key: cord-0302558-9cyba6jk
authors: Zhang, Fengwen; Zang, Trinity; Stevenson, Eva M.; Lei, Xiao; Copertino, Dennis C.; Mota, Talia M.; Boucau, Julie; Garcia-Beltran, Wilfredo F.; Jones, R. Brad; Bieniasz, Paul D.
title: Inhibition of major histocompatibility complex-I antigen presentation by sarbecovirus ORF7a proteins
date: 2022-05-26
journal: bioRxiv
DOI: 10.1101/2022.05.25.493467
sha: e5effc47cb235504395255970bd7e3298dad0674
doc_id: 302558
cord_uid: 9cyba6jk

Viruses employ a variety of strategies to escape or counteract immune responses, including depletion of cell surface major histocompatibility complex class I (MHC-I), that would ordinarily present viral peptides to CD8+ cytotoxic T cells. As part of a screen to elucidate biological activities associated with individual SARS-CoV-2 viral proteins, we found that ORF7a reduced cell surface MHC-I levels by approximately 5-fold. Nevertheless, in cells infected with SARS-CoV-2, surface MHC-I levels were reduced even in the absence of ORF7a, suggesting additional mechanisms of MHC-I downregulation. ORF7a proteins from a sample of sarbecoviruses varied in their ability to induce MHC-I downregulation and, unlike SARS-CoV-2, the ORF7a protein from SARS-CoV lacked MHC-I downregulating activity. A single-amino acid at position 59 (T/F) that is variable among sarbecovirus ORF7a proteins governed the difference in MHC-I downregulating activity. SARS-CoV-2 ORF7a physically associated with the MHC-I heavy chain and inhibited the presentation of expressed antigen to CD8+ T-cells. Speficially, ORF7a prevented the assembly of the MHC-I peptide loading complex and causing retention of MHC-I in the endoplasmic reticulum. The differential ability of ORF7a proteins to function in this way might affect sarbecovirus dissemination and persistence in human populations, particularly those with infection- or vaccine-elicited immunity.

To replicate and propagate in a host population that presents an immunologically hostile environment, viruses typically employ a variety of strategies to escape or counteract immune responses. Severe acute respiratory syndrome coronavirus-2 (SARS-CoV-2), a member of the sarbecovirus subgenus, has been shown to antagonize the innate immune response through the action of viral proteins (1-3) and to escape humoral immunity through variation in the neutralizing epitopes of the spike protein (4) (5) (6) (7) (8) . Evasion of cell-mediated immunity is accomplished by many viruses through the downregulation of surface expression of major histocompatibility complex-I, that would ordinarily present viral peptides to CD8+ cytotoxic Tcells (9) (10) (11) . For example, HIV-1 renders the virus-infected cells less visible to CD8+ T-cells through Nef-induced endocytosis of MHC-I from the cell surface (12) . In general, viruses from other families that are associated with chronic infections employ diverse mechanisms to deplete MHC-I from infected cell surfaces (10, 11) . However, viruses associated with short-term, acute infection do not typically induce MHC-I downregulation.

The ~30-kb SARS-CoV-2 genome encodes structural proteins (E, M, N, and S), nonstructural proteins (nsp1 to nsp16), and several 'accessory' open reading frames (ORF3a, ORF6, ORF7a, ORF8, ORF10, ORF3b, ORF9b, and ORF9c) (13, 14) . Analysis of coronavirus-host proteinprotein interactions, using affinity-or proximity-based approaches, has suggested that several viral proteins (ORF3a, ORF3b, ORF7a, ORF8, M, and nsp4) associate with host proteins that are enriched in endoplasm reticulum (ER) or Golgi, the organelles where viral peptides are loaded onto MHC-I molecule and transported to the cell surface for presentation to CD8+ T-cells (14, 15) . Moreover, Some reports have indicated that SARS-CoV-2 ORF8 reduces expression of MHC-I on the surface of infected cells (16, 17) . Here we show that SARS-CoV-2 ORF7a can inhibit antigen presentation by preventing the assembly of the MHC-I peptide loading complex and causing retention of MHC-I in the endoplasmic reticulum. Notably, ORF7a proteins from a sample of sarbecoviruses vary in their ability to induce MHC-I downregulation and a singleamino acid that is variable among sarbecovirus ORF7a proteins governs the differential ability to induce in MHC-I downregulation.

To elucidate biological activities associated with individual SARS-CoV-2 viral proteins we used an HIV-1-based lentiviral vector (pSCRPSY) (18) to express each SARS-CoV-2 viral open reading frame, as annotated in (13, 14) . Two days after transduction of human 293T cells, we measured MHC-I surface levels by flow cytometry using a pan-HLA class I-reactive monoclonal antibody. Expression of ORF7a reduced MHC-I levels on the cell surface by approximately 5fold, whereas expression of other individual viral proteins (notably including ORF8 (16, 17) ) had no effect on MHC-I surface levels ( Figure 1 ). We also examined the impact of SARS-CoV-2

ORFs on the expression of tetherin, a cell surface antiviral protein that traps enveloped virions from various virus families that bud through cell membranes. None of SARS-CoV-2 viral ORFs reduced the levels of tetherin stably expressed on the surface of 293T cells ( Figure S1 ), underscoring the specificity of the effect of ORF7a on MHC-I. Of note, expression of two viral proteins (nsp1 and ORF6) was not accomplished in our screen, as lentiviral vectors encoding these proteins were low titer, in line with the previous findings that nsp1 suppresses host protein synthesis (19) and that ORF6 interferes with nuclear transport machinery (14, 15) .

ORF7a caused reduced MHC-I cell surface levels in other human cells such as Huh7.5 and U2OS ( Figure 2A ) suggesting that its activity is not cell type-specific. Because the recognition of MHC-I by the W6/32 antibody could be influenced by association between heavy chains (HC) and β2-microglobulin (β2M) (20, 21) , we next confirmed that downregulation occurred, as assessed with a different antibody, specific for the HLA-A HC ( Figure 2B ). MHC-I surface levels were also depleted following SARS-CoV-2USA-WA1/2020 infection of A549/ACE2 cells, specifically After loading with peptides in the endoplasmic reticulum (ER), MHC-I molecules are transported via the Golgi to the plasma membrane and are subsequently turned over by endocytosis (22, 23) . To begin to elucidate how ORF7a might reduce cell surface MHC-I levels, we first depleted clathrin, with small interfering RNA and found that clathrin knockdown, as reported previously (24) , inhibited the downregulation of MHC-I by Nef ( Figure 2E ). Conversely, clathrin depletion had only a minor effect on MHC-I downregulation by ORF7a, suggesting that ORF7a disturbs surface MHC-I levels in a distinct, clathrin-independent manner ( Figure 2E ). Western blot analysis revealed no deficit in the total amount of MHC-I in cells expressing ORF7a. Indeed, when cells were transduced with the ORF7a lentivirus at sufficient MOI that most cells expressed ORF7a, the total amount of endogenous cellular MHC-I was increased, and displayed slightly accelerated migration, suggesting that ORF7a induces intracellular accumulation and altered post-translational modification of MHC-I ( Figure 2F ). Immunofluorescent staining of endogenous MHC-I in A549 cells showed that ORF7a profoundly altered the subcellular distribution of MHC-I molecules ( Figure 2G ). Specifically, ORF7a induced a reduction in of MHC-I HC fluorescence that could be observed at the cell surface, consistent with the flow cytometric analysis (Figure 2A ), while intracellular MHC-I HC accumulated ( Figure   2G ), The accumulated intracellular MHC-I HC was partly colocalized with a marker of the ER ( Figure 2G ). The MHC-I component β2M was similarly redistributed to intracellular locations upon ORF7a expression, and partly colocalized with ORF7a ( Figure S2 ). We conclude that SARS-CoV-2 ORF7a blocks the ability of MHC-I to move through the secretory pathway to the cell surface.

We tested whether ORF7a proteins from a panel of bat SARS-related coronaviruses (SARSr-CoV), namely BM48-31, HKU3-1, Rf1, Rs672, ZC45, ZXC21, and RaTG13 shared the ability to reduce cell surface MHC-I levels (25) . The RaTG13, Rf1, ZC45, and ZXC21 ORF7a proteins, but not those from BM48-31, HKU3-1, and Rs672 ORF7, reduced MHC-I surface levels ( Figure   3A ). Western blot and immunofluorescence analyses showed that the sarbecovirus ORF7a proteins were expressed in transduced cells at a similar level, and those that induced surface downregulation also induced intracellular accumulation and altered migration of MHC-I ( Figure   3B , Figure S3 ).

Notably, the ORF7a protein from SARS-CoV was among those that did not affect MHC-I surface levels ( Figure 3C ), despite encoding 85% identical amino acids to SARS-CoV-2 ORF7a.

Moreover, unlike SARS-CoV-2 ORF7a, and the active bat SARSr-CoV ORF7a proteins, the SARS-CoV ORF7a protein showed little tendency to colocalize with MHC-I ( Figure S4 ).

To map the determinants of the differential ability to modulate cell surface MHC-I levels, we constructed 6 chimeric SARS-CoV-2/SARS-CoV ORF7a expression vectors ( Figure 3D ).

Chimeric ORF7a proteins containing the central region (residues 46-90) from SARS-CoV-2 all downregulated MHC-I from the cell surface while those encoding the equivalent region from SARS-CoV did not ( Figure 3E ). Among the sarbecovirus ORF7a proteins, the presence of hydrophobic residues (F or I) at position 59 within this region correlated with the ability to downregulate MHC-I, while the presence of a polar residue (T or S) correlated with inactivity ( Figure 3F ). Thus, to test the importance of the amino acid at position 59, we introduced an F59T substitution into SARS-CoV-2 ORF7a and found this substitution abolished MHC-I downregulating activity, while the reciprocal substitution in SARS-CoV ORF7a (T59F) substitution led to acquisition of MHC-I downregulating activity ( Figure 3G ). We conclude that residue 59, a variable position among sarbecovirus ORF7a proteins, is a critical determinant of their ability to modulate cell surface MHC-I levels.

To test whether ORF7a proteins could physically associate with MHC-I, we immunoprecipitated the endogenously expressed MHC-I HC, in ORF7a expressing cells. The SARS-CoV-2 ORF7a protein could be co-immunoprecipitated from transduced cells with an anti-HLA-A antibody, and the amount of co-immunoprecipitated ORF7a but not the total level of ORF7a protein was reduced to background levels by the functionally inactivating F59T substitution in SARS-CoV-2 ORF7a ( Figure 4A ). Conversely, SARS-CoV ORF7 was poorly co-immunoprecipitated by the anti-HLA-A antibody but the T59F gain of function substitution increased the amount of coimmunoprecipitated ORF7a ( Figure 4A ). In reciprocal immunoprecipitation experiments, HLA-A could be immunoprecipitated by an ORF7a antibody from cells expressing wild-type SARS-CoV-2 ORF7a or SARS-CoV ORF7a(T59F) mutant ( Figure 4B ). Conversely, expression of the SARS-CoV-2 ORF7a(F59T) mutant or wild-type SARS-CoV ORF7a resulted in less efficient coprecipitation of the HLA-A protein ( Figure 4B ). We conclude that ORF7a associates with MHC-I HC and that single amino-acid substitutions in ORF7a that confer or ablate MHC-I downregulating activity simultaneously confer or ablate the ability of ORF7a to coimmunoprecipitate with MHC-I.

To determine whether SARS-CoV-2 ORF7a could inhibit antigen presentation, we engineered 293T cells to express HLA-A11 and then expressed an HIV-1 Gag protein and an ORF7a protein therein. We used these HIV-1 Gag-expressing cells as 'targets' for an HLA-A11 restricted CD8+ T-cell clone that recognizes the HIV-1 Gag peptide ACQGVGGPSHK (AK11) and measured the ability of the CD8+ T-cell clone to respond to the Gag-expressing target cells by secreting IFNγ. The AK11 CD8+ T-cell clone indeed responded to Gagexpressing 293T cells by secreting IFNγ, and this response was inhibited, almost to background levels, when the 293T target cells also expressed SARS-CoV-2 ORF7a ( Figure 4C ). The loss of function (F59T) SARS-CoV-2 ORF7a mutant and the wild-type SARS-CoV ORF7a proteins were less potent in reducing AK11 CD8+ T-cell clone responsiveness. Conversely, the SARS-CoV gain of function (T59F) mutant reduced the IFNγ response as potently as the WT SARS-CoV-2 ORF7a protein ( Figure 4C ).

We next attempted to determine more precisely how ORF7a interferes with MHC-I transport through the secretory pathway, and with antigen presentation at the cell surface. For antigen presentation, peptides are loaded into MHC-I in the ER through an interaction of MHC-I with the peptide loading complex (PLC). The PLC includes the transporter associated with antigen processing (TAP) proteins, TAP1 and TAP2, as well as chaperones/cofactors Tapasin, ERp57 and calreticulin that each bind directly or indirectly to the MHC-HC/ β2 microglobulin complex in the ER lumen (9) . Western blot analysis of proteins that coimmunoprecipitated with endogenous MHC-I revealed that expression of SARS-CoV-2 ORF7a blocked the co-immunoprecipitation of each the aforementioned MHC-I associated proteins with MHC-I HC without affecting the steady state levels of these proteins. To confirm the specificity of these effects, we employed the SARS-CoV-2 loss of function (F59T) and SARS-CoV gain of function (T59F) mutants. In similar co-immunoprecipitation assays in which endogenous HLA-B was immunoprecipitated, the co-immunoprecipitation of calreticulin and ERp57 was inhibited by the active WT SARS-CoV-2 and SARS-CoV(T59F) ORF7a proteins, but not by the inactive SARS-CoV-2(F59T) and WT SARS-CoV ORF7a proteins. We conclude that active ORF7a proteins disrupt the assembly of the MHC-I PLC in the ER and prevents the movement of peptide loaded MHC-I complexes to the cell surface for antigen presentation.

Inhibition of antigen presentation through the downregulation of MHC-I peptide complexes is a long-recognized means by which certain viruses mitigate the inhibitory action of immune responses on their replication (10) . MHC-I downregulation is frequently associated with viruses that cause persistent or chronic infection, or viruses with large DNA genomes and many immunomodulatory genes (11) . Such downregulation is less frequently associated with RNA viruses or those that cause acute, rapidly resolving infections (11) . Nevertheless, by reducing the density of viral peptides that are displayed on the surface of infected cells, viruses that cause acute infections such as SARS-CoV-2 could gain a competitive advantage through MHC-I downregulation. This advantage would be most pronounced in contexts where pre-existing T cell responses could limit virus replication, such as re-infection, and might also extend to limiting the protective effects of pre-existing T cell responses to epitopes that are conserved in distinct coronaviruses (27) . Moreover, while SARS-CoV-2 is typically viewed as an acute respiratory infection, it is increasingly recognized that SARS-CoV-2 infection can persist for months in certain anatomical sites (e.g. the gut) or in individuals with sub-optimal immune responses (28) (29) (30) . Moreover, the natural history of sarbecovirus infection in non-human species such as bats, and the degree to which it is a persistent infection is poorly documented. Thus, there are several reasons to think that MHC-I downregulation might be an evolved property in sarbecoviruses.

Some recent studies have suggested that SARS-CoV-2 ORF8 protein induces MHC-I downregulation (16, 17) . We did not observe such an effect. Nevertheless, an ORF7a-deleted SARS-CoV-2 induced MHC-I downregulation almost as efficiently as did an ORF7a-intact virus.

This finding suggests that additional redundant mechanisms are employed by SARS-CoV-2 to deplete surface MHC-I and reduce visibility to responding CD8+ T-cells. For example, a recent study has suggested that SARs-CoV-2 ORF6 inhibits transcriptional upregulation of MHC-I in response to infection (31) . To make the SARS-CoV (T59F) or SARS-CoV-2 ORF7a (F59T) expression vectors, overlapextension PCR amplification were performed with primers that incorporated the corresponding nucleotide substitutions using corresponding wild-type plasmids as template, respectively. After digestion with XhoI and NotI, the purified PCR products were then inserted into the pSCRPSY expression vector. HLA A11 gene fragment was PCR amplified and, after digestion with XhoI and NotI, subcloned into a retroviral vector (LMNi-bsd).

The human embryonic kidney HEK-293T cell line, human hepatoma-derived HuH-7. 

To assess HLA downregulation by SARS-CoV viral proteins, viral stocks were generated by To assess the effect of clathrin knockdown on MHC-I downregulation by SARS-CoV-2 ORF7a or HIV Nef, 293T cells were seeded one day prior to transfection with clathrin siRNA (Horizon Discovery, Dharmacon™) using Invitrogen Lipofectamine RNAiMax (ThermoFisher Scientific).

ORF7a or HIV Nef and harvested for FACS analysis 48 hours later. 

Cell lysates and immunoprecipitates were separated on NuPage Novex 4-12% Bis-Tris Mini Gels (Invitrogen), and NuPAGE™ MES SDS running buffer (Invitrogen, NP0002) was used.

Proteins were blotted onto nitrocellulose membranes. Thereafter, the blots were probed with primary antibodies and followed by secondary antibodies conjugated to IRDye 800CW or IRDye 680. Fluorescent signals were detected using an Odyssey scanner (LI-COR Biosciences). BioLegend), 37°C 5% CO2. One month later, wells were selected from the most dilute plate that showed growth (<1/3 of wells containing cell populations) and each was screened for specificity to Gag-AK11 peptide pool by CD107a staining and flow cytometry. Specific clones were expanded through biweekly stimulations with irradiated feeder medium. Clone specificity was further confirmed by peptide stimulation and CD107a staining on the day before performing T cell antigen recognition assays.

To measure antigen presentation, 293T cells were transduced with a retrovirus vector (LMNibsd) expressing HLA A11, or an empty vector control. The cells were subsequently transduced with empty SCRPSY or SCRPSY expressing an ORF7a protein at an MOI of 1. After selection in puromycin for 24 hours cells were transfected with an HIV-1 Gag expression plasmid. these 'target' cells were washed to remove the selection media, plated in 96 well plates and then cocultured with a Gag-AK11-specific CD8+ T-cell clone in RPMI 1640 medium supplemented with 10% fetal bovine serum (FBS), 2 mM l-glutamine, 100 units/ml penicillin, 100 μg/ml streptomycin, and 50 units/ml IL-2 (NCI BRB Preclinical Biologics Repository) at 37°C 5% CO2.

After 16 hours, cells were pelleted and supernatants were harvested for IFN-γ ELISA. Cells were then washed in 2% FBS phosphate-buffered saline + 2mM EDTA and surface stained with fluorochrome-conjugated antibodies to CD3-Brilliant Violet 785 clone OKT3, CD8-BV605 clone SK1, MHC-I-PacBlue clone W6/32, (all from BioLegend), CD4-APC R700 clone RPA-T4, (BD), Gag KC57-FITC (Beckman Coulter) and Fixable Aqua Viability Dye (Invitrogen). Counting beads were also added. Samples were analyzed by flow cytometry, and data analysis was performed with FlowJo, version 10, software.

Levels of IFN-γ in supernatants from T cell clone recognition assays were measured by ELISA using the ELISA MAX Deluxe Set Human IFN-g kit (Biolegend) following the manufacturer's instructions. Concentrations were determined by interpolating from a standard curve.

Five thousand A549 cells were seeded onto gelatin (Millipore ES-0060B) coated 8-chamber 

Following 10 min permeabilization at room temperature with 0

10% goat serum in PBS, cells were probed for 1 hr at RT with primary antibodies in 0.1% Tween20 (Fisher BP337-500), 10% goat serum in PBS as indicated: HLA-A (Sigma H1650-100STS, 0.4 µg/ml)

After three rinses with 0.1% Tween20 (Fisher BP337-500) in PBS, cells were probed with secondary antibodies for 45 min at RT in 0

10% goat serum, Hoechst 33258 (Abcam ab145596, 1 µg/mL

µg/ml or ThermoFisher Scientific A11029 4 µg/ml). Cells were rinsed twice with 0

Tween20 in PBS and twice with PBS and imaged by deconvolution microscopy (DeltaVision OMX SR imaging system)

Schematic representation of chimeras generated between SARS-CoV and SARS-CoV-2 ORF7a proteins

Flow cytometric analysis of 293T cells with a pan-HLA-ABC antibody (W6/32) following transduction with a pSCRPSY-lentiviral vector expressing chimeric ORF7a proteins shown in (D) Cells transduced (TagRFP+ population) with the ORF7a expression vector (solid line) and empty lentiviral vector

SARS-CoV-2, bat SARSr-CoV ORF7a proteins. Identical residues are shaded. Star indicates the amino acid at position 59 that were

Flow cytometric analysis of 293T cells with a pan-HLA-ABC antibody (W6/32) following transduction with a pSCRPSY-lentiviral vector expressing point mutant ORF7a proteins. Cells transduced (TagRFP+ population) with the ORF7a expression vector (solid line) and empty lentiviral vector

Activation and evasion of type I interferon responses by SARS-CoV-2

Evasion of Type I Interferon by SARS-CoV-2

SARS-CoV-2 nsp13, nsp14, nsp15 and orf6 function as potent interferon antagonists

Escape from neutralizing antibodies by SARS-CoV-2 spike protein variants

Antibody cocktail to SARS-CoV-2 spike protein prevents rapid mutational escape seen with individual antibodies

High genetic barrier to SARS-CoV-2 polyclonal neutralizing antibody escape

Complete Mapping of Mutations to the SARS-CoV-2 Spike Receptor-Binding Domain that Escape Antibody Recognition

Identification of SARS-CoV-2 spike mutations that attenuate monoclonal and serum antibody neutralization

Mechanisms of MHC class I--restricted antigen processing

MHC class I antigen presentation: learning from viral evasion strategies

Recent advances in viral evasion of the MHC Class I processing pathway

Endocytosis of major histocompatibility complex class I molecules is induced by the HIV-1 Nef protein

A new coronavirus associated with human respiratory disease in China

A SARS-CoV-2 protein interaction map reveals targets for drug repurposing

Global BioID-based SARS-CoV-2 proteins proximal interactome unveils novel ties between viral polypeptides and host factors involved in multiple COVID19-associated mechanisms

The ORF8 protein of SARS-CoV-2 mediates immune evasion through down-regulating MHC-Ι

SARS-CoV-2 variants do not evolve to promote further escape from MHC-I recognition

A diverse range of gene products are effectors of the type I interferon antiviral response

Structural basis for translational shutdown and immune evasion by the Nsp1 protein of SARS-CoV-2

Mapping of the monoclonal antibody W6/32: sensitivity to the amino terminus of beta2-microglobulin

The epitope recognized by pan-HLA class I-reactive monoclonal antibody W6/32 and its relationship to unusual stability of the HLA-B27/beta2-microglobulin complex

The quality control of MHC class I peptide loading

Endocytic Recycling of MHC Class I Molecules in Non-professional Antigen Presenting and Dendritic Cells

HIV-1 Nef-induced down-regulation of MHC class I requires AP-1 and clathrin but not PACS-1 and is impeded by AP-2

A pneumonia outbreak associated with a new coronavirus of probable bat origin

Endoglycosidase and glycoamidase release of N-linked glycans

Pre-existing polymerase-specific T cells expand in abortive seronegative SARS-CoV-2

Evolution of antibody immunity to SARS-CoV-2

Post-acute COVID-19 is characterized by gut viral antigen persistence in inflammatory bowel diseases

Persistent SARS-CoV-2 infection: the urgent need for access to treatment and trials

SARS-CoV-2 inhibits induction of the MHC class I pathway by targeting the STAT1-IRF1-NLRC5 axis

An Infectious cDNA Clone of SARS-CoV-2

Latent HIV reservoirs exhibit inherent resistance to elimination by CD8+ T cells

Acknowledgments This work was supported by a grant from the National Institutes of Allergy and Infectious Diseases P01 AI165075 (to PDB) UM1 AI164565, R01 AI147845 (to RBJ). and