In these studies the HCMV TB40E/IE2-EYFP strain expressing IE2-EYFP fusion protein (Wagenknecht et al., 2015) and the HCMV TB40E/UL84Pr-luc expressing luciferase under the control of the UL84 promoter were used.

Virus stocks were prepared by infection of human foreskin fibroblasts (HFFs) with the respective viral strain. HFFs were cultured in Dulbecco’s minimal essential medium (DMEM; Lonza, Basel, Switzerland) supplemented with 10% FCS (PAA, Cölbe, Germany), 2 mM L-glutamine (Lonza), 100 U/ml penicillin (Lonza), and 100 U/ml streptomycin (Lonza) until a complete cytopathic effect was achieved. Cellular debris was removed by centrifugation and virus containing cell culture supernatant was stored -80°C.

Virus titers for HCMV strains were determined as described elsewhere (Andreoni et al., 1989; Wagenknecht et al., 2015).

Monocyte-derived DCs were generated as described previously (Heilingloh et al., 2014). In brief, peripheral blood mononuclear cells (PBMCs) were isolated from different healthy donors using a Lymphoprep gradient (Nycomed Pharma AS, Oslo, Norway) and afterward monocytes were separated using plastic adherence. By addition of 800 U/ml granulocyte macrophage-colony stimulating factor (GM-CSF; CellGenix, Freiburg, Germany) and 250 U/ml IL-4 (Milteny, Bergisch-Gladbach, Germany) monocytes differentiated to immature DCs. Maturation was induced by adding 10 ng/ml TNF-α (Beromun, Boehringer Ingelheim, Germany), 1 mg/ml prostin E2 (PGE2; Pfizer, NY, USA); 200 U/ml IL-1β (CellGenix), 40 U/ml GM-CSF, 1000 U/ml IL-6 (CellGenix), and 250 U/ml IL-4 to the medium. The maturation status was controlled by flow cytometry, and mDCs were used for further experiments.

For infection of mDC, 5 × 10⁵ cells, as a maximum, were washed with PBS and resuspended directly in virus stock at a multiplicity of infection (MOI) of 2. For mock-infected samples, mDCs were resuspended in the equal volume of DMEM supplemented with 10% FCS, 2 mM L-glutamine, 100 U/ml penicillin, and 100 mg/ml streptomycin. Infection was performed in a microcentrifuge at 23 × g for 1 h at room temperature. Subsequently, cells were transferred into a 24 well plate and Roswell Park Memorial Institute medium (RPMI 1640, Lonza) supplemented with 1% autologous serum (Sigma Aldrich, Deisenhofen, Germany), 10 mM HEPES (Lonza), 2 mM L-glutamine, 100 U/ml penicillin, 100 mg/ml streptomycin, 40 U/ml GM-CSF, and 250 U/ml IL-4 was added to a final volume of 0.5 ml.

Mock- or HCMV-infected mDCs were allowed to adhere on poly-L-lysin (Sigma Aldrich) coated glass cover slips, fixed with 4% paraformaldehyde and permeabilized using 0.2% Triton-X-100. Blocking was performed using 1% BSA in PBS. Antibodies used for Immunofluorescence staining: anit-CD83 (HB15a; Beckmann Coulter, Brea, USA); AlexaFluor647-conjugated secondary antibody (Invitrogen, Carlsbad, CA, USA). For mounting and additional nuclear staining Roti^®-Mount FluorCare DAPI (Carl Roth) was used. Confocal microscopy was performed using a LSM780 microscope (Zeiss, Oberkochen, Germany).

Proteasome inhibitors MG-132 (Enzo Life Sciences, Lörrach, Germany), Epoxomicin (Enzo Life Sciences), and Bortezomib (Santa Cruz Biotech. Inc, Heidelberg, Germany) were added 1 h post-infection (hpi) at a final concentration of 10, 5, and 2 μM, respectively.

Cycloheximide (CHX) and actinomycin D (ActD; Sigma Aldrich) were used at a final concentration of 100 and 5 μg/ml, respectively.

Detection of YFP expression of HCMV and changes in expression of cell surface molecules were analyzed by flow cytometry using antibodies specific for CD83 (clone HB15e; BD Biosciences, Heidelberg, Germany), CD80 (clone L307.4; BD Biosciences), and HLA-DR (clone G46-6, BD Biosciences). To assure analysis of living cells, samples were additionally stained using LIVE/DEAD^® Fixable Violet Dead Cell Stain Kit (Life Technologies, Carlsbad, CA, USA). Unstained cells were used as negative control.

Cells were sorted based on their YFP expression using a BD Aria FACS Sorter (BD Biosciences).

Mature DCs (mDCs; 1 × 10⁶) were harvested and resuspended in 50 μl lysis buffer (10% glycerol, 2 mM EDTA, pH 8, 137 mM NaCl, 10 mM sodium phosphate, pH 7.2, 1% NP-40, 2 mM phenylmethanesulfonyl fluoride, 20 mM sodium fluoride). After 20 min on ice, lysates were cleared by centrifugation at 4°C and full speed for 20 min. Protein samples were separated using SDS-polyacrylamide gel electrophoresis (SDS-PAGE) and subsequently transferred onto a nitrocellulose membrane. After blocking with 5% (w/v) dry milk or 1x RotiBlock (Carl Roth, Karlsruhe, Germany) the membrane was incubated with the primary antibody at 4°C overnight. After incubation with the appropriate secondary HRP-labeled antibody, detection was performed using Amersham ECL Prime Western Blotting Detection Reagent (GE Healthcare, Solingen, Germany). The following primary antibodies were used for Western blot analyses: anti-CD83 (clone F-5, Santa Cruz Biotech Inc., Heidelberg, Germany, 1:1000), anti-β-actin (clone AC-74, Sigma Aldrich; 1:2000), anti-GAPDH (clone 6C5, Millipore, Schwalbach, Germany), and anti-CMV pp86 (IE2, clone 12E2, Santa Cruz Biotech Inc; 1:1000).

Soluble CD83 levels were determined using a sandwich ELISA. Ninety-six-well plates (Maxisorp; Nunc, Thermo Scientific) were pre-coated for 1 h at 37°C with anti-CD83 (clone HB15a, Beckman Coulter, Brea, CA, USA), blocked with 10% goat serum (Life Technologies), and finally samples were added. After an overnight incubation at 4°C wells were incubated with 10 μg/ml rabbit-anti-CD83 (Sigma Aldrich) in diluent (5% goat serum, 5% mouse serum, and 1% non-fat dry milk in PBS) for 1 h at 37°C. Next, wells were incubated with goat-anti-rabbit-biotin (Dako, Eching, Germany, 1:5000) for 1 h at 37°C following a 1 h incubation at 37°C with streptavidin-HRP (Dako, 1:2000) in 0.5% BSA in PBS. Between each antibody incubation step, plates were thoroughly washed with 0.1% Tween 20 in PBS. Standard curves for the estimation of sCD83 concentration were generated using serial dilutions of recombinant sCD83 and the results were expressed as a protein concentration based on the protein standard. Plates were developed using BD OptEIA (BD Biosciences) and measured using a Tecan infinite^® F500 plate reader.

Mature DCs were mock-infected or infected with HCMV TB40E/IE2-EYFP at an MOI of 2 in presence of 100 μg/ml CHX. Cells were cultured in RPMI 1640 supplemented with 1% autologous serum, 10 mM HEPES, 2 mM L-glutamine, 100 U/ml penicillin, 100 mg/ml streptomycin, 40 U/ml GM-CSF, 250 U/ml IL-4, and 100 μg/ml CHX. After a 6 h incubation at 37°C, mDCs were washed twice in RPMI 1640 supplemented with either 5 μg/ml ActD or the equal volume of DMSO and cultured for additional 16 h as described above in the presence of ActD or DMSO, respectively. Finally, cells were harvested and surface expression of CD83 was analyzed by flow cytometry. The efficiency of blocking early gene expression was controlled by infection of mDCs with HCMV TB40E/UL84Pr-luc treated in parallel as described above, using a luciferase reporter assay.

Luciferase activity of HCMV TB40E/UL84Pr-luc infected mDCs was performed in triplicates using the Dual-Luciferase Reporter Assay System (Promega, Mannheim, Germany) according to the manufacturer’s recommendations. The relative luminescence units (RLU) were determined using the Tecan infinite^® F500 plate reader.

Mature DC (3 × 10⁶) were electroporated with 3 μg of the IE2-expression vector pHM134 (Lang et al., 1995) using P3 Primary Cell 4D-Nucleofector^® X Kit (Pulse: EH-100) and the 4D-Nucleofector Device (Lonza) according to the manufacturer’s instructions. After electroporation, the nucleovette was immediately flushed with 500 μl warm RPMI 1640 (Lonza) and cells were transferred to a 24-well tissue culture plate. RPMI 1640 supplemented with 1% autologous serum, 10 mM HEPES, 2 mM L-glutamine, 100 U/ml penicillin, 100 mg/ml streptomycin, 40 U/ml GM-CSF, and 250 U/ml IL-4 were added at a final concentration of 0.5 × 10⁶ cells/ml and mDCs were incubated for 18–20 h.

Results are displayed as means ± standard deviation (SD). For multiple comparisons, data were analyzed using One-Way ANOVA and Bonferroni’s Multiple Comparison post hoc test. For comparison of two data sets p-values were determined using the student’s t-test. Significance was accepted if p was <0.05.

For the generation of monocyte-derived DCs from leukapheresis products of healthy donors, a positive vote from the local ethics committee has been obtained (reference number 4556). This study was carried out in accordance with the recommendations of the ethics committee of the “Friedrich-Alexander-Universität Erlangen-Nürnberg” with written informed consent from all subjects. All subjects gave written informed consent in accordance with the Declaration of Helsinki.

It has previously been reported that HSV-1-induced CD83 degradation after infection of mDCs can be prevented by inhibition of the proteasome (Kummer et al., 2007). To investigate whether a similar mechanism is responsible for HCMV-induced CD83 down-modulation, infection experiments using proteasome inhibitors were performed. First, mDCs were mock-infected or HCMV-infected and 1 hpi, cells were treated with different proteasome inhibitors (10 μM MG-132, 2 μM Bortezomib, or 5 μM Epoxomicin) or DMSO as control. After 16 h, mDCs were analyzed for their CD83 surface expression by flow cytometry (Figures 2A,B). Figure 2A shows the results of one representative experiment while Figure 2B illustrates the results of three independent experiments performed with cells of different healthy donors. Mock-samples (Figure 2A, gray histograms and Figure 2B, black bars) show no CD83 reduction whereas in the HCMV positive sample without any inhibitor a significant CD83 reduction (p < 0.001) was observed. In contrast, treatment with proteasome inhibitors MG-132, Bortezomib or Epoxomicin prevented CD83 down-modulation, demonstrating that CD83 is degraded via the proteasome (Figure 2A, gray filled histograms and Figure 2B, white bars).

In a next step, we were wondering if the observed proteasomal degradation of CD83 takes place in parallel to the proposed shedding of a soluble form of CD83 (sCD83) from the cell surface. Therefore, we determined the levels of sCD83 in the supernatant of the HCMV-infected mDC cell culture. In sharp contrast to the data reported by Senechal et al. (2004), we observed no difference regarding the sCD83 levels in HCMV-infected samples compared to mock samples. Treatment with MG-132 led to an overall and therefore unspecific increase of sCD83 levels in the supernatant of mock- as well as HCMV-infected cells (Figure 2C).

These data indicate that CD83 is not shed from the cell surface of HCMV-infected mDCs but rather degraded via the cellular proteasome.

To further elucidate the mechanism leading to CD83 reduction after HCMV infection of mDCs, we aimed to identify the effector protein/s inducing CD83 down-modulation. To narrow down the phase of the viral replication cycle in which CD83 loss occurs, an infection time course experiment was performed. Therefore, mDCs were mock- or HCMV-infected and analyzed for CD83 expression at different time points post-infection using flow cytometry (Figure 3A). HCMV infection efficacy was determined in parallel by measuring YFP-expression (Figure 3B). Figure 3A illustrates CD83 reduction in HCMV-infected mDCs (white bars) in comparison to the mock control (black bar). After 12 h a significant loss of CD83 (p < 0.001) could be observed. At this time point only 50% of the infected cell population showed CD83 expression. CD83 reduction increased over time and after 24 h only 15% of mDCs expressed CD83. This CD83 down-modulation was paralleled by the increase of YFP expression, reflecting expression of IE2 (Figure 3B).

Taken together these observations clearly suggest a correlation between IE gene expression and CD83 down-modulation, leading to the hypothesis that the viral effector might be expressed in the IE or early phase of the HCMV replication cycle.

Knowing that a gene product of the IE or early phase is most likely responsible for the down-modulation of CD83, we further wanted to differentiate a possible IE effect from an early effect. The replication cycle of HCMV, as for all herpes viruses, is a strictly regulated cascade of three phases. First, IE gene products are expressed and these IE proteins are required to initiate expression of early (E) genes. The proteins of the E phase are then needed to initiate the expression of late genes. This feature can be used to discriminate between IE and E gene effects performing a cycloheximide-ActD-chase experiment. In brief, mDCs were mock- or HCMV-infected in the presence of DMSO as control or 100 μg/ml cycloheximide (CHX). This CHX treatment prevents protein translation of viral genes but allows gene transcription. After 6 h of incubation, CHX was washed out and cells were incubated in DC-medium, containing either 5 μg/ml ActD or DMSO (as a control) for additional 16 h. ActD allows translation of already existing mRNAs, but prevents the synthesis of new mRNAs. At the end of incubation, cells were harvested and stained for their CD83 surface expression.

In Figure 4A the results of the CHX-ActD-chase experiment are illustrated. Mock controls are depicted in black, HCMV-infected samples in white. CD83 was reduced significantly in CHX-DMSO-treated HCMV positive cells as well as in DMSO-treated cells. In these settings, gene products of all phases are present. In CHX-ActD-treated samples an overall reduction of CD83 surface expression is already observed in mock-infected cells. Nevertheless, there is a significant reduction of CD83 (p < 0.05) in HCMV positive CHX-ActD-treated mDCs compared to the mock control indicating that IE gene products are involved in the loss of CD83.

To verify inhibition of the expression of early and late genes by the afore described experiment, cells were infected in parallel with virus strain HCMV TB40E/UL84Pr-luc encoding a luciferase gene under the control of the promoter of the early gene UL84, and treated as described above. At the end of the incubation period a luciferase assay was performed (Figure 4B). Mock-infected mDCs served as negative control (Figure 4B, black bars). DMSO as well as CHX-DMSO-treated samples show high luciferase activity whereas CHX-ActD-treated mDCs show a luminescence intensity similar to mock-infected cells, confirming that expression of E genes and in turn L genes is inhibited.

Since our data suggested that an IE gene product is responsible for the induction of CD83-degradation, and the knowledge that levels of the IE protein IE2 increased in parallel to the loss of CD83, we next checked whether IE2 is able to induce CD83-degradation. For this purpose mDCs were electroporated with an IE2 expression plasmid and treated with or without MG-132. As mock control, cells were pulsed but without addition of any DNA (EP only). After 20 h, cells were harvested and analyzed via flow cytometry and Western blotting (Figure 5). Figure 5A illustrates the analysis of CD83 surface expression. IE2 positive cells showed a highly significant (p < 0.0001) down-regulation of CD83 surface expression (Figure 5A, white bar) which could be prevented by the addition of the proteasome inhibitor MG-132 (Figure 5A, gray bar). CD80 and HLA-DR were used as control and were not affected in the presence of IE2. Western blot analysis confirmed the results of flow cytometry analysis and verified expression of IE2 (Figure 5B, top and middle panel, respectively). GAPDH was used as loading control (Figure 5B, bottom panel).

In order to investigate whether IE2 co-localizes with CD83 in HCMV-infected mDCs, which would suggest that IE2 induces CD83-degradation via a direct interaction, we performed immunofluorescence analyses (Figure 5C). Therefore, mDCs were mock- or HCMV-infected, treated with or without MG-132 and analyzed for their CD83 as well as IE2 expression. The treatment of HCMV-infected mDCs with MG-132 restored CD83 expression reaching levels of the respective mock-condition. However, no co-localization of IE2 and CD83 was observed in the context of an HCMV-infection hinting toward an indirect mechanism for IE2-induced CD83-degradation.

Taken together these data clearly demonstrate that HCMV infection of mDCs does not lead to a shedding of a soluble form of CD83 from the cell surface but to proteasomal degradation of CD83. Furthermore, we were able to show that the major IE protein IE2 is sufficient to induce this degradation even in the absence of any additional viral factor.