced into a wild-type clone, successfully allowing the recovery of infectious virus. Screening of a number of reporter proteins and epitope tags led to the generation of the first infectious epitope-tagged noroviruses carrying the FLAG epitope tag in either NS4 or VP2. Subsequent work confirmed that epitope-tagged fully infectious noroviruses may be of use in the dissection of the molecular interactions that occur within the viral replication complex.

Fig 1

(A) Strategy for Tn mutagenesis of the MNV-1 genome. A HyperMu transposase enzyme was used to insert one Tn per pT7:MNV 3′ RZ infectious clone to generate a Tn-mutagenized library. This was subcloned to generate five libraries containing Tns only in a well-defined region of the genome (gray shading).*, SpeI cleaves at a position within the plasmid vector; other numbers indicate genome positions. Removal of the Tns from each library leaves residual 15-nt insertions (black shading). Dashed lines, vector of the pT7:MNV 3′ RZ infectious clone. (B) Composition of the 15-nt insertion. Numbers in italics represent the 5-nt target site for the original transposition, 10 nt that originated from the transposon are shown and contain a NotI restriction site highlighted in bold, and the 5 underlined numbers represent target nucleotides duplicated during transposition. (C) Recovery of infectious virus from 15-nt insertion libraries. BSR-T7 cells were transfected with in vitro-transcribed, enzymatically capped RNA produced from each cDNA 15-nt insertion library. Virus was harvested at 48 h posttransfection, and the yield of recovered infectious virus was determined in RAW264.7 cells. The limit of detection is 50 TCID₅₀/ml. Transfections were performed in triplicate; error bars represent the standard error of the mean. Significance was tested using one-way analysis of variance and Dunnett's multiple-comparison posttest to compare each insertion virus to the WT. FS, frameshift clone; ***, P < 0.001; *, P < 0.5.

Fig 2

Schematic diagram showing selection and detection of insertions in the 15-nt insertion library. RNA was produced from each 15-nt cDNA insertion library by in vitro transcription and enzymatic capping of the DNA library, and virus was recovered using the RNA-mediated reverse genetics system, in duplicate. For selection in tissue culture, the recovered virus was sequentially passed (in duplicate) three times in RAW264.7 cells at a low MOI. RNA was extracted after each passage and alongside the input RNA was used as a template for RT-PCR amplification of overlapping fragments covering the respective mutagenized region of each library. Each fragment was used as a template in a second round of PCR with a virus-specific primer and the insertion-specific NotI miniprimer, labeled with a fluorescent tag (black circles). Fluorescently labeled products were detected and analyzed by capillary electrophoresis.

Fig 3

Example electropherograms showing the detection and positioning of 15-nt insertions in the fragment covering nt 6546 to 6931 (ORF2) during selection in tissue culture. Peak size is depicted on the x axis, and fluorescence intensity is indicated on the y axis. The similar insertion profiles of the cDNA library and the RNA input suggest that diversity was not lost during in vitro transcription. Comparison of sequential passages reveals selection occurring within the insertion mutant populations. Insertions within the solid box were maintained, whereas the dashed box highlights a group of insertions that were rapidly lost. Peaks under 65 nt correspond to primer noise and were excluded from the analysis.

Fig 4

(A) Genome-wide insertion profiling of the murine norovirus genome. Insertion sites detected in each passage (passage 1 [P1], passage 2 [P2], and passage 3 [P3]) for each library were compiled to generate a map of insertion sites across the entire genome, with the exception of the first 55 nt. A total of 2,089 insertions were detected in the input profile, with only 324 remaining after the three passages, indicating selection of viable mutants. Insertion sites found in the input and passage 3 populations at the end of NS4 (B) and the end of VP2 (C) are shown in greater detail. The insertions highlighted in red correspond to the position where a FLAG tag was later inserted in the NS4- and VP2-coding regions.

Fig 5

Structural mapping of the 15-nt insertion sites. The insertions found in the input and passage 3 profiles were mapped onto the structures of NS7 (PDB accession number 3NAH) (A and B, respectively) and the P domain of VP1 (PDB accession number 3LQ6) (C and D, respectively). Blue residues highlight positions after which an insertion was detected. Where two insertions were made in a single codon with different outcomes, the most tolerated one was marked. In the P domain of VP1, the majority of insertions tolerated after three passages lie within flexible loop regions. The catalytic center in NS7 did not tolerate any insertions, further validating the profiling. The structures were mapped using Chimera software (UCSF).

Fig 6

Validation of insertion profiling data. Tolerated insertion sites maintained for three passages were introduced into the pT7:MNV 3′ RZ (WT) infectious clone and yielded infectious virus by use of the DNA-based reverse genetics system. Numbers refer to the genome position after which the insertion was made. Insertions colored white lie within NS1-2, those in light gray are insertions in NS4, and those in dark gray are in VP2. Importantly, introduction of an insertion at position 6722, which was deemed replication incompetent in the profiling, failed to yield infectious virus by reverse genetics. The pT7:MNV 3′ RZ NS7 frameshift clone (FS) was used as a negative control for virus recovery. Significance was tested using one-way analysis of variance and Dunnett's multiple-comparison posttest to compare each insertion virus to the WT control, n = 3. **, P < 0.01.

Fig 7

Engineering infectious FLAG-tagged murine noroviruses. (A) The FLAG epitope tag was inserted into sites in NS4 and VP2. (B) Infectious virus was recovered from NS4-FLAG and VP2-FLAG using the DNA-based reverse genetics system. Significance was tested using one-way analysis of variance and Dunnett's multiple-comparison posttest to compare each insertion virus to the WT control, n = 3. **, P < 0.01. (C) Detection of the FLAG-tagged proteins in infected cell lysates by Western blotting. Detection with an anti-FLAG antibody gave proteins at the expected molecular mass for NS4 and VP2. *, a nonspecific protein with the same molecular mass as NS7 is detected by the anti-NS7 antibody. (D) Immunoprecipitation of NS4 and VP2 via the FLAG tag. RAW264.7 cells were infected with either NS4-FLAG and VP2-FLAG at a low MOI, at 36 hpi cell lysates were prepared, and anti-FLAG agarose was used to immunoprecipitate NS4 and VP2, respectively, as confirmed by Western blot analysis. NS1-2 was coimmunoprecipitated with NS4 only. Lanes M, molecular mass markers.

Fig 8

NS4 and VP2 localize with NS7 in infected cells. Permissive microglial cell line BV-2 was infected with NS4-FLAG, VP2-FLAG, or WT virus at an MOI of 1 or mock infected. At 12 hpi, cells were fixed and dual stained for the FLAG tag and NS7. Both NS4 and VP2 colocalized with NS7 in perinuclear foci.