Generation of Recombinant Rotaviruses Expressing Human Norovirus Capsid Proteins

ABSTRACT

Rotavirus, a segmented double-stranded RNA virus of the Reoviridae family, is a primary cause of acute gastroenteritis in young children. In countries where rotavirus vaccines are widely used, norovirus (NoV) has emerged as the major cause of acute gastroenteritis. Towards the goal of creating a combined rotavirus-NoV vaccine, we explored the possibility of generating recombinant rotaviruses (rRVs) expressing all or portions of the NoV GII.4 VP1 capsid protein. This was accomplished by replacing the segment 7 NSP3 open reading frame with a cassette encoding, sequentially, NSP3, a 2A stop-restart translation element, and all or portions (P, P2) of NoV VP1. In addition to successfully recovering rRVs with modified SA11 segment 7 RNAs encoding NoV capsid proteins, analogous rRVs were recovered through modification of the segment 7 RNA of the RIX4414 vaccine strain. An immunoblot assay confirmed that rRVs expressed NoV capsid proteins as independent products. Moreover, VP1 expressed by rRVs underwent dimerization and was recognized by conformational-dependent anti-VP1 antibodies. Serially passaged rRVs that expressed the NoV P and P2 were genetically stable, retaining additional sequences of up to 1.1 kbp without change. However, serially passaged rRVs containing the longer 1.6-kb VP1 sequence were less stable and gave rise to virus populations with segment 7 RNAs lacking VP1 coding sequences. Together, these studies suggest that it may be possible to develop combined rotavirus-NoV vaccines using modified segment 7 RNA to express NoV P or P2. In contrast, development of potential rotavirus-NoV vaccines expressing NoV VP1 will need additional efforts to improve genetic stability.

IMPORTANCE Rotavirus (RV) and norovirus (NoV) are the two most important causes of acute viral gastroenteritis (AGE) in infants and young children. While the incidence of RV AGE has been brought under control in many countries through the introduction of universal mass vaccination with live attenuated RV vaccines, similar highly effective NoV vaccines are not available. To pursue the development of a combined RV-NoV vaccine, we examined the potential of using RV as an expression vector of all or portions of the NoV capsid protein VP1. Our results showed that by replacing the NSP3 open reading frame in RV genome segment 7 RNA with a coding cassette for NSP3, a 2A stop-restart translation element, and VP1, recombinant RVs can be generated that express NoV capsid proteins. These findings raise the possibility of developing new generations of RV-based combination vaccines that provide protection against a second enteric pathogen, such as NoV.

INTRODUCTION

Rotavirus (RV) and norovirus (NoV) are leading causes of acute gastroenteritis (AGE) in young children and the elderly (1, 2). The introduction of effective RV vaccines—the RV1 monovalent vaccine Rotarix (GSK Biologicals) and the RV5 pentavalent vaccine RotaTeq (Merck and Company)—into the childhood immunization programs of the United States and many other countries has resulted in significant reductions of the incidence of RV hospitalizations and mortality (3, 4). The effectiveness of these vaccines has been correlated, in some instances, with the induction of neutralizing antibodies in immunized children (5–7). In countries where RV vaccines are widely used, NoV has emerged as the primary cause of diarrheal disease and diarrheal-associated hospitalizations in children during the first 5 years of life (8–11). The incidence of severe NoV disease is greatest in young children, in the elderly, and in subjects with compromised immunity (12). NoV is extremely contagious, with fewer than 10 infectious particles able to cause AGE (13). The virus is highly stable in the environment and can be shed from individuals for weeks following infection (13). The development of effective anti-NoV therapeutics or vaccines has been difficult due to the lack of suitable permissive cell lines and validated animal model systems (14, 15).

Ten genogroups (GI to GX) and 49 genotypes of NoV have been identified based on sequence analysis of the NoV capsid protein VP1 (16). Human NoV disease is associated with strains belonging to GI, GII, GIV, GVII, or GIX (17). However, viruses of the GII genogroup are primarily responsible for human NoV disease worldwide; the GII.4 genotype has dominated since the 1990s and accounts for 55 to 85% of all NoV disease (18–20). A high degree of genetic diversity exists within each genogroup. Infections due to strains belonging to one genogroup generally do not confer protection against another genogroup, which is a challenge in the development of universal NoV vaccines (21–25).

The NoV capsid is a nonenveloped icosahedron comprised of 90 dimers of VP1 (Fig. 1) (15, 22, 26). The 7.6-kbp plus-sense RNA [(+)RNA] genome of NoV is organized into three open reading frames (ORF1 to -3). VP1 is the product of ORF2 and, in baculovirus expression systems, self-assembles into virus-like particles (VLPs). Based on its positioning in the NoV capsid, VP1 can be resolved into an interior shell (S) domain and a protruding (P) domain (27, 28). Within genogroups, the S domain is relatively conserved in its primary sequence, whereas the P domain is much more variable. The P domain can be further resolved into P1 and P2 subdomains. The highly variable P2 subdomain represents the immunodominant region of VP1, serving as a target for neutralizing antibodies, and contains the receptor binding site for the NoV capsid (29, 30).

FIG 1.

Domains of the human NoV VP1 capsid protein. (A) VP1 consists of shell (S) and protruding (P) domains (26). The P domain is further resolved into P1 and P2 subdomains. (B) Surface representation of NoV capsid with the S domain (green) and P1 (cyan) and P2 (blue) subdomains of VP1. (C) Ribbon representation of a NoV VP1 dimer: S (green), P1 (cyan), and P2 (blue).

Three distinct types of NoV vaccine candidate have been developed to date: NoV VLPs and P particles and recombinant adenoviruses expressing NoV capsid proteins (25, 31). In general, studies evaluating NoV vaccine candidates have been performed in adults (32–34). However, a phase II trial of a potential GI.1 and GII.4 bivalent VLP vaccine carried out in children and infants showed that the candidate was safe and evoked a robust immune response (35). In attempting to simultaneously prevent both RV and NoV diseases, a trivalent vaccine composed of two types of NoV GII.4 VLPs (GII.4 NO-2010 and GII.4 SYD-2012) and the RV inner capsid protein VP6 was analyzed in a mouse model (36, 37). The results demonstrated significant enhancement in the production of cross-reactive anti-VP1 IgG antibodies following coimmunization of GII.4 VLPs and VP6 (35). In this vaccine formulation, RV VP6 may contribute by serving as an adjuvant, stimulating immune responses to NoV antigens, while at the same time evoking a protective anti-RV immune response (38–40). In immunized animals, the production of VP6 antibodies can also serve as a correlate of protection against RV disease (41).

Recent development of RV reverse genetics systems has resulted in the generation of recombinant RVs that can serve as expression vectors of foreign proteins (42–50). The RV genome consists of 11 segments of double-stranded RNA (dsRNA), with a total size of 18.5 kbp (1). All the segments contain a single ORF encoding a single protein, except for segment 11, which uses frameshifting to express two proteins from a single ORF. Together, the genome segments encode six structural (VP) and six nonstructural (NSP) viral proteins (1). RV segment 7 encodes the 36-kDa protein NSP3, an RNA-binding protein that acts as a translation enhancer of viral (+)RNAs in infected cells (51–53). In our previous studies, we showed that segment 7 can be engineered to express NSP3 and a separate foreign protein by inserting a 2A stop-restart translational element immediately downstream of the NSP3 ORF, followed by the ORF for the foreign protein (47, 48). This approach resulted in the production of recombinant RVs expressing various fluorescent proteins (FPs) (e.g., UnaG [green], mRuby [red], and TagBFP [blue]) from segment 7 (47). Similarly, as a step in exploring the potential usefulness of recombinant RVs as vaccine expression vectors, we showed that segment 7 could be modified to express, as separate products, portions of the severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) spike protein by using 2A translation elements (48). Analysis of recombinant RV expressing foreign proteins showed that they are generally well-growing and genetically stable and continue to produce functional NSP3 capable of dimerization and inducing nuclear localization of the cellular poly(A)-binding protein (PABP) (46, 47). These results raised the possibility of engineering RVs into vector systems that express NoV capsid proteins, enabling the production of combined RV-NoV vaccines that may provide immunological protection against the two most common causes of childhood AGE. In this study, we show that it is possible to generate recombinant RVs that express all or portions (P and P2) of the VP1 capsid protein of a human GII.4 NoV.

RESULTS

Construction of modified segment 7 plasmids encoding NoV capsid proteins.

To evaluate the possibility of using recombinant RVs as expression vectors of NoV capsid proteins, we replaced the NSP3 ORF in the pT7/NSP3SA11 plasmid with a cassette that sequentially encoded NSP3, a flexible linker (GAG), a porcine teschovirus 2A translation element (54), and the NoV P2, P, or VP1 protein (Fig. 2). Some cassettes included sequences specifying a 3×FLAG tag at the N terminus or a 1×FLAG tag at the C terminus of the NoV capsid protein (Fig. 2). Instead of FLAG tags, some cassettes encoded a NoV capsid protein with a C-terminal 6×His tag or a thrombin (LVPRGS) cleavage site (55) followed by a C-terminal 6×His tag (Fig. 2). This process yielded a set of pT7/SA11NSP3-2A-NoV vectors designed to express separately NSP3 and FLAG- or His-tagged P2 (pT7/NSP3-2A-fP2), P (pT7/NSP3-2A-fP or pT7/NSP3-2A-Ph), or VP1 (pT7/NSP3-2A-fVP1, pT7/NSP3-2A-VP1f, pT7/NSP3-2A-VP1th, or pT7/NSP3-2A-VP1h). The NoV capsid sequences were introduced into the pT7/NSP3SA11 vector at the same site used previously for making recombinant SA11 (rSA11) rotaviruses that expressed FPs and domains of the SARS CoV-2 spike protein (47, 48).

FIG 2.

Plasmids with modified segment 7 cDNAs used to generate recombinant RVs expressing NoV capsid proteins. Schematic indicates nucleotide positions of the coding sequences for NSP3, 2A element (2A), 3×FLAG (3FL), 1×FLAG (1FL), 6×histidine (His) tag, thrombin cleavage site (Th), and VP1, P2, and P. The red arrow notes the position of the 2A translational stop-restart site, and the asterisk notes the end of the ORF. Sizes (in amino acids [aa]) of encoded NSP3 and NoV VP1 products are given in parentheses. T7, T7 RNA polymerase promoter sequence; Rz, hepatitis D virus ribozyme; UTR, untranslated region.

To test the possibility that segment 7 RNAs of RV strains, other than SA11, could serve as expression platforms, we generated a RIX4414-like (RIX)-based segment 7 plasmid (pT7/NSP3RIX-2A-Ph) in which the NSP3 ORF had been replaced with a coding cassette for NSP3-GAG-2A-NoV P-6×His (Fig. 2). RIX4414 is a human RV G1P[8] strain that is used in formulating the Rotarix vaccine (3, 4).

Recovery of rSA11 rotaviruses containing NoV coding sequences.

Recombinant viruses were generated by transfecting BHK-T7 cells with 11 pT7/SA11 plasmids, each directing the synthesis of a viral (+)RNA, and pCMV-NP868R, a vector expressing the African swine fever virus capping enzyme. In the transfection procedure, the pT7/NSP3SA11 plasmid was replaced with a pT7/SA11NSP3-2A-NoV or pT7/NSP3RIX-2A-Ph plasmid. The pT7/NSP2SA11 and pT7/NSP5SA11 plasmids were included in transfection mixtures at levels 3-fold greater than the other plasmids (43). Transfected BHK-T7 cells were overseeded with MA104 cells 2 days posttransfection. Three days later, cell mixtures were freeze-thawed, and recombinant viruses in the lysates were grown by passage on MA104 cells. Recombinant viruses were recovered from virus pools by plaque selection and grown to larger volumes prior to characterization. Properties of rSA11 viruses expressing FLAG-tagged NoV capsid proteins, including their genome profiles, peak titers, plaque morphologies and titers, and NoV products, are provided in Fig. 3. Analogous properties of rSA11 viruses expressing 6×His-tagged NoV proteins are provided in Fig. 4, and properties of rSA11 viruses that use RIX segment 7 RNAs to express NoV capsid protein are presented in Fig. 5. Additional information about the recombinant viruses, including the sizes of their modified segment 7 RNAs and their complete genomes, is given in Table 1.

FIG 3.

Recombinant viruses expressing FLAG-tagged NoV capsid proteins. (A) dsRNAs were recovered from rSA11-infected MA104 cells, resolved by gel electrophoresis, and detected by ethidium bromide staining. The genome segments of rSA11-wt are labeled 1 to 11. Sizes (in kilobase pairs) of modified segment 7 RNAs (black arrows) are indicated. (B) Plaques were detected by crystal violet staining. (C) Mean diameters of 22 plaques produced by rSA11 viruses at 5 days p.i., with 95% confidence intervals indicated (black lines). Significance values were calculated using an unpaired Student's t test. **, P < 0.01; ***, P < 0.001; ****, P < 0.0001. (D) Peak titers reached by rSA11 viruses in MA104 cells were determined by plaque assay (reported in plaque-forming units [PFU]). Data represent the results of three independent experiments and are means ± standard deviations with 95% confidence intervals. Significance values were calculated using Welch's t test. ns, not significant; *, P < 0.1 (E) An immunoblot assay was used to examine mock-infected and rSA11-infected cells for the presence of NoV VP1, P, or P2 (α-FLAG antibody), NSP3, or NSP3-2A (α-NSP3), NSP3-2A (α-2A), VP6, and β-actin. Readthrough products, resulting from lack of 2A peptide function, are indicated with a red asterisk. Sizes (in kilodaltons [kDa]) of protein molecular weight markers (MWM) are indicated.

FIG 4.

Recombinant viruses expressing His-tagged NoV capsid proteins. (A) dsRNAs were recovered from rSA11-infected MA104 cells, resolved by gel electrophoresis, and detected by ethidium bromide staining. RNA segments of rSA11-wt are labeled 1 to 11. Sizes (in kilobase pairs [kbp]) of modified segment 7 RNAs (black arrows) are indicated. (B) Viral plaques assays were detected by crystal violet staining. (C) Mean diameters of 20 plaques produced by rSA11 viruses at 5 days p.i., with 95% confidence intervals indicated (black lines). (D) Peak titers reached by rSA11 viruses in MA104 cells were determined by plaque assay (plaque-forming units [PFU]). Values represent the averages of three independent experiments, with error bars showing standard deviations. Significance values were calculated as described for Fig. 3. (E) Immunoblot assay was used to examine mock- and rSA11-infected cells for the presence of NoV VP1 or P (α-His antibody), NSP3 or NSP3-2A (α-NSP3), NSP3-2A (α-2A), VP6, and β-actin. Sizes (in kilodaltons [kDa]) of protein molecular weight markers (MWM) are indicated.

FIG 5.

Recombinant RV expressing NoV P from a modified RIX segment 7 RNA. (A) dsRNA was recovered from MA104 cells infected with rSA11-wt, rSA11(RIX-NSP3)-Ph viruses, or plaque isolates (Plq1 to Plq3) of rSA11(RIX-NSP3)-Ph, resolved by gel electrophoresis, and detected by ethidium bromide staining. RNA segments of rSA11-wt are labeled 1 to 11. Sizes (in kilobase pairs [kbp]) of segment 7 RNAs (black arrows) are indicated. (B) Viral plaques were detected at 5 days p.i. by crystal violet staining. (C) Peak titers reached by rSA11 viruses in MA104 cells were determined by plaque assay (plaque-forming units [PFU]). Values represent the averages of three independent experiments, with error bars showing standard deviations. (D) An immunoblot assay was used to examine mock- and rSA11-infected cells for the presence of NoV P (α-His antibody), NSP3 (SA11 α-NSP3), NSP3-2A (α-2A), VP6, and β-actin. Readthrough products, resulting from lack of 2A peptide function, are indicated with a red asterisk. Sizes (in kilodaltons [kDa]) of protein molecular weight markers (MWM) are indicated.

TABLE 1.

Properties of recombinant rSA11 strains expressing NoV capsid proteins

Virus strain			Genome segment 7
Abbreviated virus name	Formal virus namea	Genome size/increase over wt (bp)	RNA (bp)	Protein product size				NCBI accession no.
				Uncleaved (aa)	2A cleaved (aa)	Uncleaved (kDa)	2A cleaved (kDa)
rSA11-wt	RVA/Simian-lab/USA/SA11wt/2019/G3P[2]	18,559/0	1,105	315	NDb	36.4	ND	LC178572
rSA11-fP2	RVA/Simian-lab/USA/SA11(NSP3-2A-NoV:fP2)/2019/G3P[2]	19,126/567	1,672	504	336 + 168	57.1	38.5 + 18.6	MN190002
rSA11-fP	RVA/Simian-lab/USA/SA11(NSP3-2A-NoV:fP)/2019/G3P[2]	19,642/1,083	2,188	676	336 + 340	76.1	38.5 + 37.7	MN190003
rSA11-fVP1	RVA/Simian-lab/USA/SA11(NSP3-2A-NoV:fVP1)/2019/G3P[2]	20,314/1,755	2,860	900	336 + 564	100.3	38.5 + 61.9	MN190004
rSA11-VP1f	RVA/Simian-lab/USA/SA11(NSP3-2A-NoV:VP1f)/2019/G3P[2]	20.272/1,713	2,818	886	336 + 550	98.6	38.5 + 60.1	MN201548
rSA11-Ph	RVA/Simian-lab/USA/SA11(NSP3-2A-NoV:P His)/2019/G3P[2]	19,594/1,035	2,140	660	336 + 324	74.2	38.5 + 35.8	MN201549
rSA11-VP1th	RVA/Simian-lab/USA/SA11(NSP3-2A-NoV:VP1ThHis)/2019/G3P[2]	20,278/1,719	2,824	888	336 + 552	98.8	38.5 + 60.4	MN201547
rSA11-VP1h	RVA/Simian-lab/USA/SA11(NSP3-2A-NoV:VP1 His)/2019/G3P[2]	20,260/1,701	2,806	882	336 + 546	98.2	38.5 + 59.8	MZ562305
rSA11(RIX-NSP3)-Ph	RVA/Simian-lab/USA/SA11(RIX NSP3-2A-NoV:P His)/2021/G3P[2]	19,563/1,004	2,109	655	331 + 324	74.2	38 + 35.8	MZ643978

^aFormal strain names were assigned according to the classification of Matthijnssens et al. (76).

^bND, not determined (no 2A cleavage site was present).

As expected, rSA11 viruses generated with pT7/NSP3-2A-NoV plasmids (subsequently referred to as rSA11-NoV viruses) contained segment 7 RNAs that were much larger than the segment 7 RNA of wild-type rSA11 virus (Fig. 3A and 4A). Sequence analysis confirmed that the segment 7 RNAs of the rSA11-NoV viruses matched those of the pT7/NSP3-2A-NoV plasmids. The addition of NoV P2 and P sequences into the 1.1-kbp segment 7 RNAs increased their total sizes to 1.7 kbp and 2.1 kbp, respectively (Table 1; Fig. 3A and 4A). Correspondingly, addition of NoV VP1 sequences to the segment 7 RNA increased its total size to 2.8 to 2.9 kbp, which migrated electrophoretically near genome segment 1. Insertion of NoV VP1 sequences into the segment 7 RNA increased the total size of the viral genome to 20.3 kbp, ~9.5% greater than the size of the wild-type rSA11 genome (Table 1). The 2.8- to 2.9-kbp segment 7 RNAs of rSA11 viruses containing NoV VP1 sequences were smaller than the 3.3-kbp segment 7 RNA of the rSA11-fS1 virus, which contains the SARS-CoV-2 S1 coding sequence (48).

Analysis of recombinant viruses recovered using the pT7/NSP3RIX-2A-Ph plasmid confirmed that it was possible to use the RIX segment 7 RNA to produce viruses containing coding sequences for the NoV capsid protein (Fig. 5A). The rSA11(RIX)NSP3-Ph virus contained a 2.1-kbp segment 7 RNA instead of the wild-type 1.1-kbp segment 7 RNA (Table 1).

Plaque analysis showed that plaques formed by the rSA11-NoV and rSA11(RIX-NSP3)-Ph viruses were smaller than those of wild-type rSA11 (Fig. 3B and C, Fig. 4B and C, and Fig. 5B). This was consistent with previous studies showing that recombinant RVs with modified segment 7 RNAs expressing FPs or SARS-CoV-2 proteins have a small plaque phenotype (47, 48). Although all the rSA11-NoV and rSA11(RIX-NSP3)-Ph viruses appeared to grow to peak titers in MA104 cells that were lower than those in with wild-type rSA11 virus, the titer reduction was found to be statistically significant only in some cases (Fig. 3D, 4D, and 5C).

Recombinant viruses express NoV P2, P, or VP1 as separate products.

To examine NoV protein products made by rSA11-NoV and rSA11(RIX)NSP3-Ph viruses, lysates prepared at 9 h postinfection (p.i.) from MA104 infected cells were probed by immunoblot assay using FLAG antibody (Fig. 3E). The anti-FLAG immunoblots showed that rSA11-fP2, -fP, -fVP1, and -VP1f viruses directed the expression of NoV proteins of the size predicted for a functional 2A element in the modified NSP3 ORF: -fP2 (18.6 kDa), -fP (37.7 kDa), -fVP1 (61.9 kDa), and -VP1f (60.1 kDa) (Table 1). Assays with anti-NSP3 antibody and 2A element antibody identified a 38-kDa protein, which corresponded to NSP3 (36 kDa) linked to remnant residues of the 2A peptide (2 kDa) (47, 54). Anti-FLAG immunoblot assay also detected minor amounts of large proteins, which likely represented readthrough products of NSP3-2A-NoV protein cassettes (NSP3-2A-fP2, NSP3-2A-fP, and NSP3-2A-fVP1) generated when the 2A element fails to function (Fig. 3E, red asterisks). Such readthrough product was not detected for the rSA11-VP1f virus, suggesting that direct fusion of the VP1 ORF to the upstream 2A peptide sequence perhaps increases the efficiency of 2A activity (Fig. 3E).

Immunoblot assays performed with antibody specific for a 6×His epitope showed that the rSA11-Ph virus and the rSA11-VP1h and rSA11-VP1th viruses generated 36-kDa P and 60-kDa VP1 products, respectively, sizes consistent with the presence of functional 2A peptides in the expression cassettes (Fig. 4E, Table 1). As noted above for rSA11-VP1f (Fig. 3E), the 6×His antibody did not detect readthrough products (NSP3-2A-Ph, NSP3-2A-VP1h, and NSP3-2A-Pth) (Fig. 4E), possibly because placement of NoV P and VP1 coding sequences immediately downstream of the 2A peptide sequence, without intervening tags or spacer sequences, may have improved the efficiency of 2A activity. Mirroring results described above for viruses expressing FLAG-tagged NoV protein products (Fig. 3E), immunoblot assays performed with antibodies for NSP3 and the 2A peptide revealed the expression of the 38-kDa NSP3-2A protein in MA104 cells infected with rSA11-Ph, -VP1h, and -VP1th (Fig. 4E).

Immunoblot assays with the anti-6×His antibody showed that the 36-kDa P protein was a dominant product of the rSA11(RIX)NSP3-Ph virus (Fig. 5D). A smaller amount of an NSP3-2A-PHis readthrough product was also detected with anti-6×His antibody. Due to lack of cross-reactivity, immunoblot probes with antibody raised against SA11 NSP3 failed to detect RIX NSP3 (Fig. 5D). However, reprobing the same blot with antibody specific for the 2A peptide identified expression of a 38-kDa protein, likely representing RIX NSP3 protein fused to 2A peptide (Fig. 5D). Overall, these results showed that it is possible to use segment 7 RNAs of RV strains other than SA11, including the vaccine RIX4414 strain, to direct the expression of NoV capsid proteins.

Dimerization of NoV VP1 expressed by recombinant viruses.

VP1 dimers serve as intermediates in the assembly of the NoV capsid (56). To address whether NoV proteins expressed from recombinant viruses formed dimers, lysates from rSA11-fP2, -fP, -fVP1, and -VP1f infected cells were incubated in sample buffer at a temperature in which VP1 dimers are stable (25°C) and at a temperature in which the dimers are disrupted (95°C). The samples were analyzed by non-denaturing gel electrophoresis and immunoblot assay using anti-FLAG antibody (Fig. 6). The results indicated that neither separately expressed P2 nor P formed stable dimers. In contrast, expressed VP1 proteins, with FLAG tags at the N or C terminus (fVP1 and VP1f), migrated with the expected size of VP1 dimers (Fig. 6A). Under the same electrophoretic conditions used to analyze NoV capsid proteins for dimerization, cells infected with rSA11-wt and rSA11-fP2, -fP, -fVP1, and -VP1f generated NSP3 dimers and VP6 trimers (Fig. 6A), naturally occurring multimeric forms of these proteins (57, 58). Similarly, 6×His-tagged VP1 proteins (VP1h and VP1th) expressed by rSA11-VP1h and rSA11-VP1th also formed dimers (Fig. 6B). As above, NSP3 and VP6 proteins formed stable dimers and trimers in rSA11-VP1h- and rSA11-VP1th-infected cells, respectively (Fig. 6B).

FIG 6.

Dimerization of NoV capsid proteins expressed by recombinant viruses. (A) MA104 cells were mock infected or infected with rSA11-wt, -fP2, -fP, -fVP1, or -VP1f and harvested at 9 h p.i. Cell lysates were mixed with SDS-sample buffer, incubated for 10 min at 25 or 95°C, and subjected to non-denaturing electrophoresis on 4% to 20% polyacrylamide gels. Resolved proteins were blotted onto nitrocellulose membranes and detected using antibodies specific for FLAG, SA11 NSP3, VP6, or β-actin. Primary antibodies were detected using HRP-conjugated secondary antibodies. Sizes (in kilodaltons [kDa]) of protein markers (MWM) are indicated. (B) Similar experiment as in panel A, except cells were infected with rSA11-wt, -Ph, -VP1h, or -VP1th. Blots were probed with antibodies specific for 6×His instead of FLAG.

Folding of NoV VP1 capsid proteins into native structures.

To gain insight into whether the VP1 products expressed by recombinant viruses folded into native structures, lysates prepared from MA104 cells infected with rSA11-fVP1, -VP1f, and -VP1h viruses were probed by pulldown assay using a conformation-dependent neutralizing monoclonal antibody that recognized GII.4 VP1 (NVB43.9). As shown in Fig. 7 (red arrows), the anti-VP1 antibody immunoprecipitated both FLAG and 6×His-tagged NoV VP1 proteins (fVP1 and VP1h), indicating that VP1 expressed by rSA11 viruses folded in a conformation that included an authentic neutralizing epitope found in the NoV VP1 protein. Combined with the observation that rSA11-expressed VP1 formed dimers (Fig. 6), these data suggested that at least some VP1 produced in cells infected with rSA11-fVP1 and -VP1h folded into native structures. Unlike the successful pulldown of fVP1 and VP1His with anti-GII.4 VP1 antibody, it was not certain whether the antibody likewise immunoprecipitated the VP1f product of rSA11-VP1f (Fig. 7). This may in part have resulted from the lower level of VP1f expression in infected cells compared to that of fVP1 and VP1His, or from differences in the affinity of the FLAG antibody for the N-terminal 3×FLAG tag for fVP1 versus the C-terminal 1×FLAG tag for VP1f (Fig. 7).

FIG 7.

Antibody recognition of a neutralizing epitope in expressed NoV VP1. Whole-cell lysates (WCL) were prepared from MA104 cells infected with the indicated recombinant viruses (rSA11-wt, -fVP1, -VP1f, and -VP1h) and incubated with a monoclonal antibody specific for NoV GII.4 VP1 (NVB43.9). Antigen-antibody complexes were recovered using magnetic IgA/G beads, resolved along with WCL by gel electrophoresis, and blotted onto nitrocellulose membranes. Blots were probed with anti-FLAG/6×His antibodies to detect immunoprecipitated VP1 and with anti-NSP3, -VP6, and -β-actin antibodies. Arrows indicate immunoprecipitated VP1 proteins. Ig/L, Ig light chain; Ig/H, Ig heavy chain. Positions of molecular weight markers (MWM) in kilodaltons (kDa) are indicated.

Genetic stability of rSA11 strains expressing NoV proteins.

To analyze the genetic stability of rSA11 viruses expressing NoV capsid proteins, the viruses were subjected to 5 rounds of serial passage at three dilutions (1:10, 1:100, and 1:1,000). Analysis of the dsRNAs recovered from cells infected with the rSA11-fP2, -fP, and -Ph viruses showed no changes in the sizes of any of the genome segments, including the modified genome segment 7 RNA, over 5 rounds of serial passage (P1 to P5) (Fig. 8A). Thus, rSA11 viruses carrying up to 1.1 kbp of foreign sequence were genetically stable, consistent with the findings of a previous study (47, 48). In contrast, serial passage of rSA11-fVP1 and -VP1h showed evidence of genetic instability during serial passage at all three dilutions (Fig. 8B and C). Notably, new genome segments were visible by the third round of passage that were smaller (Fig. 8B and C, red arrows) than the initial 2.8- to 2.9-kbp segment 7 RNAs of rSA11-fVP1 and -VP1h (black arrowheads). Indeed, the 2.8- to 2.9-kbp segment 7 RNAs of these viruses were no longer detectable by passage 5. Instead, the ~1.2-kbp variant RNA became a dominant genome segment for rSA11-fVP1 and -VP1h by passage 5, replacing the 2.8- to 2.9-kbp segment 7 RNA (Fig. 8B and C, red arrows).

FIG 8.

Genetic stability of rSA11 viruses expressing NoV proteins. (A to C) rSA11 viruses were serially passaged 5 times (P1 to P5) in MA104 cells with inoculum prepared from infected cell lysates diluted 1:10, 1:100, or 1:1,000 in DMEM. RNAs were recovered from infected cell lysates by TRIzol extraction and analyzed by gel electrophoresis. Positions of viral genome segments are labeled. Positions of modified segment 7 (NSP3) dsRNAs introduced into rSA11 strains are denoted with black arrows. In some cases, duplicated bands are present near segment 10 and 11 RNAs. These are likely artifacts or cellular RNA, as viruses were not recovered by plaque isolation that contained the duplicated bands instead of the expected segment 10 and 11 RNAs. Novel RNAs appearing in the RNA population during serial passage are indicated with red arrows. (D and E) Genomic RNAs of virus isolates recovered by plaque isolation from P5 virus pools, with large (L) and small (S) plaque phenotypes. Novel (rearranged) segment 7 RNAs are denoted in red (fVP1/R1, fVP1/R2) or blue (VP1h/R) arrows. (F) Organization of the fVP1/R2 segment 7 RNA of the L1 plaque isolate. Dashed lines represent sequences deleted from the starting segment 7 RNAs. Features identified include the NSP3 ORF (orange), 2A peptide (red), VP1 ORF (blue-green), and 3×FLAG tag (3FL). The red arrow indicates the 2A cleavage site. Nucleotide positions correspond to those of the starting nonrearranged segment 7 RNAs.

To understand the origin of the ~1.2-kbp variant RNA, plaque isolation was used to recover five viruses each from the P5 virus pools of rSA11-fVP1 and -VP1h, four with a large (L) plaque phenotype and one with a small (S) plaque phenotype. Gel electrophoresis showed that none of the plaque isolated viruses contained the 2.8- to 2.9-kbp segment 7 RNAs of the unpassaged virus (Fig. 8D and E). Rather, the L1 to L4 isolates of the rSA11-fVP1 P5 pool contained the fVP1/R2 RNA segment (R, rearranged), whereas the S1 isolate contained both fVP1/R1 and fVP1/R2 RNA segments (Fig. 8D). Gel electrophoresis indicated that the large (L1 and L3 to L5) and small (S1) plaque isolates of the rSA11-VP1h P5 pool contained only a single type of variant RNA, VP1h/R (Fig. 8E). The fact that all five virus isolates from the P5 pool of the rSA11-fVP1 and rSA11-VP1h viruses contained variant 7 RNAs that were much smaller than the initial 2.8- to 2.9-kbp segment 7 RNAs of these viruses, suggests that smaller RNAs provide a growth advantage (Fig. 8D and E). This is consistent with a previous report suggesting that selective pressures exist during rotavirus replication that favor minimizing segment size (48).

Sequence analysis revealed that the fVP1/R2 RNA originated from the 2.9-kbp segment 7 RNA of rSA11-fVP1 (Fig. 8F). While the fVP1/R2 RNA (1,262 bp) contained the complete 5′-untranslated region (UTR) and NSP3 ORF of segment 7, it lacked 1.6 kbp of the NoV VP1 coding sequence and 7 bp of the 3′-UTR. Although fVP1/R2 contained a near-complete deletion of the NoV VP1 ORF, the RNA retained the complete sequence of the NSP3 ORF, suggesting that NSP3 likely has an essential function in viral replication. Further analysis of the total population of viral RNAs in serial passaged virus may provide better insight into mechanisms governing the deletion of VP1 sequences from rSA11-VP1 viruses.

DISCUSSION

Through modification of the RV segment 7 RNA, we generated rSA11 viruses that expressed all (VP1) or portions (P, P2) of the major capsid protein of the NoV GII.4 MD145 isolate (59). Viruses of the GII.4 genotype are the most common cause of NoV illness globally (19). Given that VP1, P, and P2 contain immunodominant epitopes and are targeted by neutralizing NoV antibodies (29, 30), recombinant RVs expressing these proteins may be able to trigger immunoprotective responses to both RV and NoV in the immunized host. In previous studies, VLPs assembled from baculovirus-expressed VP1 were shown to induce the formation of blocking (neutralizing) antibodies in animals (60). Although we do not know whether rSA11-expressed VP1 likewise assembled into VLPs, rSA11-expressed VP1 was found to form dimers and was recognized by a conformationally dependent VP1 neutralizing antibody. Thus, rSA11-expressed VP1 may fold into native structures able to stimulate the production of neutralizing antibodies and generate immunoprotective responses in animals. Based on the dimerization and immunoprecipitation assays performed in this study, the P and P2 proteins expressed by rSA11 viruses appear not to fold into structures that mimic the P and P2 protruding domains of the NoV capsid. However, through appropriate genetic engineering of the terminal sequences of the expressed P protein, it may be possible to generate modified forms of P that self-assemble into small (12-mer) or large (24-mer) P particles (61, 62). Such particles have been shown to induce NoV-specific immunoprotective responses (63).

Recombinant rSA11 viruses expressing VP1, P, or P2 were generated though replacement of the NSP3 ORF in the segment 7 RNA with an ORF sequentially encoding NSP3, a 3-residue flexible (GAG) linker, a 19-amino acid (aa) porcine teschovirus 2A translation element, and a FLAG or 6×His tagged VP1, P, or P2 protein. Because none of the viral ORFs of the rSA11 viruses were altered except that of NSP3, these viruses are expected to produce the same complement of proteins as wild-type virus. Although NSP3s of rSA11s expressing NoV proteins contained an extra 19 aa at their C terminus (NSP3-2A), representing remnant residues of the cleaved 2A peptide (54), the NSP3-2A product was able to dimerize. This finding is consistent with previous results showing that the NSP3 product of similarly constructed rSA11 viruses expressing fluorescent proteins also dimerizes (47). Moreover, NSP3 with 2A remnant residues also retains the ability to induce nuclear accumulation of PABP (47). The fact that NSP3 functions are not affected by remnant 2A residues may be expected given that NSP3 of group C rotavirus is naturally expressed with a downstream 2A peptide (64). This allows the group C NSP3 genome segment to express an additional protein (double-stranded RNA-binding protein, dsRBP). In our analysis of rSA11 viruses, we found that teschovirus 2A elements were efficient in promoting the expression of NoV capsid proteins as independent products. In some cases, low levels of readthrough products (NSP3-2A-NoV protein) were detected, due to inefficient 2A function. In general, readthrough products were more noticeable when antibody tags, or extra amino acids, were positioned between the 2A element and NoV protein sequence. This is consistent with earlier reports indicating that 2A activity can be affected by the nature of upstream and downstream residues (65).

The maximum amount of foreign sequence that can be inserted into the RV genome has not been established (66). However, naturally occurring RV strains have been isolated that contain duplications of nearly 1.0 kbp in their segment 7 RNA, bringing the RNA’s total size to ~2.0 kbp (67, 68). In this study, generation of the rSA11-fVP1 virus required insertion of 1.8 kbp of foreign sequence into the 1.1-kbp segment 7 RNA, bringing the RNA’s total size to 2.9 kbp, well in excess of segment 7 RNAs of viruses with natural sequence duplications. The total size of the genome of rSA11-fVP1 was 20.3 kbp. Although this significantly exceeded the 18.5-kbp genome of wild-type SA11, larger recombinant RVs have been recovered. For example, a recombinant rSA11 virus has been generated with a 2.2-kbp segment 7 RNA that expresses the S1 protein of SARS-CoV-2 S1 (48).

Recombinant RVs with large segment 7 sequence insertions have a small plaque phenotype and grow to titers lower than those for wild-type virus, suggesting that the extra sequence imposes a limitation on some aspect of viral growth. The limitation may stem from the longer elongation time likely required for the viral RNA polymerase to transcribe the modified segment 7 dsRNA or the longer time likely required for translational machinery to translate the extended ORF of modified segment 7 (+)RNA. Alternatively, the limitation may result from a decreased efficiency in the assortment or packaging of the segment 7 (+)RNA. Certainly, the time required for the viral RNA polymerase to replicate the longer RNA into dsRNA is likely increased. Interestingly, rSA11 viruses expressing NoV VP1 not only had a small plaque phenotype, but also the segment 7 RNAs of these viruses were genetically unstable. During 2 to 3 rounds of serial passage, the 1.6-kbp VP1 sequence insertion in the segment 7 RNAs was lost, giving rise to variant virus populations that outcompeted the parental strains. Molecular mechanisms driving deletion of the VP1 sequences are not known, but these findings clearly indicate that rSA11 viruses with 1.6-kbp VP1 sequence insertions are under considerable pressure either to delete the foreign sequence or to reduce the overall size of the viral genome. The instability results observed here with rSA11 viruses expressing NoV VP1 are similar to those observed previously with rSA11 viruses expressing the SARS-CoV-2 S1 protein (48). In that case, the modified segment 7 RNAs of the viruses were found to delete most, if not all, of their 2.2-kbp S1 sequence insertions within a few rounds of serial passage.

In contrast to rSA11 viruses with 1.6-kbp VP1 sequence insertions, rSA11 viruses containing 1.1-kbp fP sequences or 0.6-kbp fP2 sequences were genetically stable over 5 rounds of serial passage. Likewise, rSA11 viruses with modified segment 7 RNAs containing sequences of the SARS-CoV-2 spike gene of up to 1.5 kbp were found to be genetically stable (48). Thus, recombinant RVs with foreign sequence insertions in segment 7 that do not exceed 1.5 kbp may be sufficiently genetically stable for developing vaccine candidates. The coding capacity provided by 1.5 kbp of foreign sequence is sufficient to generate recombinant RVs that express the NoV P protein (~35 kDa) or modified forms of NoV P with localization or affinity tags that may increase immunogenicity potential. For example, inclusion of affinity ligands for Fc-immunoglobulin G1 (Fc-IgG1) or for the neonatal cell surface (FcRn) receptor may help to direct expressed P protein that will promote antigen recognition and presentation (69–71).

This work represents further illustrates the potential usefulness of RV as an expression platform for the capsid protein of another pathogenic virus. In an earlier study, through similar modification of the segment 7 RNA, we showed that it was possible to generate recombinant RVs that expressed regions of the SARS-CoV-2 spike protein, including the immunodominant S1 region. In unpublished work, we have also generated rSA11 viruses that express portions of the capsid proteins of astrovirus and hepatitis E virus (reference 72 and data not shown). Importantly, we found in the current study that the segment 7 RNAs of other RV strains, such as the RIX4414 (G1P[8]) vaccine strain, can be engineered to serve as expression platforms of nonrotaviral proteins. Given that reverse genetics systems have been developed for human G1P[8] and G4P[8] RVs, it seems likely that RV strains that formulate human RV strains can be developed into combination vaccines that target other enteric or mucosal viruses. With the development of reverse genetics systems for animal RV strains, it may be possible to develop similar combination vaccines for use in livestock and farm animals.

Our studies indicate that recombinant RVs can be used as vector systems for the expression of the capsid proteins of other enteric and mucosal viruses (references 48 and 72 and the current study), providing a possible path for developing combination vaccines that induce immunological protection against more than one pathogen. Studies evaluating whether foreign capsid proteins expressed by recombinant RVs can trigger the production of neutralizing antibodies targeting another virus are key to the development of combination vaccines. Although such studies are only beginning, a preliminary report by Kawagishi et al. (73) provided evidence that recombinant RVs using modified segment 7 RNAs to express NoV capsid proteins are cable of inducing NoV blocking (pseudo-neutralizing) antibodies in infant mice, a promising early result.

MATERIALS AND METHODS

Cell culture.

Embryonic monkey kidney (MA104) cells were grown in Dulbecco’s modified eagle medium (DMEM) containing 5% fetal bovine serum (FBS) and 1% penicillin-streptomycin (74). Baby hamster kidney cells constitutively expressing T7 RNA polymerase (BHK-T7) were provided by Ulla Buchholz, Laboratory of Infectious Diseases, NIAID, NIH, and were propagated in Glasgow minimum essential media (GMEM) containing 5% heat-inactivated FBS, 10% tryptone-peptide broth, 1% penicillin-streptomycin, 2% nonessential amino acids, and 1% glutamine (75). BHK-T7 cells were grown in a medium supplemented with 2% Geneticin (Invitrogen) in every other passage.

Plasmid construction.

rSA11 viruses were prepared using the plasmids pT7/VP1SA11, pT7/VP2SA11, pT7/VP3SA11, pT7/VP4SA11, pT7/VP6SA11, pT7/VP7SA11, pT7/NSP1SA11, pT7/NSP2SA11, pT7/NSP3SA11, pT7/NSP4SA11, and pT7/NSP5SA11 (https://www.addgene.org/Takeshi_Kobayashi/) and pCMV/NP868R (42). The plasmid pT7/NSP3-2A-fUnaG was produced by fusing a DNA fragment containing the ORF for 2A-3×FL-UnaG to the 3′-end of the NSP3 ORF of pT7/NSP3SA11 by using a TaKaRa In-Fusion cloning kit. A plasmid (pUC57/MDA145_VP1) containing a full-length cDNA of the VP1 gene of the NoV GII.4 MD145-12 strain (GenBank accession number AY032605.1) was purchased from Genewiz. The plasmids pT7/NSP3-2A-fP2, pT7/NSP3-2A-fP, and pT7/NSP3-2A-fVP1 were made by replacing the UnaG ORF in pT7/NSP3-2A-fUnaG with ORFs for the P2, P, and VP1 regions, respectively, of the NoV VP1 capsid protein, by In-Fusion cloning. The backbone of the plasmids was generated through PCR amplification of pT7/NSP3-2A-fUnaG with the primer pairs Vector_For and Vector_Rev (Table 2). DNA fragments containing P2, P, and VP1 coding sequences were amplified from pUC57/MDA145_VP1 using the primer pairs fP2_For and fP2_Rev, fP_For and fP_Rev, and fVP1_For and fVP1_Rev, respectively (Table 2). The plasmids pT7/NSP3-2A-VP1f, pT7/NSP3-2A-Ph, pT7/NSP3-2A-VP1h, and pT7/NSP3-2A-VP1th were constructed in a similar way. The backbone of the plasmids was generated by amplifying pT7/NSP3-2A-fUnaG with the primer pairs Vector P2A_For and Vector P2A_Rev (Table 2). DNA fragments containing VP1 with a C-terminal FLAG or P or VP1 with a C-terminal His tag were produced through PCR amplification of pUC57/MDA145_VP1 with the primer pairs VP1-fFor and VP1-fRev, P-His_For and P- His_Rev, VP1-ThHis_For and VP1-ThHis_Rev, and VP1-His_For and VP1-His_Rev, respectively (Table 2). A puc19 plasmid containing a RIX/NSP3-2A-P-His insert under the control of a T7 transcription promoter (puc19/T7/RIX/NSP3-2A-Ph) was purchased from Bio Basic Canada Inc. Transfection quality plasmids were prepared commercially (www.plasmid.com) or by using Qiagen plasmid purification kits. Primers were provided by and sequences determined by EuroFins Scientific.

TABLE 2.

Primers used to produce pT7/NSP3-2A-NoV plasmids

Primer name	Sequence
Vector_For	CCATTTTGATACATGTTGAACAATCAAATACAGTGT
Vector_Rev	GCTAGCCTTGTCATCGTCATCCT
fP2_For	GATGACAAGGCTAGCACTACCCAGCTGTCAGCTG
fP2_Rev	CATGTATCAAAATGGTCACAGGTGCACATTATGACCAGTTCT
fP_For	GATGACAAGGCTAGCAAACCATTCACCGTCCCAATCT
fP_Rev	CATGTATCAAAATGGTTATAATGCACGCCTGCGCCC
fVP1_For	GATGACAAGGCTAGCATGAAGATGGCGTCGAGTGAC
fVP1_Rev	CATGTATCAAAATGGTTATAATGCACGCCTGCGCCC
Vector P2A_For	CCATTTTGATACATGTTGAACAATCAAATACAGTGT
Vector P2A_Rev	AGGACCGGGGTTTTCTTCCAC
VP1-fFor	GAAAACCCCGGTCCTGCTAGCGTGAAGATGGCGTCGAG
VP1-fRev	CATGTATCAAAATGGTTACTTGTCATCGTCATCCTTGTAATCTAATGCACGCCTGCGC
P- His_For	AGAAAACCCCGGTCCTGCTAGCAAACCATTCACCGTCC
P-His_Rev	CATGTATCAAAATGGTTAGTGGTGGTGGTGGTGGTGTAATGCACGCCTGCGCC
VP1-ThHis_For	GAAAACCCCGGTCCTGTGAAGATGGCGTCGAGTGAC
VP1-ThHis_Rev	ACATGTATCAAAATGGTTAGTGGTGGTGGTGGTGGTGCGAGCCACGGGGGACCAATAATGCACGCCT
VP1-His_For	GAAAACCCCGGTCCTGTGAAGATGGCGTCGAGTGAC
VP1-His_Rev	CATGTATCAAAATGGTTAGTGGTGGTGGTGGTGGTGTAATGCACGCCT

Isolation and amplification of recombinant viruses.

The reverse genetics protocol used to generate recombinant rotaviruses was described in detail previously (46, 47). Briefly, BHK-T7 cell monolayers in 12-well plates were transfected with SA11 pT7 plasmids and pCMV-NP868R using the Mirus TransIT-LT1 transfection reagent. Transfection mixtures contained 0.8 μg of each of the 11 pT7 plasmids except for pT7/NSP2SA11 and pT7/NSP5SA11, which were used at levels 3-fold higher. Two days after transfection, the BHK-T7 cells were overseeded with MA104 cells (75) and the trypsin in the medium was adjusted to a final concentration of 0.5 μg/mL. Three days later, the BHK-T7/MA104 cell mixture was freeze-thawed 3 times and the lysates were clarified by low-speed centrifugation (800 × g, 5 min). To amplify recombinant viruses present in clarified lysates, the lysates were adjusted to 10 μg/mL trypsin (final concentration) and incubated for 1 h at 37°C (74). MA104 monolayers in 6-well plates were then infected, using 300 μL of the trypsin-treated lysate as inoculum, and incubated at 37°C in a CO₂ incubator until all cells were lysed (typically 5 to 7 days). Recombinant viruses were recovered from clarified cell lysates by plaque isolation on MA104 cell monolayers (74, 75). Plaque-isolated viruses were initially grown on MA104 cell monolayers in 6-well plates and then, to generate larger virus pools, grown on MA104 cell monolayers in T75 tissue culture flasks (~1 × 10⁷ cells per flask) at low multiplicity of infection (<1 PFU per cell). Briefly, 500 μL of plaque-amplified viral lysates was activated with trypsin (10 μg/mL, final concentration). The activated viral lysates were diluted to 2.5 mL with serum-free DMEM and then used as inoculum to infect MA104 cell monolayers in T75 flasks. The flasks were placed at 37°C in a CO₂ incubator for 1 h, with intermittent rocking (every 15 min) to ensure equal coverage of inoculum over monolayers. Following adsorption, the inoculum was removed and 12 mL of serum-free DMEM containing 0.5 μg/mL trypsin was added to each flask. Flasks were left in the incubator until the cell monolayers were fully disrupted (5 to 7 days). Infected cell lysates were collected from the flasks and clarified by low-speed centrifugation (800 × g, 5 min, 4°C). Clarified lysates were stored at −80°C. Virus titers in clarified lysates were determined by plaque assay on confluent MA104 cells in 6-well plates (74, 75). Peak titers represented the titer of virus (PFU per milliliter) reached in flasks when the MA104 cell monolayer was fully disrupted. Statistical differences between the peak titers reached by viruses were evaluated using an unpaired Student's t test with Welch's correlation. Titers are presented as means ± standard deviations with 95% confidence intervals (GraphPad Prism, version 8). Viral dsRNAs were recovered from infected-cell lysates by TRIzol extraction (Thermo-Fisher), resolved by electrophoresis on 10% polyacrylamide gels in Tris-glycine buffer, detected by staining with ethidium bromide, and visualized using a Bio-Rad ChemiDoc MP imaging system (47, 74).

Plaque assay.

Rotavirus plaque assays were performed as described before (74, 75). To visualize plaques, cell monolayers with agarose overlays were incubated overnight with phosphate-buffered saline (PBS) containing 3.7% formaldehyde. Afterward, agarose overlays were removed, and the monolayers were stained for 3 h with a solution of 1% crystal violet dissolved in 5% ethanol. Monolayers then were rinsed with water and air dried. Plaque images were captured using a Bio-Rad ChemiDoc imaging system and diameters were measured using ImageJ software, and the results were analyzed with GraphPad Prism, version 8. Statistical significance of plaque size differences was determined using an unpaired Student's t test and are presented as means ± standard deviations with 95% confidence intervals.

Immunoblot analysis.

MA104 cells were mock infected or infected with 5 PFU per cell of recombinant virus and harvested at 9 h p.i. Cells were washed with cold PBS, pelleted by centrifugation (5,000 × g, 10 min), and lysed by incubation for 30 min on ice in nondenaturing lysis buffer (300 mM NaCl, 100 mM Tris-HCl [pH 7.4], 2% Triton X-100, and 1× EDTA-free protease inhibitor cocktail [Roche Complete]). For immunoblot assays, lysates were resolved by electrophoresis on 10% polyacrylamide gels and transferred to nitrocellulose membranes. After blocking with PBS containing 5% nonfat dry milk, blots were probed with mouse monoclonal FLAG M2 (F1804, Sigma, 1:2,000), mouse monoclonal His tag antibody (MCA1396GA, Bio-Rad, 1:1,000), mouse 2A antibody (NBP2-59627, Novus, 1:1,000), guinea pig polyclonal NSP3 (lot 55068, 1:2,000), VP6 (lot 53963, 1:2,000) antisera, or rabbit monoclonal β-actin (8457S, Cell Signaling Technology [CST], 1:1,000) antibody. Bound primary antibodies were detected using 1:10,000 dilutions of horseradish peroxidase (HRP)-conjugated secondary antibodies (goat anti-mouse IgG [CST], goat anti-guinea pig IgG [KPL], or goat anti-rabbit IgG [CST]) or Alexa Fluor-conjugated antibody (goat anti-mouse Alexa 647 antibody [CST] in 2.5% nonfat dry milk [Carnation]). HRP signals were developed using Clarity Western ECL substrate (Bio-Rad) and detected using a Bio-Rad ChemiDoc imaging system, whereas Alexa Fluor signals were visualized directly using a Bio-Rad ChemiDoc imaging system.

To evaluate the dimerization capacity of NoV proteins expressed by rSA11 viruses, cell lysates were adjusted to a final concentration of 1.5% sodium dodecyl sulfate and 3% β-mercaptoethanol and incubated for 10 min at 25°C or 95°C. Afterward, proteins in the samples were resolved by electrophoresis on 4–20% gradient polyacrylamide gels and detected by immunoblot assay.

Immunoprecipitation assay.

Whole-cell lysates were prepared at 9 h p.i. from MA104 cells either mock infected or infected with rSA11 virus, as described above. Rabbit anti-NoV GII.4 monoclonal antibody (Absolute Antibody, NVB43.9, final dilution of 1:150) was added to cell lysates. After incubation at 4°C with gentle rocking for 18 h, antigen-antibody complexes were recovered using Pierce magnetic IgA/IgG beads (ThermoScientific), resolved by gel electrophoresis, and blotted onto nitrocellulose membranes. Blots were probed with antibodies specific for FLAG (1:2,000) and 6×His tags (1:1,000) to detect FLAG- and 6×His-tagged VP1 proteins, respectively.

Genetic stability of rSA11 viruses.

Viruses were serially passaged five times on MA104 cell monolayers using 1:1,000, 1:100, or 1:10 dilutions of infected cell lysates prepared in serum-free DMEM. When cytopathic effect reached completion (4 to 5 days), the cells were freeze-thawed three times in their own medium, and the lysates were clarified by low-speed centrifugation. dsRNAs were recovered from the clarified lysates by TRIzol extraction. The purified dsRNAs were resolved by electrophoresis on 10% polyacrylamide gels, and the bands of dsRNA were detected by ethidium bromide staining.

Isolation and sequencing of unstable variants.

Individual rSA11 variants were recovered from pools of the serially passaged virus by plaque isolation (74, 75). The variants were amplified by a single round of passage on MA104 cells and their genomic dsRNA was recovered by TRIzol extraction. The full-length genome segment 7 RNAs in samples were amplified with gene-specific primer pairs NSP3_5’UTR (5′-GGCATTTAATGCTTTTCAGTG-3′) and NSP3_3’UTR (5′-GGCCACATAACGCCCCTATAG-3′), and a shorter fragment from the C terminus of NSP3 ORF to the 3'-UTR was amplified with the primer pairs NSP3 C termF (5′-CATTGCACGCTTTTGATGACTTAG-3′) and NSP3_3'UTR (5'-GGCCACATAACGCCCCTATAG-3′) similarly using the Superscript III One-Step RT-PCR system with Platinum Taq DNA polymerase (Invitrogen). Amplified PCR products were resolved by electrophoresis on 0.8% agarose gels in Tris-acetate-EDTA buffer, products were gel purified using Nucleospin gel and PCR Clean-up (TaKaRa), and the sequences were determined by EuroFins Scientific.

Data availability.

Segment 7 sequences in rSA11 viruses have been deposited in GenBank and assigned the following accession numbers: wild type (LC178572), NSP3SA11-2A-fP2 (MN190002), NSP3SA11-2A-fP (MN190003), NSP3SA11-2A-fVP1 (MN190004), NSP3SA11-2A-VP1f (MN201548), NSP3SA11-2A-Ph (MN201549), NSP3SA11-2A-VP1th (MN201547), NSP3SA11-2A-VP1h (MZ562305), and NSP3RIX-2A-Ph (MZ643978) (see also Table 1).