Chikungunya virus infection disrupts MHC-I antigen presentation via nonstructural protein 2

Infection by chikungunya virus (CHIKV), a mosquito-borne alphavirus, causes severe polyarthralgia and polymyalgia, which can last in some people for months to years. Chronic CHIKV disease signs and symptoms are associated with the persistence of viral nucleic acid and antigen in tissues. Like humans and nonhuman primates, CHIKV infection in mice results in the development of robust adaptive antiviral immune responses. Despite this, joint tissue fibroblasts survive CHIKV infection and can support persistent viral replication, suggesting that they escape immune surveillance. Here, using a recombinant CHIKV strain encoding the fluorescent protein VENUS with an embedded CD8+ T cell epitope, SIINFEKL, we observed a marked loss of both MHC class I (MHC-I) surface expression and antigen presentation by CHIKV-infected joint tissue fibroblasts. Both in vivo and ex vivo infected joint tissue fibroblasts displayed reduced cell surface levels of H2-Kb and H2-Db MHC-I proteins while maintaining similar levels of other cell surface proteins. Mutations within the methyl transferase-like domain of the CHIKV nonstructural protein 2 (nsP2) increased MHC-I cell surface expression and antigen presentation efficiency by CHIKV-infected cells. Moreover, expression of WT nsP2 alone, but not nsP2 with mutations in the methyltransferase-like domain, resulted in decreased MHC-I antigen presentation efficiency. MHC-I surface expression and antigen presentation was rescued by replacing VENUS-SIINFEKL with SIINFEKL tethered to β2-microglobulin in the CHIKV genome, which bypasses the requirement for peptide processing and TAP-mediated peptide transport into the endoplasmic reticulum. Collectively, this work suggests that CHIKV escapes the surveillance of antiviral CD8+ T cells, in part, by nsP2-mediated disruption of MHC-I antigen presentation.


Introduction
Since its re-emergence in 2004 in the Indian Ocean region [1,2], chikungunya virus (CHIKV) has spread to numerous subtropical and tropical regions of Asia [3], the Caribbean [4], the Americas [2] and Europe [5] causing large-scale mosquito-borne epidemics that challenge public health systems.In the Americas, over three million CHIKV cases in 45 countries have been documented, and the global spread of CHIKV places approximately 1.3 billion people at risk for infection [6].CHIKV infection causes an acute debilitating febrile illness accompanied by high fever, severe myalgia, and arthralgia.Musculoskeletal tissue pain and inflammation can become chronic in up to two thirds of patients [7,8,9].Due to the increasing incidence of CHIKV disease, the unpredictable nature of its outbreaks, and the lack of approved vaccines or antiviral medications, it remains critical to elucidate CHIKV pathogenesis and immune restriction and evasion mechanisms.
CHIKV is one of ~32 Togaviridae family members [10] and is comprised of an 11.8 kilobase, positive-sense, single-stranded RNA genome that contains two open reading frames encoding nonstructural and structural polyproteins [11].After an alphavirus particle binds a host cell, the virion is internalized by clathrin-mediated endocytosis.Upon fusion of the viral and endosomal membranes, the virion disassembles, and viral RNA (vRNA) is released into the cytoplasm.The vRNA is translated by host ribosomes to generate the nonstructural polyprotein P123 or P1234 via read-through of the opal termination codon in nsP3 [11].The protease activity of nonstructural protein 2 (nsP2) cleaves P1234 into P123 and nsP4, thus allowing for formation of a replication complex that synthesizes full-length negative-strand RNA from the template viral genome.Subsequent processing of P123 into nsP1, nsP2, and nsP3 results in formation of a replicase that synthesizes positive sense genomic vRNA as well as a subgenomic mRNA that is translated into the viral structural polyprotein.Though nsP1-4 are synthesized at equal molar ratios in their polyprotein form, only a single nsP2 for every twelve nsP1s is found within the replicase complex [12,13], suggesting nsP2 has auxiliary functions apart from the replicase.Indeed, nsP2 has been reported to enter nuclei and export STAT1 [14], suppress host mRNA transcription through the induction of RPB1 ubiquitination and degradation [15], induce translational shutoff by phosphorylation of eEF2 [16], disrupt type I interferon (IFN) signaling [17,18], and suppress the unfolded protein response (UPR) [19].
CD8 + T cells are key executors of the adaptive immune response against many viral infections, including some arbovirus infections [20,21,22,23,24,25].Their capacity to surveil, recognize, and lyse virus-infected cells is well documented [26,27,28,29].Early innate immune responses, including inflammatory cytokines [30,31], and the presence of viral peptides displayed by major histocompatibility complex class I (MHC-I) molecules on cell surfaces dictate the breadth and duration of the CD8 + T cell response.Importantly, CD8 + T cells scan for the presence of viral peptide-loaded MHC-I (pMHC-I) complexes on the surface of infected cells via their T cell receptor (TCR) [32,33].Ligation of the TCR leads to activation and release of cytotoxic effectors such as perforin 1 and granzyme B and the subsequent destruction of the infected target cell [34].Numerous viruses have evolved mechanisms to subvert this crucial immune surveillance system through downregulation, suppression, or elimination of pMHC-I complexes from the surface of infected cells, thereby evading CD8 + T cell-mediated clearance [35].Although this paradigm has been extensively illustrated for DNA viruses such as herpesviruses and poxviruses [36], some RNA viruses also have evolved mechanisms for interfering with CD8 + T cell recognition.For example, the human immunodeficiency virus 1 Nef protein diverts pMHC-I complexes to the protein degradation pathway [37,38], and infection with influenza A and B viruses also suppresses surface MHC-I expression [39].SARS-CoV-2 infection was shown to disrupt MHC-I cell surface expression [40,41], in part, by disrupting nuclear transport of NLRC5 [41], the principal MHC-I transcription factor.
Recently, we observed that during CHIKV infection, CD8 + T cells accumulate in joint-associated tissue and lyse exogenously peptide-loaded splenocytes that had been adoptively transferred into the tissue.Despite this finding, CHIKV-specific CD8 + T cells appear to have little to no impact on the clearance of CHIKV-infected cells, as the viral burden in joint tissues of WT and Cd8α -/-mice are similar throughout the course of infection, suggesting that CHIKVinfected cells evade the CD8 + T cell response [42].
Here, using a recombinant chimeric protein composed of the reporter VENUS fused to a well characterized CD8 + T cell epitope (ovalbumin 257-264; SIINFEKL), embedded in the CHIKV genome, we observed a marked loss of both MHC-I cell surface expression and antigen presentation by CHIKV-infected joint tissue fibroblasts.Mutations within the methyl transferase-like (MTL) domain of the CHIKV nsP2 restored MHC-I cell surface expression and antigen presentation efficiency by CHIKV-infected cells.In contrast to cell surface MHC-I expression, intracellular MHC-I levels were unaffected by CHIKV infection.Moreover, MHC-I antigen presentation by CHIKV-infected cells was rescued by replacing the VENUS-SIINFEKL in the CHIKV genome with SIINFEKL tethered to β2-microglobulin (β 2 m), which obviates the need for peptide processing and TAP-mediated peptide transport into the endoplasmic reticulum (ER).Finally, we find that expression of wild-type (WT) nsP2 alone, but not nsP2 with mutations in the MTL domain, is sufficient to disrupt MHC-I antigen presentation efficiency in primary joint tissue fibroblasts.In contrast, expression of nsP2 with mutations known to abrogate helicase or protease activity disrupts MHC-I antigen presentation similar to WT nsP2, suggesting a specific role for the MTL domain in disrupting MHC-I antigen presentation.Collectively, this work suggests that CHIKV-infected cells escape surveillance by antiviral CD8 + T cells, in part, by nsP2-mediated disruption of MHC-I antigen presentation.

CHIKV-infected primary joint tissue fibroblasts display decreased MHC-I cell surface expression
To determine if virus-encoded peptides are presented by MHC-I on CHIKV-infected cells, a recombinant CHIKV strain (CHIKV-VENKL) was engineered to express the chimeric protein VENUS-SIINFEKL (VENKL) wherein VENUS, a fluorescent reporter protein [43], and the CD8 + T cell peptide SIINFEKL from ovalbumin [44] were encoded between the capsid autoproteolysis domain [45] and an engineered 2a ribosomal skip site in-frame in the structural open reading frame of the CHIKV genome (Fig 1A), a design we have used previously [46].Following the C-terminus of VENUS, SIINFEKL is flanked by the amino acid sequences LEQLE and TEW, which are important for efficient peptide processing and MHC-I loading of SIINFEKL [47].This design allows for the measurement of MHC-I antigen presentation efficiency because we can measure SIINFEKL presentation as a function of source protein (i.e., VENUS) across all cells on a per-cell basis [47].To assess MHC-I antigen presentation by physiologically relevant CHIKV target cells, we generated primary ex vivo ankle (EVA) cell cultures by adapting a previously established protocol in which single cell suspensions were generated from murine ankle tissue [42], depleted of erythrocytes and nonadherent cells, and then cultured in growth medium.
After 24 hours (h) in culture, EVA cells were assessed for cellular composition, the capacity to support CHIKV infection, and MHC-I cell surface expression by flow cytometry.More than 95% of EVA cells were CD45 -(Figs 1B and S1A) and within that population most cells (99%) expressed the fibroblast marker CD29 (Figs 1C and S1A), a marker of ankle tissue fibroblasts shown previously to support CHIKV infection in vivo [48].In addition, 75-80% of the CD45 -cells had detectable H2-K b and H2-D b MHC-I molecules on the cell surface (Fig 1D-1E).To evaluate their capacity to support CHIKV infection, EVA cells were mock-inoculated or inoculated with 10 6 plaque forming units (PFU), a multiplicity of infection (MOI) of ~0.67 FFU/cell, of CHIKV-VENKL and at 24 h post-inoculation (hpi) cells were evaluated for VENUS expression as an indicator of productive CHIKV infection.Consistent with prior studies [48], most CHIKV-infected cells were CD45 -and expressed CD29 on their surface ( Figs 1B, 1C, and S1A).Next, we assessed surface levels of MHC-I on CHIKV-infected VENUS + cells and control cells.VENUS + CD45 -EVA cells showed marked reduction of H2-D b and H2-K b cell surface expression compared with VENUS -CD45 -EVA cells exposed to CHIKV, but not mock-infected CD45 -EVA cells, as measured by geometric mean fluorescence intensity (gMFI) (Fig 1D -1E).The percentage of VENUS + CD45 -EVA cells that displayed cell surface H2-K b and H2-D b was reduced compared with mock-infected CD45 -EVA cells (Fig 1D -1E).In contrast, the percentage of VENUS + CD45 -EVA cells that expressed cell surface CD29 was unchanged compared with control cells (

Determinants in nsP2 promote impairment of MHC-I antigen presentation by CHIKV-infected cells
CHIKV nsP2 interferes with numerous cellular functions, including host cell transcription by promoting degradation of RPB1 [15], type 1 IFN responses by nuclear export of STAT1 [17] and inhibition of Jak/STAT phosphorylation [18,49], and the unfolded protein response (UPR) [19], all of which can be important for MHC-I antigen presentation [50,51,52,53].Given this, we hypothesized that CHIKV nsP2 impairs MHC-I antigen presentation in CHIKV-infected cells.In prior studies, Akhrymuk et al. demonstrated that mutations in the MTL domain of CHIKV nsP2 at ATL 674-676 QMS (QMS) abrogate nsP2-mediated inhibition of cellular transcription and type I IFN responses, as well as the cytopathic effects induced by CHIKV infection, without compromising CHIKV replication in vitro [54].We therefore incorporated the QMS mutations into CHIKV-VENKL (CHIKV QMS -VENKL) (Fig 2A) and confirmed that murine fibroblasts supported comparable viral replication of CHIKV-VENKL and CHIKV QMS -VENKL, particularly in the absence of type I IFN signaling (Fig 2B).EVA cells were inoculated with 10 6 PFU of CHIKV-VENKL or CHIKV QMS -VENKL and at 24 hpi cells were evaluated by flow cytometry for VENUS and cell surface H2-K b (Figs 2C and S2).In addition, we quantified cell surface SIINFEKL-loaded H2-K b (H2-K b -SIINFEKL) using the 25D1.15T cell receptor mimic monoclonal antibody [55].CHIKV-VENKL and CHIKV QMS -VENKL both infected CD45 -EVA cells, albeit with modestly different efficiencies (Figs 2D  and S2).The surface level of CD29 was similar on cells infected with either CHIKV-VENKL or CHIKV QMS -VENKL (Figs 2E and S2).In contrast, the surface level of H2-K b on cells infected with CHIKV-VENKL was reduced compared with cells infected with CHIKV QMS -VENKL (Fig 2F).Moreover, the percentage of VENUS + cells that were double positive for both H2-K b -SIINFEKL (i.e., viral peptide-loaded MHC-I) and H2-K b was significantly lower among cells infected with CHIKV-VENKL compared with CHIKV QMS -VENKL, and these cells showed markedly lower pMHC-I presentation efficiency (Fig 2G -2H).These data indicate that MHC-I presentation of viral peptides is less efficient in CHIKV-infected cells and that this can be improved by mutations in nsP2.
To further assess whether CHIKV infection interferes with functional MHC-I antigen presentation, and validate our direct assessment of cell surface H2-K b -SIINFEKL detected by the 25D1.16antibody, we developed a co-culture system utilizing highly specific and sensitive transgenic OT-I CD8 + T cells, which upregulate cell surface expression of CD69 and CD25 upon recognition of and activation by H2-K b -SIINFEKL complexes [56].To determine if OT-I CD8 + T cells are activated by CHIKV-infected cells, EVA cells were inoculated with mock inoculum or 10 6 PFU of CHIKV-VENKL or CHIKV QMS -VENKL.At 24 hpi, all groups of cells were washed and co-cultured for 6 h with a 1:1 mixture of 10 6 purified OT-I CD8 + T cells and 10 6 bystander CD8 + T cells with >95% purity and minimal background activation prior to addition (S3A-S3C indicating that their activation was not specific to TCR ligation.These data suggest that CHIKV-infected cells are inefficiently recognized by antigen-specific CD8 + T cells and that this evasion is promoted by determinants in nsP2.In addition, these findings demonstrate that the lower signal detected by the H2-K b -SIINFEKL antibody on CHIKV-VENKLinfected cells functionally impacts antigen presentation to CD8 + T cells.

MHC-I antigen presentation is inefficient in CHIKV-infected cells in vivo
To evaluate whether CHIKV-infected cells display poor MHC-I antigen presentation in vivo, WT C57BL/6 mice were inoculated in the left-rear footpad with 10 3 PFU of either CHIKV-VENKL or CHIKV QMS

Cell surface, but not intracellular, MHC-I expression is reduced in CHIKVinfected cells
CHIKV nsP2 impairs cellular transcription in vitro [57,58] which could explain the poor pMHC-I presentation efficiency by CHIKV-infected cells and the increase in pMHC-I presentation by cells infected with CHIKV encoding the QMS mutations in nsP2.To investigate whether the suppression of cell surface MHC-I expression in CHIKV-infected cells is due to a reduction in total cell-associated MHC-I, we developed a flow cytometry-based intracellular staining protocol for H2-K b to allow for concurrent assessment of cell surface and cell-associated MHC-I on single EVA cells (Fig 5A).The intracellular quantity of H2-K b and percentage of cells expressing intracellular H2-K b was similar for CHIKV-VENKL-and CHIKV QMS -VENKL-infected cells.In contrast, fewer CHIKV-VENKL-infected cells displayed H2-K b on their surface, and those that did had reduced expression levels (Fig 5B and 5C).These data suggest that CHIKV infection does not downregulate expression of H2-K b by infected cells, but instead disrupts a step of the MHC-I antigen presentation pathway required for the display of mature pMHC-I on the cell surface.

Bypassing peptide processing and ER transport restores MHC-I antigen presentation by CHIKV-infected cells
Given our findings that CHIKV-infected cells inefficiently present pMHC-I, and that intracellular expression of H2-K b is normal during CHIKV-VENKL infection, we sought to determine the step(s) at which MHC-I maturation is disrupted in CHIKV-infected cells.Mature MHC-I complexes are composed of a membrane anchored heavy chain (e.g., H2-K b ), in association with beta-2-microgobulin (β 2 m), bound to an 8-10 amino acid peptide.β 2 m and peptide of proper length are required for the heavy chain of MHC-I to traffic to the cell surface [59], and the absence of an appropriate peptide destabilizes MHC-I during transit and at the cell surface [60,61].Prior studies found that the covalent linkage of an antigenic peptide to β 2 m via a flexible linker allows for assembly of mature, peptide-loaded MHC-I complexes independent of peptide processing, transporter associated with antigen processing (TAP)-mediated peptide translocation into the ER lumen, and the peptide loading complex (PLC) [62,63,64].To evaluate if CHIKV-infected cells are deficient for these early steps in MHC-I antigen presentation, the VENKL coding sequence in CHIKV-VENKL was replaced with SIINFEKL-GSSGSSGS-  infected cells displayed H2-K b -SIINFEKL on the cell surface, resulting in improved pMHC-I presentation efficiency that was elevated compared with CHIKV QMS -VENKL-infected cells (Fig 6B and 6C).Likewise, the inoculation of mice with CHIKV-SIINFEKLβ 2 m elicited an increased percentage of antigen presenting CHIKV-infected cells and increased pMHC-I presentation efficiency compared to CHIKV-VENKL at 2 dpi, even above that seen by cells infected with CHIKV QMS -VENKL (Fig 6D and 6E).These findings suggest CHIKV infection disables MHC-I antigen presentation at a step prior to β 2 m-peptide stabilization of the heavy chain within the ER.
To further demonstrate that presentation of viral antigen by MHC-I can occur more efficiently in CHIKV-infected cells when the peptide is linked to β 2 m, EVA cells were inoculated with mock inoculum or 10 6 PFU of CHIKV-VENKL or CHIKV-SIINFEKLβ 2 m.At 24 hpi, all groups of cells were washed and co-cultured for 6 h with a 1:1 mixture of 10 6 purified OT-I CD8 + T cells and 10 6 bystander CD8 + T cells.Like OT-I CD8 + T cells co-cultured with CHIKV QMS -VENKL-infected EVA cells (Fig 3 CD25 compared with bystander levels (Fig 7A and 7B) and OT-I CD8 + T cells co-cultured with CHIKV-VENKL-infected EVA cells.Furthermore, these OT-I CD8 + T cells expressed elevated levels of IRF4, a transcription factor activated in response to TCR ligation strength [65], suggesting that specific engagement of CHIKV-infected cells with OT-I CD8 + T cells is restored when the virus-encoded peptide is directly linked with β 2 m.Notably, OT-I CD8 + T cells cocultured with CHIKV-SIINFEKLβ 2 m-, but not CHIKV-VENKL-infected EVA cells, showed a marked loss of CD8a expression (Fig 7B), which has been observed on activated CD8 + T cells [66], further suggesting that antigen-specific CD8 + T cells poorly engage cells infected with CHIKV-VENKL.

CHIKV nsP2 impairs antigen presentation efficiency
Our data shows that mutations in the MTL domain of nsP2 increased MHC-I viral antigen presentation by CHIKV-infected cells.Based on these data, we hypothesized that nsP2 expression alone would be sufficient to disrupt pMHC-I presentation efficiency.To test this idea, EVA cells were co-transfected with a CHIKV nsP2 expression vector along with an expression vector encoding VENKL.Control cells were co-transfected with a control vector plus VENKL.At 24 h post-transfection, ~20-30% of EVA cells were VENUS + (S6A Fig) .By Western blot analysis, we detected expression of both nsP2 and nsP2 QMS , with nsP2 QMS expression increased compared with nsP2 (Fig 8A).In cells transfected with nsP2, VENUS + CD45 -EVA cells displayed a reduced percentage of cells double-positive for H2-K b and H2-K b -SIINFEKL, and decreased pMHC-I presentation efficiency (Fig 8B and 8C) compared with control cells.In contrast, EVA cells transfected with nsP2 QMS and VENKL had a similar frequency of double-positive (H2-K b+ H2-K b -SIINFEKL + ) cells and pMHC-I presentation efficiency as control cells (Fig 8B and 8C).These findings indicate that nsP2 expression alone can impair pMHC-I presentation efficiency similar to CHIKV-infected cells.
Given that alphavirus nsP2 has additional functional domains (e.g., helicase and protease domains) in addition to the MTL domain, and that expression of nsP2 can inhibit host cell transcription and translation, we tested the capacity of other forms of nsP2 to impair pMHC-I presentation efficiency.As shown in Fig 8D -8F, we found that expression of nsP2 with a mutation in the protease active site (i.e., C478S) or the highly conserved lysine in the Walker A motif (i.e., K192A) results in lower levels of cell surface H-2K b -SIINFEKL compared with control cells or cells expressing nsP2 QMS and reduced pMHC-I presentation efficiency to the same extent as WT nsP2.Importantly, all forms of nsP2 were expressed to similar levels (Fig 8D ) and cells from all groups showed similar levels of VENUS expression (S6B Fig) .Notably, the K192A mutation has been previously reported to abrogate RBP1 degradation and eEF2 phosphorylation by nsP2 which are required for inhibition of cellular transcription [15] and translation [16], respectively.These findings suggest that nsP2 interferes with MHC-I antigen presentation by a mechanism that is independent of effects on host cell transcription or translation, and nsP2 protease activity.

Discussion
Replication of arthritogenic alphaviruses in musculoskeletal tissues of mice, nonhuman primates, and humans leads to the recruitment and infiltration of inflammatory cells [67,68,69,70,71,72].Some of these cellular infiltrates, such as Ly6C hi monocytes, contribute to control of alphavirus infection [73,74], whereas others, such as CD4 + T cells, mediate joint tissue pathology [75,76].These tissue infiltrates also include CD8 + T cells [42], which are critical for the control and clearance of many viral infections.Remarkably, prior studies found that CD8 + T cells have little to no impact on CHIKV or RRV replication in joint-associated tissues [42,77].However, CD8 + T cells in joint-associated tissues of CHIKV-infected mice were capable of lysing adoptively transferred splenocytes exogenously loaded with peptide [42], suggesting that CHIKV-infected cells in joint tissue escape CD8 + T cell surveillance.Thus, we investigated MHC-I antigen presentation by CHIKV-infected cells.Indeed, we found that joint-associated CHIKV-infected fibroblasts displayed reduced cell surface MHC-I expression and inefficiently presented a virus-encoded peptide in MHC-I both in vitro and in vivo.Moreover, CHIKV-infected fibroblasts were unable to activate viral antigen-specific CD8 + T cells, further indicating that MHC-I antigen presentation by CHIKV-infected cells is impaired.
Viruses have evolved numerous strategies to evade the MHC-I antigen presentation pathway, including disruption of MHC-I machinery by suppression of transcription [35,36,78,79,80], such as the suppression of NLRC5 function by the ORF6 protein of SARS-CoV-2 [41], highlighting the importance of CD8 + T cells in the elimination of virus-infected cells and control of virus infection.Infection of cells with CHIKV and other alphaviruses can inhibit host cell transcription [81].Mechanistically, for CHIKV and several other alphaviruses, inhibition of transcription is mediated by translocation of the viral nsP2 protein into the nucleus where it promotes polyubiquitination and proteasomal degradation of RPB1 [15], the catalytic subunit of cellular DNA-dependent RNA polymerase II.Recently, CHIKV strains that lack the capacity to induce degradation of RPB1 but retain high levels of replication were developed-these strains encode specific mutations in the V-loop of the MTL domain of nsP2 [54].We introduced these mutations into CHIKV-VENKL and found that cells infected with this virus retained cell surface expression of MHC-I, more efficiently presented MHC-I antigen, and better activated antigen-specific CD8 + T cells.Moreover, we found that transfection of primary murine ankle tissue fibroblasts with WT CHIKV nsP2, but not CHIKV nsP2 with mutations in the V-loop, was sufficient to diminish MHC-I surface expression and pMHC-I presentation efficiency.Thus, our studies identified a new role for nsP2 in immune evasion.
Our findings suggest that CHIKV nsP2 interferes with MHC-I antigen presentation by a mechanism that prevents display of mature peptide-loaded MHC-I on the cell surface (Fig 9).Prior studies by other groups found that MHC-I peptides that are covalently tethered to β 2 m are translocated into the ER by the β 2 m signal sequence, thus, bypassing the requirement for TAP-mediated transport [62].In addition, this approach bypasses the need for peptide processing [64].We engineered CHIKV to express SIINFEKL tethered to the N-terminus of β 2 m via a Gly/Ser linker [62] (CHIKV-SIINFEKLβ 2 m) and found that cells infected with this virus efficiently display SIINFEKL-loaded MHC-I on the cell surface.Furthermore, the capacity for viral expression of β 2 m-SIINFEKL to restore peptide-loaded MHC-I on the cell surface and promote activation of antigen-specific CD8 + T cells indicates that MHC-I trafficking is not altered by CHIKV-infected cells and further supports the idea that steps prior to MHC-I heavy chain stabilization and trafficking are disrupted.Mechanistically, these findings indicate that CHIKV infection and nsP2 expression disrupt MHC-I antigen presentation by altering peptide generation, peptide transport into the ER, or the peptide loading complex (PLC), and future studies are needed to resolve the precise mechanism of action.It is possible that nsP2 directly interacts with components of the antigen processing pathway to disrupt MHC-I antigen presentation.For example, ubiquinilin 4 (UBQLN4) directly interacts with nsP2 [82] and is known to target mis-localized membrane proteins to the proteasome [83].Given the critical importance of the proteasome for the generation of CD8 + T cell epitopes [84], it is possible that the interaction between nsP2 and UBQLN4 leads to a reduction of pMHC-I antigenic epitopes for surveillance of CHIKV-infected cells by CD8 + T cells.Alternatively, in contrast to the MHC-I heavy chain, other components of the MHC-I antigen presentation pathway may be more sensitive to nsP2-mediated transcriptional or translational inhibition.For example, HCMV infection has been shown to disrupt transcription of tapasin more so than other components of the PLC [85].However, as described below, our results suggest that nsP2 disrupts MHC-I antigen presentation by a mechanism distinct from the inhibition of host cell transcription or translation.
Given that CHIKV infection and nsP2 expression can inhibit host cell transcription, and mutations in the nsP2 MTL domain that abrogate transcriptional inhibition [54] enhanced MHC-I antigen presentation by CHIKV-infected and nsP2-expressing cells, we developed a flow cytometry-based assay for quantifying cell-associated and cell surface MHC-I levels on individual cells to determine if CHIKV infection specifically alters cell surface levels of MHC-I or induces downregulation of total MHC-I expression.These studies revealed that while the cell surface level of MHC-I was decreased on CHIKV-infected cells, intracellular levels of MHC-I were unaffected, indicating that CHIKV infection does not result in a general loss of MHC-I expression due to inhibition of host cell transcription.That intracellular levels of MHC-I are unaffected by CHIKV infection is further supported by our studies utilizing CHIKV-SIINFEKLβ 2 m which demonstrated that nascent MHC-I heavy chains could be stabilized and shuttled to the cell surface by expression of the chimeric SIINFEKLβ 2 m protein.To further investigate a role for transcription or translation inhibition in nsP2-mediated impairment of MHC-I antigen presentation, we evaluated nsP2 with a mutation in the helicase domain (nsP2 K192A ) that abrogates nsP2-mediated RBP1 degradation and eEF2 phosphorylation required for inhibition of cellular transcription [15] and translation [16], respectively, for the capacity to impair pMHC-I antigen presentation efficiency.Remarkably, in contrast with expression of nsP2 QMS , expression of either nsP2 or nsP2 K192A similarly impaired pMHC-I antigen presentation, suggesting a role for the MTL domain in antagonism of MHC-I antigen presentation that is independent of effects on cellular transcription and/or translation.In addition to these findings, we also found that expression of nsP2 harboring a mutation in the protease site (nsP2 C478A ) disrupted pMHC-I antigen presentation efficiency, suggesting that protease activity is not required for nsP2-mediated disruption of MHC-I antigen presentation.In summary, we find that through an nsP2-mediated mechanism, CHIKV-infected joint tissue fibroblasts exhibit inefficient MHC-I presentation of virus-encoded peptides.These data provide one explanation for the limited role of CD8 + T cells in controlling CHIKV infection.

Ethics statement
This study was conducted in accordance with the recommendations in the Guide for the Care and Use of Laboratory Animals and the American Veterinary Medical Association (AVMA) Guidelines for the Euthanasia of Animals.All animal experiments conducted at the University of Colorado Anschutz Medical Campus were performed with the approval of the Institutional Animal Care and Use Committee (IACUC) of the University of Colorado School of Medicine (Assurance Number: A3269-01) under protocols 00026 and 00215.Experimental animals were humanely euthanized at defined endpoints by exposure to isoflurane vapors followed by bilateral thoracotomy.

Mice and mouse experiments
An animal biosafety level 3 laboratory (ABSL3) with biosafety cabinet was used for all mouse experiments.C57BL/6 mice (stock #000664) were obtained from The Jackson Laboratory.CD45.1 + OT-I (C57BL/6-Tg(TcraTcrb)1100Mjb/J) mice were bred and housed at the University of Colorado School of Medicine under specific pathogen-free conditions.CD45.1 + OT-1 mice were maintained as hemizygous animals and selected based on flow cytometry phenotyping of transgene positive CD8 + cells expressing both Vα2 and Vβ5.1/2 TCRs.All mice were used in experiments at 4 weeks of age.Mice were anesthetized with isoflurane vapors and subcutaneously inoculated with 10 3 PFU of CHIKV in the left rear footpad in 10 μL of PBS/1% FBS using a Hamilton syringe and 30G needle.Mice were euthanized by exposure to isoflurane vapors and bilateral thoracotomy, and then intracardially perfused with phosphate buffered saline (PBS) at the indicated time points.

Cells
BHK-21 cells (American Type Culture Collection, ATCC CCL-10) were cultured at 37˚C in Minimum Essential Medium Alpha (MEMα) supplemented with 10% fetal bovine serum (FBS), 10% tryptose phosphate broth (TPB) and 100 U/mL of penicillin-streptomycin (P/S).Mouse embryonic fibroblasts (derived from WT and Ifnar1 -/-C57BL/6 mice) were a gift from Michael Diamond (Washington University at St. Louis) and cultured in Dulbecco's Modified Eagle medium (DMEM; HyClone 11965-084) supplemented with 10% FBS and 100 U/ml P/S (D:10).All cell lines were passaged when the monolayer was <90% confluent and fewer than 10 times.To generate primary ex vivo ankle (EVA) cells, 4-week-old C57BL/6 mice were euthanized, and ankle tissues dissected.A single-cell suspension was generated by horizontal shaking of ankle tissue dissections at 37˚C for 2 h with 5 mm glass beads in digestion media containing D:10, 2.5 mg/mL collagenase type I (Worthington Biochemical), and 17 μg/mL DNase I (Roche).The cell suspension was filtered through a 100 μm sterile filter, washed with 1x PBS and centrifuged at 350 x g for 5 min at room temperature.The supernatant was decanted, and the cell pellet was resuspended in D:10.Approximately 1.5 x 10 6 ankle tissue cells were added to a single well of a 6-well plate in 4 mL of D:10 medium and cultured at 37˚C for 24 h.EVA cells were transfected using Lipofectamine 3000 (L3000) according to the manufacturer's instructions.Briefly, EVA cells were plated 24 h prior to direct addition of 50 μL containing: 0.2 μg of either pCMV, pCMV-VENKL, pCMV-nsP2, pCMV-nsP2 K192A , pCMV-nsP2 C478A or pCMV-nsP2 QMS combined with 1.5 μL of L3000 and 2 μL of P3000 reagent in Opti-MEM.Twenty-four h later cells were harvested by EDTA dissociation, stained with designated antibodies, and assessed by flow cytometry.For virus infection studies, medium was aspirated, and cells were inoculated with 1 x 10 6 PFU of the indicated virus in 200 μL of in vitro diluent (PBS with 1% FBS, 1 mM Ca 2+ , and 1 mM Mg 2+ ).After one hour with gentle rocking, the inoculum was removed, cells were washed with 1x PBS and then incubated for 16 h at 37˚C in 4 mL D:10.

Viruses
CHIKV cDNA clones of AF15561 were generated as described previously [68,86].The recombinant CHIKV-VENKL cDNA clone was constructed by insertion of a tandem sequence encoding the fluorescent GFP derivative, VENUS, linked to LEQLE-SIINFEKL-TEW followed by the 2A protease sequence of foot-and-mouth disease virus (FMDV2a) and inserted inframe between capsid and E3 of CHIKV [46].CHIKV-SIINFEKLβ 2 m was generated similarly by inserting SIINFEKL with a GSSGSSGS linker between the β 2 m signal sequence and transcriptional start site of human β 2 m.CHIKV QMS -VENKL was generated by site-directed mutagenesis of the sequence encoding amino acids 674 ATL 677 of the nsP2 using primers 5' CACTA GGTCATACCTACCACTCATCTGTGGTAGACCCAgCTCTAGG-3' and 5'-CCTAGAGC TGGGTCTACCACAGATGAGTGGTAGGTATGACCTAGTG-3'.The PCR amplicon was inserted into the WT vector.The complete viral genome sequence in all plasmids was confirmed by Sanger sequencing.Stocks of infectious CHIKV were generated from cDNA clones and titered by plaque assay on BHK-21 cells [68] or focus formation assay using Vero cells as previously described [87].

In vitro virus replication assay
Triplicate wells of MEFs were inoculated with virus at a MOI of 0.01 FFU/cell.Viruses were absorbed onto cells for 1 h at 37˚C with gentle rocking.Cells were washed two times with 1 mL of 1x PBS, then 1 mL of D:10 was added to each well, and cells were incubated at 37˚C.At the indicated time points, 100 μL samples of culture supernatants were collected, and an equal volume of D:10 was added to maintain a constant volume within each well.Samples were stored at −80˚C before analysis by focus formation assay on Vero cells.

Focus formation assays
As previously described [87], Vero cells were seeded in 96-well plates and inoculated with 10-fold serial dilutions of samples for 2 h at 37˚C.Cells were overlayed with 0.5% methylcellulose in MEM-alpha plus 10% FBS and incubated for 18 h at 37˚C.Following fixation with 1% paraformaldehyde (PFA), CHIKV-infected cells were detected using CHK-11 monoclonal antibody [88] at 500 ng/mL, and foci were visualized with TrueBlue Peroxidase substrate (Ser-aCare 5510-0030) and counted using a CTL Biospot analyzer and Biospot software (Cellular Technology).

CD8 + T cell activation assays
Splenocytes from CD45.1 + OT-I mice were acquired as previously described [42].CD8 + T cells were enriched by magnetic bead separation with negative selection following the StemCell protocol (StemCell, 19853) and only utilized with >90% purity.After enrichment, CD45.1 + OT-I CD8 + T cells were added to primary murine ankle cell cultures that had been previously infected for 24 h.After 6 h of co-culture, non-adherent cells were collected, stained with designated antibodies, and analyzed by flow cytometry.

Quantification and statistical analysis
Antigen presentation efficiency was calculated utilizing the Flowjo "derive parameter" function taking the mean fluorescence intensity of H2-K b -SIINFEKL channel and dividing it by VENUS or E2 mean fluorescence intensity [47].For statistical analysis, one-way analysis of variance (ANOVA) with multiple comparisons, and student's t-test were performed on data using GraphPad Prism 10.02.P values of <0.05 were considered significant.

Fig 1 .
Fig 1. CHIKV-infected primary joint tissue fibroblasts display decreased MHC-I cell surface expression.(A) Schematic of the recombinant CHIKV-VENKL.The coding sequence for the VENUS-SIINFEKL chimeric protein was inserted in-frame into the structural ORF of the CHIKV genome with known cleavage motifs flanking SIINFEKL.(B-E) EVA cells were inoculated with PBS (mock) or 10 6 PFU CHIKV-VENKL.(B) At 24 hpi, the percentages of CD45 -(open circles) and CD45 + (open squares) cells among mock-inoculated cells and VENUS + cells were determined.(C) The magnitude and frequency of expression of CD29 on CD45 -cells was determined by flow cytometry.(D) Representative flow cytometry plots of H2-K b and VENUS expressing cells from mock (black) and CHIKV-VENKL-infected cell populations (VENUS + , red; VENUS -, grey).(E) Quantification of surface expression and frequency of H2-K b and H2-D b from mock and CHIKV-VENKL-infected EVA cells.Data are pooled (B), or representative (C, E) from 3-4 independent experiments.P values were determined by unpaired student's t-test (B) or one-way ANOVA with Tukey's test for multiple comparisons (C,E).*, P<0.05; **, P<0.01; ***, P<0.001; ****, P<0.0001.https://doi.org/10.1371/journal.ppat.1011794.g001 Fig 1C).To assess the extent to which VENUS expression correlated with cell surface levels of MHC-I, cells were stratified into gates based on the intensity of VENUS expression (S1B Fig) and then assessed for H2-K b and H2-D b expression.As VENUS expression increased, H2-K b and H2-D b expression decreased (S1C Fig), suggesting cells with more CHIKV infection express less cell surface MHC-I.Collectively, these data suggest that CHIKV infection suppresses cell surface MHC-I expression but not expression of other cell surface proteins.

Fig 9 .
Fig 9. Proposed model for CHIKV disruption of MHC-I antigen presentation.CHIKV infection disables MHC-I antigen presentation at a step prior to β 2 m-peptide stabilization of the heavy chain within the ER. Figure made with Biorender.com.https://doi.org/10.1371/journal.ppat.1011794.g009