We have previously shown that MCMV protein m20.1 (Lenac Rovis et al., 2016), just like its counterpart, the HCMV protein UL141 (Tomasec et al., 2005), retains CD155 in the endoplasmic reticulum (ER) in an immature form, leading to its proteasomal degradation. However, we have also observed that CD155 accumulates outside the ER compartment in MCMV-infected cells (Figure 1A, Figure 1B, upper panel), irrespective of the ER-resident m20.1 protein (Figure 1B, lower panel, Figure 1—figure supplement 1). Thus, we aimed to determine if there is an additional MCMV regulator of CD155.

By screening a library of MCMV deletion mutants, we have identified a new viral regulator of CD155 surface expression. Figure 1C shows the CD155 surface expression upon infection with MCMV mutants lacking genes of the region m144-m158. Based on the ability of the mutant virus to restore CD155 surface expression, we identified m154 gene of MCMV as the one responsible for the downregulation of CD155 (Figure 1C). We also tested if this downregulation has a functional impact in terms of binding of CD155 to its immune receptors. Infection of cells with WT MCMV led to a decrease in TIGIT-Fc and DNAM-1-Fc binding while the viral mutant lacking m154 gene (Δm154) restored the binding of CD155 interaction partners on the surface of infected cells (Figure 1D). The restoration of binding was more pronounced for TIGIT-Fc, since the binding of TIGIT to CD155 is of higher affinity and faster kinetics as compared to DNAM-1 (Stanietsky et al., 2013).

To determine whether m154 is involved in CD155 intracellular accumulation, we first analyzed its kinetics of expression and localization. In line with the published results (Zarama et al., 2014), we observed early kinetics of expression of the m154 protein on the plasma membrane of infected cells. However, we also observed that a significant amount of m154 is localized intracellularly (Figure 2A). The intracellular expression of m154 overlapped with CD155 intracellular accumulation (Figure 2B). After testing a panel of markers for cellular compartments, we determined that in MCMV-infected cells, the accumulated CD155 and m154 colocalize in the adaptor protein-1 (AP-1) positive compartment (Figure 2C). The AP-1 protein complex is associated with cargo selection and sorting in trans-Golgi network (TGN), endosomes, and clathrin-coated vesicles (Ghosh and Kornfeld, 2003; Nakatsu et al., 2014). We also generated stable transfectants expressing m154 N-terminal HA tagged protein and showed that m154 N-HA occupies the expected location in the AP-1 compartment (Figure 2D).

Since HCMV also downregulates CD155 surface expression (Tomasec et al., 2005), we assessed whether human CD155 colocalizes with AP-1 compartment upon infection. By using human foreskin fibroblasts (HFF) infected with HCMV TB40/E strain, we observed that indeed, human CD155 accumulates intracellularly in the AP-1 compartment (Figure 2E). Thus, the AP-1 accumulation of CD155 is a feature of both MCMV and HCMV infection.

We showed that in WT MCMV infection m154 localizes to the plasma membrane and AP-1 positive compartment where it also colocalizes with CD155. We examined if the progression of newly synthesized CD155 to the plasma membrane was arrested by m154 on the level of AP-1 compartment. In the cells infected with the virus lacking m154, CD155 was still accumulating in the AP-1 compartment (Figure 2—figure supplement 1), indicating that m154 does not directly participate in CD155 retention in this compartment. Next, we investigated the possibility that the accumulated CD155 molecules in AP-1 compartment of MCMV-infected cells were mature, endocytosed CD155 molecules originating from the plasma membrane. To assess this, we stained the plasma membrane CD155 on uninfected and MCMV-infected cells and followed its internalization after 20 hr (Figure 3A, schematic model). To avoid the binding of the antibody to the viral Fc receptor, we used the MCMV mutant that lacks m138 gene encoding for the viral Fc receptor (Crnković-Mertens et al., 1998). We observed that CD155 surface staining is quite stable on the plasma membrane of uninfected cells even after 20 hr, while the MCMV-infected cells showed a more pronounced reduction of the initial CD155 surface staining (Figure 3A). Using the same surface staining model, but with a permeabilisation step (Figure 3B, schematic model), we followed the intracellular trafficking of membrane CD155 by confocal microscopy and found that a significant portion of plasma membrane-stained CD155 (PM-CD155) 20 hr post-infection (p.i.) resided in the AP-1 compartment in MCMV-infected cells, confirming the membrane origin of the intracellularly accumulated CD155 (Figure 3B, upper panel). Consistently, the internalized CD155 portion also colocalized with the m154 intracellular expression (Figure 3B, lower panel). By using the same surface staining model (Figure 3C, schematic model) on m154-N-HA transfectants we showed that surface labeled m154 is also internalized into the AP-1 compartment (Figure 3C).

Since the AP-1 clathrin-adaptor is involved in lysosomal enzymes sorting via its TGN-endosome transport function, we next analyzed whether the PM-CD155 molecules of MCMV-infected cells are further processed in lysosomes. To that aim, we stained the PM-CD155, treated cells with lysosomal inhibitor leupeptin and followed the trafficking of internalized CD155 after 20 hr (Figure 4A, schematic model). PM-CD155 staining was quite stable on the surface of uninfected cells, regardless of the leupeptin treatment (Figure 4A). The MCMV-infected cells showed an increased amount of internalized PM-CD155 after the leupeptin treatment (Figure 4A). The rescued portion of PM-CD155 also colocalized with late endosomes and lysosomes, as shown by staining with the acidic organelle probe DAMP. We questioned whether the redirection of PM-CD155 into lysosomes instead of recycling back to the plasma membrane is a direct consequence of m154 action. ∆m154 virus possesses viral Fc receptor that would unspecifically bind anti-CD155 antibody in our 20 hr surface staining model. Thus, we labeled plasma membrane proteins with biotin (Figure 4B, schematic model) and followed the degradation of the surface portion of CD155 both in WT and Δm154 infection by western blot. The biotinylation assay confirmed the loss of CD155 molecules from the plasma membrane of MCMV-infected cells (Figure 4B) and the partial rescue of PM-CD155 under leupeptin treatment. More importantly, the degradation of surface CD155 molecules was much less pronounced in the case of Δm154 infection, and there was no further increase of PM-CD155 with leupeptin treatment. We also demonstrated the involvement of lysosomal enzymes in the degradation of m154 protein in MCMV-infected cells (Figure 4C). In addition, PM-m154 accumulated in m154-N-HA transfectants when lysosomal degradation was blocked (Figure 4D). Thus, we concluded that m154 interferes with CD155 trafficking by redirecting the PM-CD155 from the AP-1 vesicles to the lysosomal compartment.

Given the fact that numerous proteins are subjects of AP-1 cargo selection and sorting, we suspected that additional cellular proteins might be affected by the above-described m154 mechanism of action. Therefore, we compared the expression of various surface molecules in cells infected with WT and Δm154 virus and observed an effect of m154 on the surface expression of a remarkable set of molecules involved in innate and adaptive immune responses (Figure 5A). This comprised signaling lymphocytic activation family molecules (CD229, CD84), cell adhesion molecules and their receptors (CD18, CD54, CD162), members of tumor necrosis factor receptor family (CD262, CD270), and an integrin-associated protein CD47. The selectivity of m154 downregulation was further shown by examining the surface expression of another set of immunologically relevant molecules that were affected by WT virus but not rescued in the case when m154 gene was missing (Figure 5B). We also showed that m20.1, which regulates immature CD155 molecules on the level of ER, has no role in regulating a broad spectrum of targets affected by m154 (Figure 5—figure supplement 1).

Next, we wanted to determine whether some of the newly discovered m154 targets follow the same pattern of MCMV-induced sequestration from the cell membrane into the intracellular AP-1 compartment and m154-mediated rerouting for degradation in lysosomes, as observed for CD155. To this end, we used the same model as described in Figure 3B, which allowed us to follow the trafficking of surface-stained molecules by confocal microscopy 20 h p.i.. In uninfected cells, the plasma membrane-stained CD229 (PM-CD229) remained quite stable on the cell membrane after 20 hr (Figure 5C), while in MCMV-infected cells, the PM-CD229 molecules were internalized and accumulated in the AP-1 compartment (Figure 5C). When the lysosomal degradation was inhibited in the infected cells by leupeptin, additional intracellular accumulation of PM-CD229 was detected in infected cells, on the top of its colocalization with the AP-1 compartment (Figure 5D). The additional amount of PM-CD229 colocalized with lysosomes in infected cells treated with leupeptin, while the colocalization was minimal when lysosomes were functioning normally (Figure 5D). In addition, MCMV induced the accumulation of plasma membrane fraction of other m154 targets, CD18 and CD47, in the AP-1 compartment (Figure 5—figure supplement 2).

To conclude, m154 affects the surface expression of multiple targets involved in the antiviral immune response by employing the same mechanism of redirection of the plasma membrane fraction of a protein from the AP-1 compartment to the lysosomal degradation. The AP-1-mediated TGN export of m154 target proteins is blocked by m154, leading to the accumulation of a large fraction of m154 targets in the TGN and a reduction of m154 target proteins on the plasma membrane. A fraction of m154 targets is transported to lysosomes and degraded (Figure 5—figure supplement 3).

To confirm that AP-1-dependent sorting is involved in the downregulation of m154 targets we silenced AP-1µ subunit by siRNAs. We showed that AP1MI siRNA targeting reduced the ability of MCMV to downregulate CD155 by 50% (Figure 6A and Figure 6—figure supplement 1). Accordingly, the ability of MCMV to downregulate CD155 was also reduced in cells which do not constitutively express functional AP-1 complex (Meyer et al., 2000; Medigeshi et al., 2008) (Figure 6—figure supplement 2). In the screen of potential functional motifs, the Minimotif Miner database (Mi et al., 2012) predicted an AP-binding motif ([DE]D) in the cytoplasmic tail of m154. To investigate the possible involvement of this motif in the protein’s function, we have generated a new MCMV mutant (Figure 6B) where two aspartic acid residues of the motif (DD) were substituted with two alanine (AA) amino acids (m154-DDAA viral mutant). Currently, the only available antibody that recognizes m154 protein (Zarama et al., 2014), was generated by the immunization with a synthetic peptide corresponding to its cytoplasmic tail. Therefore, to be able to characterize the mutated protein, we introduced an HA tag at the end of the m154 sequence in m154-DDAA mutant. We have confirmed the expression of the mutated form of m154 in m154-DDAA virus by immunoblotting HA tag in lysates of infected cells (Figure 6—figure supplement 3A). The m154-DDAA mutant displayed plaque morphology and in vitro growth kinetics that were indistinguishable from those of the ∆m154 and WT virus (Figure 6—figure supplement 3B). Next, we showed that m154 protein with a mutated DD motif has reduced ability to localize to the AP-1 compartment in infected cells (Figure 6C) and consequently, reduced ability to downregulate the surface expression of its targets (Figure 6D). The wild-type m154 with the C-terminal HA tag remained fully functional and downregulated its targets from the surface of infected cells (Figure 6—figure supplement 4). In conclusion, we have characterized the DD motif in the cytoplasmic tail of m154 that is responsible for its localization to AP-1 compartment and immunomodulatory function.

It was previously shown that the viral mutant lacking m154 is attenuated early after infection in NK-dependent manner, attributing this phenotype to CD48, the first described target of m154 (Zarama et al., 2014). Considering that the attenuation was reported for intraperitoneal (i.p.) viral inoculation, we also confirmed the attenuation of Δm154 four days post-infection (p.i.) in an intravenous (i.v.) route of infection (Figure 7—figure supplement 1A). Moreover, we observed an even stronger attenuation of Δm154 virus at a later time point of infection (Figure 7—figure supplement 1B). Then, we compared the in vivo phenotype of Δm154 and m154-DDAA virus that bears a mutation in the cytoplasmic tail motif. We confirmed the functional role of the DD motif by showing the same levels of attenuation of Δm154 and m154-DDAA viral mutants in different organs and at different times points of infection (Figure 7A).

Given the fact that we observed the attenuation of m154-DDAA virus at 7 d p.i. and having showed how the m154 protein downregulates surface molecules involved in CD8⁺ T cell response, we next determined the role of CD8⁺ T cells in the control of this virus. While the depletion of NK cells had an effect on the titer of m154-DDAA virus seven days p.i., the additional depletion of CD8⁺ T cells resulted in a substantial increase of viral titer, supporting the role of CD8⁺ cell subset in the observed phenotype (Figure 7B). Furthermore, these results were comparable to ∆m154 infection with NK and CD8⁺ T cell depletion (Figure 7—figure supplement 1C). Of note, CD4⁺ cell depletion did not further significantly contribute to the observed phenotype (Figure 7—figure supplement 1C).

To gain more insight into the role of DD motif in CD8⁺ T cell response, we next compared CD8⁺ T cells from WT and m154-DDAA-infected mice. We observed increased frequency and number of MCMV pp89-specific CD8⁺ T cells in m154-DDAA-infected mice as compared to WT MCMV (Figure 7C). To assess the effector function, CD8⁺ T cells from WT and m154-DDAA-infected mice were stimulated in vitro with MCMV pp89 peptide. Upon 5 hr of stimulation, we detected increased frequency and number of IFN-γ-producing CD8⁺ T cells in mutant infection compared to WT MCMV (Figure 7D) indicating that the DD motif of m154 protein impairs CD8⁺ T cell response.

To address the effect of m154 on the ability of antigen-presenting cells to induce CD8⁺ T cell response, we co-cultivated infected bone marrow-derived dendritic cells (BDMCs) with naive CD8⁺ T cells (Figure 7E, schematic model). As a source of CD8⁺ T cells, we used splenocytes from T cell receptor (TCR)-transgenic mice in which 90% of CD8⁺ T cells are specific for MCMV M38 peptide (Maxi mice) (Torti et al., 2011). m154-DDAA-infected BMDCs induced significantly stronger IFN-γ response by CD8⁺ T cells as compared to the WT MCMV infection, confirming the role of m154 protein, in particular, its DD motif, in the evasion of CD8⁺ T cell immune response (Figure 7E). All of the indicated findings were also relevant for Δm154 infection (Figure 7—figure supplement 1D–F).

Altogether, we found that m154 impairs CD8⁺ T cell response and the virus lacking this protein or its functional motif is attenuated in vivo due to the increased capacity to stimulate virus-specific CD8⁺ T cells.

B12 (SV40-immortalized BALB/c fibroblasts), MEF (mouse embryonic fibroblasts from BALB/c mice) and HFF (human foreskin fibroblasts) were cultivated in Dulbecco’s modified Eagle’s medium supplemented with 10% or 3% of fetal calf serum (FCS) and 100 U/ml of penicillin, 100 μg/ml streptomycin and 2 mM L-glutamine. DC2.4 cells (immortalized dendritic cells from C57BL/6 mouse) were cultivated in Roswell Park Memorial Institute (RPMI) 1640 medium supplemented with 10% of FCS and 100 U/ml of penicillin, 100 μg/ml streptomycin and 2 mM L-glutamine, w/o β-mercaptoethanol. To obtain bone marrow-derived dendritic cells (BMDCs), cells from the bone marrow of C57BL/6 mice were cultured for 7 days in RPMI supplemented with 10% of FCS and 20% of supernatant from J558 plasmacytoma cells as a source of GM-CSF cytokine (Alloatti et al., 2016). Peritoneal macrophages were elicited from peritoneal exudate cells (PECs) four days following i.p. injection of 1 ml of 3% thioglycollate into BALB/c mice. PECs were collected by peritoneal lavage, plated out at 2 × 10⁵ cells/ml in supplemented RPMI 1640 medium and incubated for 2 hr at 37°C, 5% CO₂, after which non-adherent cells were washed away with phosphate buffered saline (PBS). HFF, MEF, B12 and DC2.4 have been tested for mycoplasma. There is no information about cell sex for B12 or DC2.4 cell lines available. MEF cells were mixed sex since they were obtained by homogenisation of total embryos 17 days after fertilization. BMDC cells were obtained from female C57BL/6 mice, and PECs from both male and female BALB/c mice. M154-HA fusion protein was constructed as described in Zarama et al. Stable transfectants were obtained by transfecting B12 cells with Extreme Gene nine transfection reagent, according to the manufacturer's protocol. HA-high population was sorted using BD FACSAria II and selection was performed with 0.7 mg/ml of G418. Since μ1Α‐deficiency is embryonic lethal at embryonic day 13.5, the AP‐1µ1A deficient cell lines (here referred AP1-KO) were previously established by continuous passaging of fibroblasts from embryonic day 11.5 mouse embryos (Meyer et al., 2000). The AP1-KO cells and the rescued cell line which served as the control cell line (here referred AP1-WT) were cultivated in 10% DMEM without or with hygromicin selection (200 µg/ml) (Meyer et al., 2000; Medigeshi et al., 2008). All cells were grown in an incubator that enables conditions of 5% CO₂ and 37⁰C.

All viruses were propagated on primary BALB/c MEFs and titrated by standard plaque assay as described in Brizić et al. (2018). The BAC pSM3fr-derived MCMV, C3X, based on the MCMV Smith strain (ATCC VR-1399) (Wagner et al., 1999) and a recombinant MCMV-GFP (Mathys et al., 2003) were used as wild type MCMVs. The generation of the following recombinant MCMV mutants was described elsewhere: mutants which lack different sets of genes from the region m144-m158 (also called MCMV-Δ6 region) were described in Brune et al. (2006), the Δm138/fcr-1 mutant in Crnković-Mertens et al. (1998) and Δm154 and ∆m154Int mutant in Zarama et al. (2014). The MCMV mutant m154-DDAA, containing the DD to AA substitution and the sequence encoding HA-tag at the C-terminal portion of the m154 gene, was constructed by en-passant mutagenesis (Tischer et al., 2010), in cells containing pSM3fr BAC. The following primers were used: m154-CTermDelFlag-F: 5'-TCACCGTTGTGATTTTATCTGGGATCGCCGCGGGAGTACTCCTGATCACATACCCATACGATGTTCCAGATTACGCTTGAAGGATGACGACGATAAGTAGGG-3'; m154-CtermDelFlag-R: 5'-AAACACCGCACCAGAGACCAAGTATAAAGCAGTTTTATTGAGCTGATGAGTCAAGCGTAATCTGGAACATCGTATGGGTATGTGATCAGGAGTGTATATCTGGCCCGTACATCGATCT-3'; m154-CTermDelFlag-short-F: 5'-TCACCGTTGTGATTTTATCT-3'; m154-CtermDelFlag-short-R: 5'-AAACACCGCACCAGAGACCA-3'; Pep-Kan-F: 5'-AGGATGACGACGATAAGTAGGG-3'; Pep-Kan-R: 5'-GTATATCTGGCCCGTACATCGATCT-3': ORF-Kan-F: 5'-ATGAGCCATATTCAACGGGA-3'; ORF-Kan-R: 5'-CTCATCGAGCATCAAATGAAA-3'; m154-AA-Flag-F: 5'-GATCGCCGCGGGAGTACTCCTGATCACACACCGTTGGGAAGCAGCAAAGGGTGGGGAGGTGGCACTCGGGGAAGGTTATGACGAGTCTTATGTGTACCCATACGATGTTCCAGATTACGCTTGAAAGGATGACGACGATAAGTAGGG-3'; m154-AA-Flag-R: 5'- AAACACCGCACCAGAGACCAAGTATAAAGCAGTTTTATTGAGCTGATGAGTCAAGC GTAATCTGGAACATCGTATGGGTACACATAAGACTCGTCATAACCTTCCCCGAGTGCCACCTCCCCACCCTTTGCTGCGTATATCTGGCCCGTACATCGATCT-3'; m154-AA-Flag-F-short: 5'-GATCGCCGCGGGAGTACTCC-3' and m154-AA-Flag-R-short: 5'- AAACACCGCACCAGAGACCAA-3'. Using analogous approach, a virus encoding HA-tag at the C-terminal portion of the wild-type m154 allele (WT-HA) was constructed using primers: m154-HA-R-long: 5’-AAACACCGCACCAGAGACCAAGTATAAAGCAGTTTTATTGAGCTGATGAGTCAAGCGTAATCTGGAACATCGTATGGGTACACATAAGACTCGTATATCTGGCCCGTACATCGATCT-3’, KanF: 5’-AAGGATGACGACGATAAGTAGGG-3’, m154-HA-F-long: 5’-ATAAGGGTGGGGAGGTGGCACTCGGGGAAGGTTATGACGAGTCTTATGTGTACCCATACGATGTTCCAGATTACGCTTGAAAGGATGACGACGATAAGTAGGG-3’ and m154-HA-short-R: 5’-AAACACCGCACCAGAGACCAAG-3’. The accuracy of desired genomic modifications at the target loci in the recombinant viruses was verified by sequencing. Subsequently, all recombinant viruses were reconstituted by transfection of recombinant BAC into primary MEFs, and the expression of the mutated form of the m154 protein encoded by this virus was confirmed by immunoblotting HA-tag from lysates of infected cells. Plaque morphologies and growth kinetics were determined and compared to the parental virus. WT MCMV strains and various MCMV mutants used in this study were propagated on MEF, and virus stocks were prepared as described previously (Brizić et al., 2018). Viral titers were determined by standard plaque assays on MEFs. For infection of HFF cells, we used the HCMV strain TB40/E (Sinzger et al., 2008).

Escherichia coli strain GS1783 (Tischer et al., 2010) was used for the preparation of MCMV BAC pSM3fr (Wagner et al., 1999).

BALB/c, C57BL/6, and Maxi mice (Torti et al., 2011) (MHC-I-restricted TCR-transgenic mice with specificity for the inflationary M38 epitope) were housed and bred under specific-pathogen-free conditions at the Central Animal Facility, Faculty of Medicine, University of Barcelona and Central Animal Facility, Faculty of Medicine, University of Rijeka. All procedures involving animals and their care were approved by the Ethics Committee of the University of Barcelona (protocol number CEEA 308/12) and the Animal Welfare Committee at the University of Rijeka and National ethics committee (Croatia) (Approval reference 525-10/0255-17-4). All procedures were conducted in compliance with institutional guidelines as well as with national (Generalitat de Catalunya decree 214/1997, DOGC 2450) and international (Guide for the Care and Use of Laboratory Animals, National Institutes of Health, 85–23, 1985) laws and policies. Female and male mice aged 8–16 weeks were used in the study.

The immunogen was a recombinant mouse CD155 protein expressed in Fc mut pIRESpuro vector (Glasner et al., 2012) and purified from the supernatant of transfected HEK 293 T cells. The recombinant protein was designed in such a way that the CD155 ectodomain is fused with the human IgG1 Fc fragment (mCD155-Fc). It was purified using prepacked chromatography Protein G HP column (GE) for AKTA systems (AKTA purifier). We used the Dark Agouti (DA) rat strain for immunization. DA rats were injected with 50 µg of the immunogen in complete Freund’s adjuvant and 2 weeks later reinjected in incomplete Freund’s adjuvant. After 2 weeks, the sera were screened for the antibody titer. The best responders were boosted with the immunogen in PBS. Three days later, spleen cells were collected, and after lysis of red blood cells, fused with SP2/0 cells. The cells were seeded in 20% RPMI 1640 medium containing hypoxanthine, aminopterin and thymidine for hybridoma selection and screened for mAbs production using ELISA. mPVR.01 antibody clone was further selected based on performance in immunofluorescence, flow cytometry and immunoprecipitation methods.

In vitro viral growth was analyzed by infecting MEFs with WT MCMV, ∆m154 or m154-DDAA virus at 0.1 PFU/cell. At indicated days p.i. culture supernatants were collected and frozen at −80°C. Further, the amount of extracellular infectious virus present in the culture supernatant was determined by a standard plaque-forming assay (Brizić et al., 2018) in two technical replicates.

Cell lysates of DC2.4 cells were prepared using NP-40 lysis buffer (10 mM Sodium phosphate pH 7.2, 150 mM sodium chloride, 2 mM EDTA, 1% NP-40, protease inhibitors). Cells were infected with 3 PFU/cell of WT MCMV or indicated viral mutants. Following infection, cells were supplemented with lysosomal inhibitor leupeptin (75 μg/μl) and lyzed after 20 hr. 75–100 μg of lysate proteins were separated by 10–12% SDS-PAGE electrophoresis and transferred onto 0.45 μm PVDF membranes. Membranes were blocked with 5% w/v nonfat dry milk and incubated with anti-m154 (clone m154.4.113); anti-actin (clone C4), anti-HA Peroxidase (clone 3F10) and anti-MCMV IE1 (clone IE1.01). Protein signals were visualized using UVITec imaging system. For immunoprecipitation of biotinylated mCD155, cells were biotinylated via standard protocol (Thermo Scientific Pierce Cell Surface Protein Isolation/biotinylation), lysates were prepared as above and incubated overnight at 4°C under rotation with anti-mCD155 mAb (clone mPVR.01) followed by 1 hr incubation with protein G sepharose beads (50 μl). The precipitates were washed five times (1 ml each) with IP buffers before the samples were subjected to the western blot analysis. The membranes were incubated with peroxidase-conjugated streptavidin and visualized as described above. At least two independent experiments were performed for each subject of analysis.

B12 and DC2.4 cells were infected with 1–3 PFU/cell of WT MCMV or indicated viral mutants. HFF cells were infected with 1 PFU/cell of HCMV strain TB40/E for 72 hr. In siRNA silencing assays (Figure 6A and Figure 6—figure supplement 1), short interfering RNAs (siRNAs) pre-designed by Thermofisher: Stealth siRNAs MSS202065 (here referred siRNA65) and MSS273190 (here referred siRNA90) were used to silence adaptor-related protein complex AP-1, mu subunit one gene, alone or in combination. The procedure followed Lipofectamine RNAiMAX reagent protocol with the OPTI MEM transfection medium, LIPOFECTAMINE RNAIMAX transfection reagent, BLOCKIT ALEXA FLUOR RED OLIGO positive control of transfection and STEALTH RNAI NEG CTL HI GC negative control siRNA. In the antibody internalization assays (Figures 3, 4 and 5 and Figure 5—figure supplement 2), the m154 N-terminal HA transfectant was used, or in the case of infection, the virus lacking m138 gene (Δm138) that encodes for a viral homolog of Fc receptor was used to exclude the possibility of Fc-antibody binding to the viral Fc receptor during 20 hr of incubation. The equivalence of the WT MCMV and Δm138 virus in terms of specific protein regulation was confirmed in a separate set of analyses. Following infection, cells were treated with leupeptin (75 μg/μl) where indicated and 20 h p.i. fixed and analyzed for target proteins localization. The following in-house produced antibodies were used: anti-mCD155 (clone mPVR.01), anti-hCD155 (clone hPVR.16), anti-MCMV m20.1 (clone m20.1.02), anti-m154 (clone m154.4.113), anti-CD229 (clone 7.144.2). The following purchased antibodies were used: anti-mouse calnexin, anti-mouse γ1-adaptin, anti-HA, anti-mouse CD18 (clone M18/2), anti-mouse CD47 (clone miap301). The matching isotype controls were used. For lysosomal visualization, an acidic compartments probe DAMP (N-(3-[2,4-dinitrophenyl amino] propyl)-N-(3-aminopropyl)methylamine) was used. The samples were stained with secondary antibodies coupled to fluorescein isothiocyanate (FITC) or tetramethylrhodamine (TRITC) fluorophores, mounted using Mowiol mounting medium and analyzed at room temperature with Olympus FV300 confocal laser scanning microscope using Olympus-Microscope-PlanApo-60X-NA1.4-oil objective and FluoView acquisition software without gamma adjustments. At least two independent experiments were performed for each subject of analysis and at least two technical replicates/experiment were analyzed.

Flow cytometry was performed according to the Guidelines for the use of flow cytometry and cell sorting in immunological studies (Cossarizza et al., 2017). Uninfected or cells infected with indicated MCMV strains were stained for cell surface markers: CD155 (clone mPVR.01), CD229 (clone 7.144.2), CD18 (clone M18/2), CD48 (clone HM48-1), CD47 (clone miap301), CD54 (clone 1H4), CD84 (clone mCD84.7), CD162 (clone 2PH1), CD270 (clone HMHV-1B18), CD80 (clone MEM-233), CD262 (clone MD5-1), CD112 (clone 502–57), MULT-1 (clone 1D6), or anti-RAE-1ɛ mAb (Clone 205001). The matching isotype controls were used. In siRNA silencing assays (Figure 6A) we monitored CD155 protein expression on B12 cells following the protocol described in the Immunofluorescence microscopy section. For the viral strains expressing GFP, infected cells were gated as GFP⁺. Otherwise, where needed, infected cells were gated based on the expression of MCMV protein m06 or m04, revealed by staining with anti-m06 (clone croma229) or anti-m04 (clone m04.17). The following secondary antibodies were used: anti-rat IgG F(ab’)2-PE or FITC, anti-mouse IgG F(ab’)2-PE, FITC, APC or PE-streptavidin. Fixable Viability Dye coupled to Alexa Fluor 780 was used to exclude dead cells. When needed, the blocking with 20% fetal rabbit serum and 1% fetal bovine serum was performed before immunostaining. For IFNγ staining, splenic leukocytes were prepared using standard protocols and incubated in RPMI medium supplemented with 10% of FCS and 1 µg/well of pp89-derived peptide for 5 hr at 37°C. Brefeldin A at 1 µg/ml was added immediately to the cells. Before staining, Fc receptors were blocked with anti-mouse CD16/CD32 (clone 2.4G2) (Yokoyama and Kim, 2008). Subsequently, cells were surface stained, fixed and permeabilized, followed by intracellular staining with anti-IFNγ. The following reagent was obtained through the National Institutes of Health (NIH) Tetramer Core Facility: MCMV-specific tetramer pp89-PE. MCMV pp89-derived peptide was obtained from JPT Peptide Technologies. Flow cytometry data were acquired on FACSAriaIlu, FACScan (BD Biosciences) or LSRFortessa (BD Biosciences), and analyzed using FlowJo_v10 (Tree Star) software. At least three independent experiments were performed for each subject of analysis.

For each experiment, mice were pooled and randomized in experimental groups of five animals. Female mice aged 8–12 weeks were i.v. injected with tissue culture-grown recombinant MCMV strains with 2 × 10⁵ PFU/animal, in a volume of 500 μl of pure DMEM. Depletion of NK cells was performed by intraperitoneal (i.p.) injection of 20 μl of anti-asialo GM1, depletion of CD8⁺ T cells by i.p. injection of 150 μg of anti-CD8 antibody (YTS 169.4) and depletion of CD4⁺ T cells by i.p. injection of 150 μg of anti-CD4 antibody (YTS 191.1). Animals were sacrificed at indicated time points, organs were harvested and viral titers were determined by a standard plaque-forming assay (Brizić et al., 2018) in two technical replicates. At least two independent experiments were performed for each subject of analysis.

Bone marrow-derived dendritic cells (BMDCs) from C57BL/6 mice were obtained as described above and infected with 3 PFU/cell of indicated viruses for 24 hr and used as antigen-presenting cells. Further, they were co-cultured at indicated ratios with splenocytes isolated from Maxi mice during 6 hr in the presence of Brefeldin A at 1 μg/ml. The number of IFNγ⁺ CD8⁺ T cells within the total splenocyte population was determined by flow cytometry. At least three independent experiments were performed for each subject of analysis with at least two technical replicates/experiment.

Results are expressed as mean + /- standard deviation of at least three independent experiments. For in vivo experiments, mice were pooled and randomized in experimental groups of five animals. The sample size was determined by pilot studies and according to accepted practice in previous literature utilizing the MCMV model. The selection of the appropriate statistical test was based on the number and distribution of data points per group. Confidence interval was set to 95% and a level of p<0.05 was considered to be statistically significant and assigned as *p<0.05, **p<0.001, ns p>0.005. Exact p values and statistical details of experiments are indicated in the figure legends. Statistical differences between the two study groups were evaluated using paired or unpaired, two-tailed t-test or the non-parametric Mann-Whitney test. Statistical differences between more than two study groups were evaluated using the Kruskal-Wallis bidirectional test. All statistical analyses were performed using GraphPad Prism eight software.

The Key Resources Table (Supplementary file 1) lists the reagents and resources used in this study.

Further information and requests of the materials and resources described in the article should be directed to and will be fulfilled by the Correspondence Contact, Tihana Lenac Rovis (tihana.lenac@medri.uniri.hr) or Stipan Jonjic (stipan.jonjic@medri.uniri.hr).