key: cord-0920209-esipez86
authors: Kurt-Jones, Evelyn A.; Dudek, Tim; Watanabe, Daisuke; Mandell, Leisa; Che, Jenny; Zhou, Shenghua; Cao, LuCheng; Greenough, Thomas; Babcock, Gregory J.; Diaz, Fernando; Oh, Hyung Suk; Zhou, Changhong; Finberg, Robert W.; Knipe, David M.
title: Expression of SARS coronavirus 1 spike protein from a herpesviral vector induces innate immune signaling and neutralizing antibody responses
date: 2021-04-21
journal: Virology
DOI: 10.1016/j.virol.2021.04.006
sha: d3b12edfea92db333016ea24fb6ee38b1d312098
doc_id: 920209
cord_uid: esipez86

SARS coronavirus 1 (SARS-CoV1) causes a respiratory infection that can lead to acute respiratory distress characterized by inflammation and high levels of cytokines in the lung tissue. In this study we constructed a herpes simplex virus 1 replication-defective mutant vector expressing SARS-CoV1 spike protein as a potential vaccine vector and to probe the effects of spike protein on host cells. The spike protein expressed from this vector is functional in that it localizes to the surface of infected cells and induces fusion of ACE2-expressing cells. In immunized mice, the recombinant vector induced antibodies that bind to spike protein in an ELISA assay and that show neutralizing activity. The spike protein expressed from this vector can induce the expression of cytokines in an ACE2-independent, MyD88-dependent process. These results argue that the SARS-CoV1 spike protein intrinsically activates signaling pathways that induce cytokines and contribute directly to the inflammatory process of SARS.

Severe acute respiratory syndrome (SARS) coronavirus 1 (SARS-CoV1) causes acute inflammation and cytokine storm in the respiratory tract, leading to acute respiratory distress (Drosten et al., 2003; Ksiazek et al., 2003; Peiris et al., 2003) . The SARS-CoV genome encodes 4 structural proteins, S (spike), M (membrane), E (envelope protein), and N (nucleoprotein), and a series of non-structural and accessory proteins (Perlman and Masters, 2020) . The spike (S) is the major surface protein responsible for binding to the host cell receptor and promoting entry of the virus. S protein is the target of neutralizing antibodies against the virus. (Ksiazek et al., 2003; Peiris et al., 2003) , and this could contribute to viral pathogenesis, but the mechanisms of the full pathogenic process including the cytokine storm remain to be explained fully.

SARS-CoV1 is known to inhibit type 1 interferon signaling by blocking IRF-3 and NF-κB signaling (Devaraj et al., 2007; Frieman et al., 2009 ) by the action of several proteins (Perlman and Masters, 2020) , but the mechanism(s) of induction of pro-inflammatory cytokines have not been defined. Two studies have found weak induction of IL-6 and TNF-ɑ by soluble fragments of SARS spike protein or induction of IL-8 by baculovirus expressing SARS spike protein (Chang et al., 2004) , although high concentrations of soluble protein (e.g., 1-20 µg/ml) were needed for the effects observed.

Several viral recombinant vectors using modified vaccinia Ankara virus, parainfluenza virus, and adenovirus have been constructed that express the SARS J o u r n a l P r e -p r o o f spike protein as potential vaccines for SARS (Taylor, 2006) . We have constructed herpes simplex virus 1 (HSV-1) replication-defective mutant viruses (Murphy et al., 2000; Watanabe et al., 2007) that serve as vaccine vectors for simian immunodeficiency virus (SIV) in non-human primates (Murphy et al., 2000) , human immunodeficiency virus (HIV) in humanized mice (Claiborne et al., 2019) , and West Nile virus in mice (Taylor et al., 2016) . We specifically utilized the HSV-1 d106 recombinant virus that has deletions in the two copies of the ICP4 (R S 1) gene, and deletions in the ICP22/ICP47 promoter sequence (Samaniego, Neiderhiser, and DeLuca, 1998 ) as a vaccine vector (Liu et al., 2009; Watanabe et al., 2007) .

Viral recombinant vectors can also serve as vectors for expression of microbial proteins to define the effects of the expressed protein on host cells under reduced containment conditions. In this study we constructed an HSV-1 d106 replicationdefective mutant virus that expresses the SARS CoV1 spike protein, evaluated its immunogenicity in mice, and studied the effect of spike protein expression on cytokine expression in host cells.

Comparison of the vector potential of HSV-1 d106-GFP with Ad5-GFP.

To assess the vector potential of HSV-1 d106 virus, we compared expression of green fluorescent protein (GFP) from identical expression cassettes (CMV-eGFP) contained in the HSV-1 d106 virus ( Figure 1A ) or an adenovirus 5 recombinant virus, Ad5-GFP, under conditions where there were equal numbers of viral genomes of the two viruses in infected cells. To define conditions where equal numbers of GFP sequences were in the infected cells, we infected human foreskin fibroblast (HFF) cells with varying amounts of the two viruses using different multiplicities of infection (MOI) and measured GFP gene sequences in the infected cells by PCR. We determined that at MOI = 1.25 for d106 and MOI = 0.125 for Ad5-GFP that similar GFP DNA copy numbers were observed in infected cells (not shown). We therefore infected sets of cells with the two viruses under these conditions and measured GFP expression by immunoblotting ( Figure 2A ) or flow cytometry ( Figure 2B ). We observed higher expression of GFP protein in d106 infected cells as compared with Ad5-GFP virus ( Fig.   2A ). Furthermore, when we quantified fluorescent fluorescence using flow cytometry, we observed that GFP fluorescence was 10-fold higher with d106 as compared with Ad5-GFP (Fig. 2B ). We concluded that d106 is a very efficient vector for expression of a transgene.

Construction and testing of an HSV-1 vector expressing SARS coronavirus 1 spike protein.

We constructed an HSV-1 d106 recombinant vector, d106-SARS-CoV-1-S (Fig. J o u r n a l P r e -p r o o f 1B), expressing SARS CoV-1 spike protein by homologous recombination as described previously (Taylor et al., 2016) . SARS spike protein coding sequences from the SARS-SOptimized plasmid (Li et al., 2003) , generously provided by Michael Farzan, were inserted into the CMV expression vector pCIΔAflII (Murphy et al., 2000) , and the expression cassette was cloned into the transfer plasmid pPs27pd1 plasmid (Rice, Su, and Knipe, 1989) , which contains the HSV genomic sequences surrounding the U L 54 (ICP27) gene to generate the pd27SARS-S plasmid. The pd27SARS-S plasmid was linearized and co-transfected into E11 complementing cells with infectious HSV-1 d106 viral DNA. Progeny viruses were harvested, and potential recombinants were screened for by the formation of non-fluorescent plaques by visualization with an inverted fluorescence microscope. Recombinants were confirmed by the detection of S protein in cell lysates by Western blot analysis. The resulting d106-SARS-CoV-1S recombinant virus was triple-plaque purified, and stocks were grown in the complementing E11 cells.

To monitor expression of S protein, we infected human foreskin fibroblast (HFF) cells with d106-SARS-CoV-1S virus or mock-infected, prepared cell lysates and extracellular supernatant, and detected S protein by immunoblotting with the monoclonal antibody SW111 that we described previously (Petit et al., 2005) . In d106-SARS-CoV-1S virus-infected cells, we detected a major protein with apparent molecular weight of 180,000, similar to that in SARS virus-infected cells (Babcock et al., 2004) To test if the S protein expressed on the surface of d106-SARS-CoV-1S virusinfected cells was functional, we infected ACE2-293T cells that express the SARS virus receptor, ACE2 (Fig. 4A ). We observed that the infected cells formed syncytia ( Fig.   4B ), indicating that the S protein expressed by d106-SARS-CoV-1S virus was functional and at least part was localized on the surface of infected cells.

To define the immunogenicity of the recombinant vector, we immunized Balb/C mice with d106-SARS-CoV-1S or d106 virus subcutaneously in the rear flank at 0 and 3 weeks and obtained sera for testing at 3, 6, and 9 weeks. We performed an ELISA test using baculovirus-expressed S protein (Petit et al., 2005) , as described previously and found that mice immunized with d106-SARS-CoV-1 virus had spike protein binding antibodies (Fig. 5A ). Sera from C57BL/6 mice immunized with d106-SARS-CoV-1 developed SARS-CoV1 neutralizing activity (protection from CPE). With a single dose of d106-SARS-CoV-1, titers of 1:40-1:80 were achieved (data not shown). Following a second dose of d106-SARS-CoV-1, neutralizing titers increased to ≥1:640 for wild type C57BL/6 mice (Fig. 5B ). Neutralizing titers in TLR4 -/mice were indistinguishable from titers in WT mice. (Fig. 5B ). In contrast, the d106 vector alone induced minimal SARS-CoV1 neutralizing activity at all time points (titer <1:4). 

In these studies, we constructed an HSV-1 replication-defective recombinant vector strain expressing SARS spike protein, tested its immunogenicity, and used it to probe the effects of spike protein on host cells. The spike protein expressed from this vector is functional in that it localizes to the surface of infected cells and induces fusion of ACE2-expressing cells. In immunized mice, the recombinant vector induced antibodies that bind to spike protein in an ELISA assay and show neutralizing activity.

The spike protein expressed from this vector can induce the expression of cytokines in an ACE2-independent, MyD88-dependent process. These results argue that the SARS spike protein intrinsically activates signaling pathways that induce cytokines and could contribute directly to the inflammatory process of SARS.

Efficiency of HSV-1 d106 as a vector. We compared the HSV-1 d106 vector for expression of GFP protein with an Ad5 recombinant virus expressing the same GFP expression cassette under infection conditions where both vectors were delivering similar amounts of GFP DNA sequences into the host cells. We observed that HSV-1 d106 expressed more viral protein as detected on an immunoblot and 10-fold more fluorescence than Ad-GFP under these conditions. Thus, d106 virus is a very efficient vector for expression of a transgene in normal human fibroblasts. This is likely due, at least in part, to the expression of the HSV-1 ICP0 immediate-early protein, which promotes gene expression from the d106 genome (Samaniego, Neiderhiser, and DeLuca, 1998) by promoting the degradation of a number of host restriction factors (Everett et al.; Orzalli, DeLuca, and Knipe, 2012) and de-silencing the viral genome J o u r n a l P r e -p r o o f (Cliffe and Knipe, 2008; Lee, Raja, and Knipe, 2016) . In contrast, the Ad5 vector is deleted for E1A, which in part serves the same purpose of combatting the host cell epigenetic silencing of the viral genome. Further studies are needed to determine the mechanisms that define the relative efficiencies of gene expression by these vaccine vectors.

Expression of SARS spike protein. RANTES, and IL-6, relative to the d106 GFP vector. This argues that the spike protein can induce these cytokines, likely independently, but possibly in conjunction with an HSV virion protein. Induction of the cytokines requires MyD88, suggesting that Toll-like receptors or other innate pathways are involved in the cytokine induction. We observed a partial dependence on TLR2 or TLR4, two innate receptors that signal via MyD88, in macrophages challenged with d106-SARS-CoV1S (not shown), suggesting that multiple pathways may be contributing to the induction of inflammation. In contrast to inflammatory cytokine production, we noted that TLR4 was not required for Ab responses to S protein. These results argue that S protein is part of the mechanism of induction of inflammatory cytokines by SARS virus. This is ACE2-independent because it is observed in murine cells.

A previous study using transfection of lung epithelial and fibroblast cells expressing SARS spike protein showed activation of the IL-8 gene promoter (Chang et al., 2004) . Soluble S protein, transfection of plasmid encoding S protein, or infection with a baculovirus vector expressing S protein induced IL-8 expression (Chang et al., 2004) . Their results argued for AP-1 induction and not NF-κB and a dependence on ACE2 for the induction. Another study using soluble S protein incubation with murine RAW264.7 macrophages showed induction of IL-6 and TNF . This induction involved IKB degradation and NF-κB signaling. Therefore, different systems have yielded different mechanisms and different cytokine activation, but in total, the results argue that SARS spike protein may contribute to or even be a major contributor to the cytokine storm observed in SARS.

Induction of cytokines could be due, at least in part, to over-expression or even normal levels of spike protein by the d106 vector and activation of the unfolded protein response (UPR) as observed previously (Chan et al., 2006) . The UPR has been associated with cytokine induction (reviewed by Smith, 2018) . SARS CoV-1 viral infection has been shown to activate UPR (Chan et al., 2006) and this has been speculated to increase viral assembly. Further studies should define the relationships between these observations. The implications of these results are two-fold. Approaches that block the signaling induced by the SARS spike protein could reduce the inflammation and immune pathology. Furthermore, the results point to the possibility that immunization with a vector expressing SARS protein or with SARS protein itself might lead to local inflammatory cytokine production.

HSV-1 d106 virus (Samaniego, Neiderhiser, and DeLuca, 1998) was kindly provided by Neal DeLuca, University of Pittsburgh. Infectious full-length HSV-1 d106 viral DNA was purified from infected E11 cell lysates by sodium iodide gradient centrifugation (Walboomers and Schegget, 1976) , as described (Knipe, Ruyechan, and Roizman, 1979 ).

An adenovirus 5 recombinant expressing GFP (Ad5CMV GFP) was obtained from the Viral Vector Core at the University of Iowa (VVC-U of Iowa-4 Ad5CMVeGFP).

Antibodies.

Anti-ACE2 antibody (R&D, Cat#: AF933, 1:1000) and Anti-GAPDH ([6C5], Mouse monoclonal, abcam, Cat. #: ab8245, 1:5000) were used in this study.

Cell-cell fusion assay.

Human embryonic kidney 293T/ACE2 cells (kind gift from Michael Farzan) (5x10 5 ) were plated in 6-well plates. On the next day, the cells were infected with d106-SARS-CoV-1S virus (3x10 6 PFU/well), and cell-cell fusion was assessed at 20 h post infection (hpi). Images were acquired using a Nikon TE200 microscope.

Immunoblotting was performed as described previously (Oh et al., 2014) . Briefly, cells were lysed in 1x RIPA buffer (10 mM Tris-Cl (pH 8.0), 1 mM EDTA, 0.5 mM EGTA, J o u r n a l P r e -p r o o f 1% Triton X-100, 0.1% sodium deoxycholate, 0.1% SDS, 140 mM NaCl) with protease inhibitor (Complete Protease Inhibitor Cocktail, Millipore Sigma). The proteins were resolved in NuPAGE 4-12% Bis-Tris Gels (Life Technologies) and transferred to a Nitrocellulose Membrane (Bio-Rad, #1620112). The membranes were then incubated in Odyssey Blocking Buffer (LI-COR) for 1 h at room temperature and incubated with antibodies specific for ACE2 (1:1000, polyclonal goat, R&D, Cat#: AF933) or GAPDH

(1:10,000, mAb, Abcam, Cat#: ab8245). The membranes were incubated with secondary antibodies, IRDye 680RD or IRDye 800 (LI-COR), for 45 min at room temperature. Near-infrared fluorescence was detected using Odyssey (LI-COR) and imaged using ImageStudio V4 (LI-COR).

Immunofluorescence.

HFF cells (4-8x10 4 ) were seeded on glass coverslips in a 24-well plate. On the next day, the cells were infected with d106-SARS-CoV-1S virus at an MOI of 3 for 20 h.

To detect spike protein expression on the cell surface, unfixed or fixed, unpermeabilized cells were used. Fixation of cells was with 3.7% paraformaldehyde in PBS for 10 min, and permeabilization was performed using 0.1% Triton X-100 in PBS for 10 min, both as indicated. Permeabilized cells were blocked in 10% normal goat serum (Jackson ImmunoResearch) + human IgG (0.16%) in DMEM for 1 h at room temperature. Cells were incubated with monoclonal antibody SW111 specific for spike protein (Petit et al., 2005) for 1 h at room temperature. Secondary antibodies conjugated to the dye Alexa For immunization studies, 8wk old C57BL/6J WT or TLR4-/-mice were infected ip with 10 5 pfu d106 or d106-SARS (primary immunization). Four weeks later, a cohort of mice were boosted with a second ip infection of 10 5 pfu d106 viruses (secondary immunization). Sera were collected at 4 wks (post-primary immunization) and 7 wks (i.e., 3 wks post-secondary boosting). Sera were tested for SARS spike protein binding by ELISA and for SARS neutralization activity in CPE reduction assays.

Neutralization assays (CPE reduction).

The neutralizing activity of serum was measured as described previously (Greenough et al., 2005) . Vero E6 cells were seeded at 5000 cells/well, in 96-well J o u r n a l P r e -p r o o f microtiter plates, on assay day -1 in a volume of 100 µL. On assay day 0, two-fold serum dilutions were preincubated for 1 h with 100 TCID50 of virus stock (Urbani strain; generously provided by Larry Anderson [Centers for Disease Control and Prevention, Atlanta]; low passage in Vero E6 cells). These mixtures of virus and serum dilutions then were added to cells in duplicate. One additional set of serum dilutions without virus was included as a control to detect toxicity. Virus stock was back-titrated in each assay, to ensure that the inoculum was 30-300 TCID50/well. Presence or absence of cytopathic effect (CPE) after 72 to 96 h of incubation was determined by microscopy.

The dilution of serum that completely prevented CPE in 50% of the wells was calculated by means of the Reed-Muench formula. After microscopic visualization of CPE, the medium was replaced by PBS, CellTiter96 reagent (Promega) was added, and plates were incubated for 2-4 h until gradations of color between uninfected and infected controls were easily distinguished visually. CellTiter96 is metabolized to a soluble, colored product, the concentration of which is proportional to the number of viable cells in the culture. Absorbance is reduced in wells with significant CPE. To inactivate virus, 10% SDS was added, and the absorbance (optical density measured at 490 nm) was read by use of a universal plate reader (EL 800; Bio-Tek Instruments). Percent protective effect was calculated as follows: 100 (observed CPE-maximum CPE)/(minimum CPE-maximum CPE), where "maximum CPE" refers to absorbance in control wells with virus and no serum and "minimum CPE" refers to absorbance in control wells with no virus and no serum. 

Amino acids 270 to 510 of the severe acute respiratory syndrome coronavirus spike protein are required for interaction with receptor

Modulation of the unfolded protein response by the severe acute respiratory syndrome coronavirus spike protein

Induction of IL-8 release in lung cells via activator protein-1 by recombinant baculovirus displaying severe acute respiratory syndrome-coronavirus spike proteins: identification of two functional regions

Herpes simplex virus ICP0 promotes both histone removal and acetylation on viral DNA during lytic infection

Regulation of IRF-3-dependent innate immunity by the papain-like protease domain of the severe acute respiratory syndrome coronavirus

A Randomized, Double-Blinded, Placebo-Controlled, Phase 1 Study of a Replication-Defective Herpes Simplex Virus (HSV) Type 2 Vaccine, HSV529, in Adults With or Without HSV Infection

Identification of a novel coronavirus in patients with severe acute respiratory syndrome

Use of murine embryonic fibroblasts to define Toll-like receptor activation and specificity

Herpesviral ICP0 Protein Promotes Two Waves of Heterochromatin Removal on an Early Viral Promoter during Lytic Infection

The epitope study on the SARS-CoV nucleocapsid protein

Genetic engineering of a modified herpes simplex virus 1 vaccine vector

Vaccine protection against simian immunodeficiency virus by recombinant strains of herpes simplex virus

A targeted RNA interference screen reveals novel epigenetic factors that regulate herpesviral gene expression

Nuclear IFI16 induction of IRF-3 signaling during herpesviral infection and degradation of IFI16 by the viral ICP0 protein

Clinical progression and viral load in a community outbreak of coronavirus-associated SARS pneumonia: a prospective study

Coronaviridae. In "Fields Virology

Genetic analysis of the SARS-coronavirus spike glycoprotein functional domains involved in cell-surface expression and cell-tocell fusion

Herpes simplex virus alpha protein ICP27 possesses separable positive and negative regulatory activities

Persistence and expression of the herpes simplex virus genome in the absence of immediateearly proteins

Regulation of Cytokine Production by the Unfolded Protein Response; Implications for Infection and Autoimmunity

Obstacles and advances in SARS vaccine development

Production of immunogenic West Nile virus-like particles using a herpes simplex virus 1 recombinant vector

A new method for the isolation of herpes simplex virus type 2 DNA

Up-regulation of IL-6 and TNF-alpha induced by SARScoronavirus spike protein in murine macrophages via NF-kappaB pathway

Properties of a herpes simplex virus multiple immediate-early gene-deleted recombinant as a vaccine vector

187-90. the boost. Neutralizing titers of anti-SARS-CoV1 Ab measured by in CPE protection assay and are expressed as the inverse titer conferring 50% protection from SARS-CoV1 infection induced CPE

WT d106-SARS-CoV1S, ≥640; TLR4-/-d106-SARS-CoV1S, ≥640

Figure 6. Induction of innate cytokines in PBMC by infection with d106-SARS-CoV-1S virus

PBMCs were plated at 10 6 per well and infected with d106-SARS-CoV-1S vector or d106 vector alone at MOI 1 and 10. Controls included challenge with medium alone, LPS (TLR4 ligand, 10ng/ml), Pam 2 CSK 4 (TLR2 ligand, 100ng/ml), Pam 3 CSK 4 (TLR2 ligand, 100ng/ml), or IL-1β (10ng/ml)

6 per well) were infected with d106-SARS-CoV-1S vector or d106 vector alone at varying MOI. (A) MCP-1 and (B) RANTES cytokine levels in culture supernatants were measured 18 h later by ELISA. (C, D) Controls include challenge with medium alone

This research was supported by grants from the New England Regional Center for Excellence in Biodefense and Emerging Infectious Diseases (DMK and RWF) and by a grant from the Massachusetts Consortium for Pathogen Readiness (DMK).