HEK293T and HeLa cells were cultured in DMEM (Thermo Fisher) supplemented with 10% fetal calf serum (FCS; Thermo Fisher) and 1% Penicillin/Streptomycin (Life Technologies). Huh7.5 and Huh7-Lunet cells, originally obtained from Charles Rice (Rockefeller University, New York), were cultured in DMEM (Thermo Fisher) supplemented with 5% fetal calf serum (FCS; Thermo Fisher), 1% Penicillin/Streptomycin, 1% Non-essential amino acids and 1% Sodium pyruvate (all Life Technologies). Stably transduced cells were cultured with additional 1 µg/ml puromycin. Cells were serum starved by culturing them in medium without FCS.

To assess cell surface receptor modulation, the LEGENDScreen™ Human Cell PE Kit (Biolegend) was used. Huh7.5 cells were electroporated with Jc1_NS5A-mtagBFP. 48 h later, cells were detached and washed, before they were ntibody stained (5x10⁴-2x10⁵ cells per well). The staining and fixing procedure, as well as data analysis was performed as described previously (Businger et al., 2021), except that measurement of the samples was conducted with a BD FACS Canto II with high-throughput sampler. The complete measured mean fluorescence intensities (MFIs) and calculations of three biological replicate screens are summarized in Supplementary Table 1 ( Table S1 ). In brief, for each receptor/antibody, the PE MFI of the non-fluorescent, i.e. HCV-negative cell population was divided by the MFI of the BFP, i.e. HCV-expressing cell population to calculate X-fold receptor downmodulation and vice versa to calculate receptor upregulation.

Plasmids were amplified in chemocompetent NEB10 or NEB Stbl3 (for CRISPR constructs) E. coli and isolated using the PureYield™Plasmid Midiprep System (Promega) according to the manual.

To generate pFK_Jc1_NS5A-mScarlet, the eGFP fluorescent protein of pFK_Jc1_NS5A-GFP (Schaller et al., 2007) was replaced by mScarlet. In brief, the mScarlet insert was amplified from pmScarlet-C1 (Addgene #85042) with primers adding XbaI (XbaI-mScarlet_fw; 5’- GTtctagaCCTCGAGCTATGGTGAGCAAGGGCGA-3’) and PmeI (meI-mScarlet_rev; 5’-CACgtttaaacCCCTTGTACAGCTCGTCCATGC-3’) restriction sites at the 5’- and 3’-ends, respectively. pFK_ Jc1_NS5A-GFP was digested with XbaI, PmeI and FastAP (Thermo Fisher) according to manufacturer’s instructions, separated by agarose gel electrophoresis, and the backbone band was cut out and isolated using NucleoSpin Gel and PCR cleanup Kit (Macherey-Nagel). Backbone and insert were ligated using T4 Ligase (Thermo Fisher) for 1 h at RT. Next, NEB10 chemocompetent bacteria were transformed with the ligation mix and plated on LB agar with 100 µg/ml ampicillin. Colonies were picked and a 5 ml culture was grown over night, followed by plasmid isolation (GeneJET Plasmid Miniprep System; Thermo Fisher). Isolated plasmids were test digested and sequenced.

pFK_Jc1_E2-mScarlet was generated according to Lee et al. (Lee et al., 2019), using a HCV genome with E2 N-terminally tagged with GFP. To generate a corresponding mScarlet expressing viral genome, a nucleotide sequence was synthesized (Genescript) starting at the Pfl23II restriction site in the E1 coding region, encoding mScarlet between E1 and E2 with the 3C peptide sequence connecting mScarlet and E2, flanked by XbaI. We additionally introduced an EcoRI restriction site between E1 and mScarlet and a BglII restriction site between mScarlet and the 3C peptide. The plasmid was created by cleaving the vector backbone [Jc1_E1(A4)_XbaI; similar to (Steinmann et al., 2007)] with Pfl23II and XbaI (Thermo Fisher), and ligated with the synthesized sequence as described above.

The fluorescent reporter construct where a fluorescent protein is N-terminally attached to the core coding region via a 2A self-cleaving peptide (pFK_Jc1_mScarlet-2A) analogous to the already described pFK_Jc1_R2A genome (Reiss et al., 2011) was generated via restriction-free cloning. The principle is described at https://www.rf-cloning.org/ and primers were designed according to this protocol (Bond and Naus, 2012). In brief, primers were designed that where half complementary to the plasmid insertion site, and half complementary to mScarlet ( Table 1 ). A PCR was performed with the insert pFK_Jc1_E2-mScarlet as template to generate a megaprimer such as that the mScarlet, between EcoRI and BglII restriction sites, is flanked by sequences homologous to the targeted insertion site. After purification of the megaprimer, it was mixed with the target plasmid, and a rolling circle PCR was performed to generate a plasmid with the insertion. The template was digested with DpnI (NEB) and the new viral genome was transformed into bacteria. To increase the chance of successful rolling circle PCR, a truncated version of pFK_Jc1_R2A was generated that only encoded the HCV genome until the end of E2. For this, pFK_Jc1_R2A was digested with SdaI and Pfl23II, and the insert was purified for ligation into pFK_Jc1_p7-half (Steinmann et al., 2007) to generate pFK_Jc1_R2A_p7-half. Then, pFK_Jc1_R2A_p7-half was digested with BcuI, HindIII and XbaI (Thermo Fisher), and the longest fragment (containing the backbone plus the HCV genome until the end of E2) was purified. Then BcuI and XbaI matching overhangs were ligated, giving rise to Jc1_R2A_short, which was then used as template for the rolling circle PCR. The HCV genome with the new reporter gene (pFK_Jc1_mScarlet-2A_short) was then digested with SdaI and Pfl23II and ligated back into a full genome plasmid using pFK_Jc1_R2A as backbone. Correct sequence was confirmed by sequencing ( Table 2 ).

To generate 293T, HeLa or Huh7.5 knock-out cells for tetraspanins CD63 and CD81, the LentiCRISPRv2 plasmid was used as described (Sanjana et al., 2014; Shalem et al., 2014). In brief, oligonucleotides with the targeting sequence and specific overhangs for ligation were ordered (Metabion international); complementary oligonucleotides were annealed and phosphorylated, and then ligated into the LentiCRISPRv2 vector. The targeting sequences 5’-GAGGTGGCCGCAGCCATTGC-3’ (CD63) and 5’-CATCGGCATTGCTGCCATCG-3’ (CD81) were used. Subsequently, respective LentiCRISPRv2 constructs (3 µg/well) were transfected with lentiviral packaging (psPAX2; 2.25 µg/well) and envelope plasmids (pMD2G; 0.9 µg/well) in HEK293T cells using JetPRIME transfection reagent (Polyplus) according to the manufacturer’s instructions in a 6-well format. LentiCRISPRv2 without an integrated targeting sequence was used as control. 24-36 h after transfection, supernatant was harvested and spun at 3200 g for 10 min at RT to get rid of cellular debris. Cells were incubated with spun supernatant for 24 h and selected with puromycin (1 µg/ml) for 2 weeks.

HEK293T cells were transfected using polyethylenimine (PEI). Cells were seeded 24 h prior to transfection until they reached a confluency of 70-80%. In brief and exemplarily for a 12-well format, 1-2 µg plasmid DNA were added to 50 µl OptiMEM (Thermo Fisher) and another 50 µl PEI mix (OptiMEM with double the amount of PEI than DNA) was added. The transfection mixture was vortexed, spun down and incubated for 15 min at RT. 100 µl of the transfection mixture was dropped onto cells, and a medium change was performed 4-6 h later or the next day. DNA and PEI amounts were adjusted accordingly for transfection in other well formats.

HeLa and Huh7.5 were transfected using JetPRIME transfection reagent (Polyplus) according to manufacturer’s instructions.

For in vitro transcription of viral RNA, the respective DNA vector was linearized by digestion using MluI (Thermo Fisher) for 1 h at 37°C ( Table 3 ). Linearized vector was then purified using the Wizard® DNA Clean-Up System (Promega) according to manufacturer’s instructions. Complete linearization was checked by agarose electrophoresis, and DNA concentration was measured. 1 µg linearized vector was used for in vitro transcription using the T7 RiboMAX™Express Large-Scale RNA Production System (Promega) according to the manual. After DNA vector digest, a phenol chloroform RNA extraction was performed. Samples were filled up to 200 µl with nuclease-free water and 200 µl phenol:chloroform:isoamylalcohol (25:24:1; Thermo Fisher) was added, then vortexed for 1 min and spun at max speed for 2 min. The upper phase was transferred to a new tube, and 200 µl chloroform:isoamylalcohol (24:1; Sigma Aldrich) was added. Again, the sample was vortexed for 1 min, spun at max speed for 2 min, and the upper phase was transferred to a new tube. Subsequently, 20 µl 3 M sodium acetate (pH 5.2) and 200 µl isopropanol were added and the sample placed in ice for 5 min. Then, the sample was spun for 10 min at max speed to pellet the RNA, supernatant was discarded and the pellet washed with 70% EtOH. Finally, the pellet was dried at 37°C for 5 min and resuspended in 40 µl nuclease-free water. RNA was stored at -80°C.

Electroporation was performed using the Neon Electroporation System (Thermo Fisher). For each electroporation 1-4x10⁵ (10 µl tip) or 1-4x10⁶ (100 µl tip) cells were used. Cells were seeded at the respective density 24 h prior to electroporation. Then, cells were detached and washed three times with PBS. Subsequently, cells were resuspended in an appropriate volume of PBS (containing Ca²⁺ and Mg²⁺; Thermo Fisher), and viral RNA was added (0.25-1 µg/1x10⁶ cells). The reaction chamber was filled with buffer E (buffer E2 for 100 µl tips). The cell/RNA mixture was put into an electroporation tip, and one pulse with 1300 V for 30 ms was applied. Electroporated cells were seeded accordingly in medium that did not contain any antibiotic.

Cellular RNA of 2-5x10⁵ cells was extracted using the RNeasy Mini Kit (Qiagen) according to manufacturer’s instructions. For lysis, 1% 2-mercaptoethanol was added to the lysis buffer. 200 ng of extracted cellular RNA was used for cDNA transcription using the QuantiTect Reverse Transcription Kit (Qiagen) according to manufacturer’s instructions. cDNA samples were filled up to 60 µl with nuclease-free water. qRT-PCR measurements were carried out on a Lightcycler 480 (Roche) using Lightcycler 480 multiwell plates (Roche) and Luna Universial qPCR Master Mix (NEB) according to the manual. In brief, a master mix containing primers (final concentration 0.3 µM), Luna Universial qPCR Master Mix and nuclease free water was generated and added to the wells together with 2 µl of diluted cDNA in duplicates ( Table 4 ). The ΔΔCp method was used for analysis.

Cells were lysed with RIPA buffer (10 mM Tris-HCl (ph 7.4), 1 mM EDTA, 0.5 mM EGTA, 140 mM NaCl, 0.1% (v/v) Na-deoxychalate, 0.1% (v/v) SDS, 1% (v/v) Triton X-100, 1x protease and phosphatase inhibitor) for 20 min at 4°C. Subsequently, 6x sample buffer (0.5 M Tris (pH 6.8), 0.6 M DTT, 30% (v/v) glycerol, 10% (w/v) SDS, 2% (w/v) bromphenol blue) was added accordingly, and the lysate was heated to 95°C for 10 min. Then, samples were loaded on a 12% acrylamide gel and separated at 80-140 V for 90-150 min. Blotting was performed in a wet blotting chamber (BioRad) at 80 V for 90 min. Then, the membrane was blocked for 1 h at RT using 5% milk or 1% BSA in PBS or TBS. Primary antibodies were applied over night at 4°C. Secondary antibodies were applied 1h at RT. Between steps, the membrane was washed three times with PBS-T or TBS-T (PBS or TBS with 0.1% Tween 20; Table 5 ). An Odyssey Fc Imaging System (LI-COR Biosciences) was used for visualization.

For flow cytometry, cells were detached, washed and fixed with 2% PFA for 10 min at RT. For intracellular staining, cells were permeabilized with 0.2% saponin in PBS or 80% acetone for 10 min at RT. Staining with primary antibodies was performed for 30min at 4°C, followed by two washing steps with FACS buffer (1% FCS in PBS). If applicable, staining with secondary antibody was performed for 30 min at 4°C in the dark, also followed by two washing steps with FACS buffer ( Table 6 ). Measurements were performed using a MACSquant VYB flow cytometer (Miltenyi Biotech).

HEK293T cells were seeded in a 12-well format one day prior to transfection. Two plasmids that encoded eCFP- or eYFP-tagged proteins of interest were used for transfection ( Table 7 ). Cells were transfected with 1 µg of each plasmid using PEI as described above. Cells were harvested 24 h after transfection by detaching them with a pipet in 1 ml PBS. They were directly transferred to a 5 ml tube on ice. Cells were spun at 300 g for 5 min, supernatant was discarded, cell pellet was resuspended in 350 µl FACS buffer and cells were immediately measured for FRET signal using at a MACSquant VYB flow cytometer (Miltenyi Biotech).

HEK293T control and CD81KO (sorted for low CD81 expression) cells were transfected with plasmids encoding for fragments of p50 and p65 that were fused to parts of Kusabira green. 48h after transfection, cells were harvested as described above for flow cytometry-based FRET analysis and green fluorescence intensity was measured. Details are described in the CoralHue Fluo-chase Kit (MBL Int.corp.) manual.

Immunofluorescence experiments were performed in a 96-well format. Cells were fixed with 2% PFA for 10 min at RT before permeabilizing them with 80% acetone for 10 min at RT. Cells were then washed three times with PBS and blocked with 10% normal goat serum in PBS for 30 min at RT. After three washing steps, primary antibody (dilution 1:250) was applied for 1 h at RT, followed by three washing steps and secondary antibody incubation (dilution 1:250) for 1 h at RT. Then, nuclei were stained with DAPI (1:20,000, 10min, RT; Sigma Aldrich) or SiR-DNA dye (1 µM, 90 min, RT; Tebu-Bio). Three final washing steps were performed, and fluorescence was measured with a Cytation 3 plate imager (BioTek) or Incucyte plate imager (Sartorius).

Luciferase assay experiments were performed in a 96-well format and measured in a 96-well white opaque plate using a Cytation 3 plate reader (BioTek) or a TriStar2 S microplate reader (Bertold Technologies). In case of Jc1_R2A Renilla luciferase reporter constructs, cells were electroporated with Jc1_R2A viral RNA and seeded in a 96-well plate. After incubation, cells were washed and lysed as described elsewhere (Fischl and Bartenschlager, 2013). For NF-κB reporter activity, cells were transfected with several plasmids ( Table 8 ). This included a NF-κB reporter plasmid (pNF-κB(3x)-FLuc), a Gaussia luciferase encoding plasmid as transfection control (pCMV-Gluc), and different plasmids encoding inducer proteins of NF-κB and IFN signaling cascades (p_human_IKKβ_ca, p(N)FLAG-CMV2 MAVS, pEF-Bos-RIG-I 1.211-flag). In some experiments, different chemical inducers of given pathways were used [phorbol 12-myristate 13-acetate (PMA; 10-100 ng/ml), TNFα (10 ng/ml), Lipopolysaccharide (LPS; 100 ng/ml), Ionomycin (0.25 µM), PolyIC (5 µg/ml, transfected)]. Cells were additionally transfected with a control plasmid (pWPI_BLR) or a plasmid encoding CD81 (pWPI_hCD81-HAHA_BLR), and plasmids encoding eYFP or eYFP-core (pEYFP, pEYFP-HCV-core). 4 h after transfection, cells were treated with chemical inducers and luciferase activity was determined 24 h after transfection. For this, cell culture supernatant was transferred to a 96-well white opaque plate and mixed with coelenterazine (final conc. 5 µM) to determine transfection efficiency. Then, cells were washed with PBS and lysed with 60 µl FLuc lysis buffer (0.1 M KH₂PO₄/K₂HPO₄ (pH 7.8), 1% (v/v) Triton X-100, 1 mM DTT before use) for 10 min at RT. 40 µl lysate was transferred to a 96-well white opaque plate and mixed with 40 µl FLuc assay buffer (0.1 M KH₂PO₄/K₂HPO₄ (ph 7.8), 15 mM MgSO₄, 5 mM ATP) and 40 µl FLuc substrate buffer (0.28 mg/ml D-Luciferin in FLuc assay buffer). Firefly luciferase signal was measured immediately.

Live cell imaging was performed in a 96-well format using an Incucyte plate imager (Sartorius). Images were taken every 2-4 h in the respective channels. Imaging of fixed plates was performed using a Cytation 3 plate imager (BioTek) for cells stained for HCV core and with an Incucyte plate imager for cells stained for p65. Translocation of p65 in images was analyzed using the Incucyte analysis software. In brief, nuclear area was defined by red fluorescence above an arbitrary threshold over background. Then, integrated green fluorescence intensity within the nuclear area was calculated and normalized to the nuclear area in total.

Design and alignment of DNA plasmids was done using SerialCloner v2.6.1 (SerialBasics) unless stated differently. Western blot membranes were analyzed using ImageStudio lite (LI-COR biosciences). Flow cytometry data was analyzed using Flowlogic v8.3 (Inivai Technologies). Microscopy images were analyzed using SoftWoRx 7.0 (Cytiva), Gen5 v3.10 (BioTek Instruments), IncuCyte GUI v2021A (Sartorius) according to instrument and subsequently handled with ImageJ. Statistical analysis and creation of graphs was done using GraphPad Prism 9 (GraphPad Spftware LLC) and Excel 2019 (Microsoft corp.). Arrangement of figures was done using CorelDraw X7 (Corel Corporation).

Viruses manipulate the plasma membrane of infected cells for antiviral immune evasion and manipulation of signaling cascades. In order to study alterations of the plasma membrane proteome upon HCV expression, a flow cytometry-based surface expression screen was performed. In brief, Huh7.5 hepatoma cells were electroporated with viral genomic RNA, which encodes for a fluorescently labeled NS5A fusion protein (Jc1_NS5A-mtagBFP) to distinguish cells with active HCV replication (BFP+) from non-HCV expressing cells (BFP-). Cells were then stained with an arrayed set of antibodies in a 96-well format, and the surface expression of 332 proteins was measured and compared between non-HCV and HCV expressing cells, to calculate fold receptor modulation. The raw data of the three independent biological replicate screens is summarized in Supplementary Table 1 ( Table S1 ). Of surprise, receptor modulation was in general not very pronounced, with only seven proteins out of 332 whose levels were significantly lower upon HCV genome expression ( Figure 1A ). The two most down-regulated proteins, CD63 and CD81, which is one of the HCV entry receptors, were independently confirmed upon electroporation of Huh7.5 hepatoma cells with a viral genome RNA encoding for a fluorescently labeled NS5A fusion protein (Jc1_NS5A-eGFP) ( Figure 1B ). Notably, both are members of the family of tetraspanins. To get first insights on viral proteins involved in this phenotype, we transfected HEK293T cells to express single HCV proteins fused to eYFP and analyzed cell surface ( Figure 1C ) and total ( Supplementary Figure 1A ) CD63 and CD81 levels by flow cytometry. We employed E2, since it is known to directly interact with CD81, NS5A as multifunctional HCV accessory protein and the viral ion channel p7, that is incorporated into membranes. Both, cell surface and less pronounced total tetraspanin levels were differentially reduced by the viral proteins. However, efficient and dose dependent downregulation was observable only upon NS5A expression ( Figure 1C ). Therefore, HCV-mediated tetraspanin modulation seems to be exerted mainly by the action of NS5A.

To assess the functional role of tetraspanins for viral replication, an infectious HCV genomic RNA encoding a luciferase reporter was used [Jc1_R2A (Reiss et al., 2011)]. Huh7.5 hepatoma cells harboring gene knock-outs in CD63 and CD81 ( Supplementary Figure 1B ) were electroporated to express Jc1_R2A, and luciferase activity was measured as a proxy for viral genome replication and spread. Huh7.5 transduced with non-gRNA expressing lentivirus served as negative control (control cells, compare M&M). As expected, given the role of CD81 as HCV entry receptor, cells lacking CD81 showed a strongly decreased luciferase signal as compared to parental cells (WT), controls, or cells with CD63KO ( Figure 1D ).

As knock-out of CD63 did not have any impact on HCV replication ( Figure 1D ), we followed up on CD81 and analyzed in detail the kinetic of total cellular CD81 modulation in HCV-expressing cells. For this, we used an HCV reporter virus expressing the red fluorescent protein mScarlet, instead of the luciferase in front of the polyprotein (Jc1_mSc-2A), or Jc1-based constructs with internal E2 or NS5A mScarlet fusion proteins (Jc1_E2-mSc, and Jc1_NS5A-mSc, respectively). Monitoring mScarlet fluorescence allowed us to determine total CD81 levels in HCV replicating and bystander cells over time via intracellular CD81 staining and flow cytometry on a single cell level ( Figure 1E ; Supplementary Figure 2A ). While total CD81 levels were not altered 24h post EP, a decrease was observed at 48h which stayed similar at 72h. In conclusion, HCV reduces not only cell surface CD81, but lowers total CD81 levels.

We next investigated if HCV degrades CD81 by proteasomal or lysosomal degradation. We electroporated Huh7.5 cells with Jc1_NS5A-mScarlet and treated the cells with a proteasomal (MG132) or a lysosomal inhibitor (Bafilomycin A1), stained for total CD81 and analyzed CD81 modulation by flow cytometry. Treatment with none of the inhibitors rescued CD81 levels, indicating that CD81 is not degraded by the proteasome or the lysosome in HCV-expressing cells ( Figure 1F ). Instead, RT-qPCR revealed that the levels of CD81 mRNA were reduced by approximately half 48 h after electroporation of Huh7.5 cells ( Figure 1G ). Of note, we here used non-modified infectious HCV Jc1, demonstrating that CD81 is also modulated by non-tagged viral genomes. Taken together, HCV reduces total CD81 expression at the mRNA level in Huh7.5 cells.

CD81 is transcriptionally silenced and downregulated upon the onset of viral genome replication ( Figures 1E, G ). This suggests, that downregulation of CD81 plays additional roles in HCV biology beyond serving as entry receptor. A common strategy of viruses is to prevent superinfection by downmodulating its entry receptors and this has been described for HCV and CD81 (Tscherne et al., 2007), even though another study reported superinfection exclusion in the absence of CD81 reduction (Schaller et al., 2007). In order to analyze if reduction of CD81 by HCV confers additional benefits in the viral life cycle, we took advantage of HCV reporter genomes producing virions that are either severely (Jc1_NS5A-mScarlet) or completely (Jc1_E2-mScarlet) compromised in their ability to de novo infect cells, due to the expression of fluorescent fusion proteins of mScarlet with either NS5A or E2 (Merz et al., 2011; Bayer et al., 2016). This allows to monitor intracellular viral RNA transcription and protein translation in the absence of viral spread. Infectious Jc1_mScarlet-2A was used as a positive control.

Crispr-control and CD81KO Huh7.5 cells were electroporated with the aforementioned genomes and then 96-well plate-based fluorescence microscopy live cell imaging was conducted over a period of six days to monitor the efficiency of HCV protein expression and viral spread using mScarlet as a surrogate marker for viral replication. This revealed, as expected, that infectious Jc1_mScarlet-2A was not able to spread in CD81-negative Huh7.5 cells ( Figure 2A , left) whereas non-infectious Jc1_NS5A-mScarlet and Jc1_E2-mScarlet failed to induce spreading infection in both, control as well as CD81KO Huh7.5, confirming their loss of infectivity ( Figure 2A , middle and right panel). However, of note, when using the two viral genomes devoid of inducing a spreading infection, we consistently observed higher numbers of HCV-expressing cells in the CD81KO cells in comparison to the CD81-positive Huh7.5 controls ( Figure 2A , middle and right panel).

Moreover, this phenotype was recapitulated in CD81-negative Huh7 Lunet cells that were engineered to express hCD81 ( Figure 2B ). Similar to the conditions in which CD81 was knocked-out in Huh7.5, the number of HCV Jc1_E2-mScarlet replicating cells was high in CD81-negative parental Lunet cells and reduced upon reconstitution of hCD81. In addition, cellular proliferation and growth was reduced in Lunet cells upon expression of hCD81 ( Figure 2B ). In conclusion, CD81 negatively regulates intracellular HCV protein expression and cellular growth, explaining why HCV transcriptionally silences CD81 post entry. Altogether, as expected and considering its role as essential entry receptor, CD81 is clearly important for overall HCV infection and spread. However, at the cellular level and post entry, its absence seems beneficial for HCV genome replication and viral protein expression.

Among other functions, CD81 serves as a scaffold protein which is why we hypothesized that reduced CD81 levels might have an impact on intraviral protein interactions. We therefore used our established FRET assay (Banning et al., 2010; Hagen et al., 2014; Lim et al., 2022) to monitor HCV intraviral protein interactions and found that these were generally not altered in CD81KO cells ( Supplementary Figure 2B ). This indicates that higher intracellular HCV protein expression in CD81KO cells is due to other mechanisms.

Interferon (IFN) treatment triggers an antiviral state that renders cells largely resistant to HCV replication and viral protein expression. Nevertheless, HCV replicating cells can overcome early IFN-mediated antiviral immune response by expression of viral proteins that counteract induction of interferon stimulated genes (ISGs) (Lin et al., 2006). We hypothesized that CD81 might possibly alter IFN-signaling and thereby restrict intracellular HCV replication and protein expression. Therefore, Huh7.5 control and CD81KO cells were electroporated with Jc1_NS5A-mScarlet, stimulated with IFNα and stained for the cell surface expressed interferon-stimulated gene (ISG) tetherin (CD317), to analyze if Huh7.5 are responsive to IFNα despite the lack of PRRs and if HCV counteracts this response in our system ( Figure 2C ). Mock electroporated cells showed a clear induction of tetherin expression 24 h after IFNα stimulation. A similar effect was observed for Jc1_NS5A-mScarlet electroporated cells that were mScarlet-negative and did hence not express viral proteins (bystander cells, gated as mSc-, Figure 2C ). In contrast, cells that actively replicate HCV, as evident by NS5A-mScarlet expression, had tetherin levels comparable to cells that were not treated with IFNα, demonstrating that the IFN-mediated antiviral immune response is indeed suppressed in HCV-expressing cells ( Figure 2C ). In addition, CD81 did not impair the ability of HCV to counteract the IFN-response, as tetherin induction was suppressed in control cells to a similar extent as in CD81KO Huh7.5. A remarkable difference was baseline induction of tetherin upon IFNα treatment. CD81KO cells that were mock electroporated or the mScarlet-negative bystander cells, both expressed higher tetherin levels as control cells, indicating that the absence of CD81 could sensitize cells for ISG induction ( Figure 2C ; Supplementary Figure 3 ).

The interferon signaling cascade involves activation of TBK1, which has a central role in innate immunity and ISG induction. In addition, TBK1 is involved in pro-survival and anti-apoptotic signaling and can activate NF-κB as downstream target (Shin and Choi, 2019). As HCV blunts ISG induction ( Figure 2C ), higher pro-survival signaling of TBK1 in CD81KO cells could be beneficial for HCV-gene expression and cellular growth. To test for this, we electroporated Huh7.5 control and CD81KO cells with Jc1_NS5A-mScarlet and treated with a TBK1 inhibitor (BX795). Then, live cell imaging was conducted and cellular growth as well as HCV-expression based on mScarlet-positive cells was monitored. Inhibition of TBK1 led to decreased viral gene expression in both, Huh7.5 control and CD81KO cells ( Figure 2D , left). However, of note, HCV-expressing Huh7.5 CD81KO cells proliferated after being treated with BX795, whereas CD81-positive control cells showed impaired growth ( Figure 2D , right, purple line as compared to blue line). This suggests that the absence of CD81 can compensate for TBK1-mediated inhibition of proliferation of Huh7.5 hepatoma cells.

Pro-survival signaling of TBK1 is mediated via NF-κB (Shin and Choi, 2019) and CD81KO cells showed enhanced survival and growth upon TBK1 inhibition ( Figure 2D ). We therefore hypothesized that CD81 suppresses NF-κB signaling which might be one of the reasons why CD81KO cells are resilient towards TBK1 inhibition. Furthermore, NF-κB is activated via HCV Core inducing proliferation of hepatoma cells and possibly tumor formation (You et al., 1999; Yoshida et al., 2001; Nguyen et al., 2006; Sato et al., 2006; Selimovic et al., 2012; Simonin et al., 2013). To study effects of Core and CD81 on NF-κB we first used HEK293T transfected to express a NF-κB luciferase reporter plasmid together with a control or CD81 plasmid. We then induced NF-κB signaling by co-transfection of HCV YFP-core plasmid with or without TNFα treatment. Indeed, HCV Core alone as well as TNFα induced NF-κB reporter activity ~20-fold, while both together led to a ~60-fold induction ( Figure 3A ). Of note, transfecting cells to express CD81 reduced NF-κB reporter activity induced by HCV Core and TNFα nearly to background levels. This confirms that the HCV Core protein induces NF-κB signaling and this activity can be counteracted by high levels of CD81.

We next sought to more closely investigate effects of CD81 on NF-κB signaling induced by either IKKβ, MAVS or RIG-I. Activity of the NF-κB reporter was induced by all of them, with IKKβ being the most potent inducer and again upon expression of CD81 in this system, a clear reduction of NF-κB activity could be detected ( Figure 3B ). Subsequently, we decided to analyze cell type dependency by using HEK293T as well as HeLa cells, varied CD81 levels by employing cells with and without CD81KO that were additionally transfected to express CD81 or a control plasmid and used Phorbol 12-myristate 13-acetate (PMA) as an exogenous NF-κB trigger in addition to IKKβ ( Figures 3C, D ). In general, irrespective of the cell line or inducer used (IKKβ or PMA), increasing amounts of CD81 reduced NF-κB reporter activity, albeit the effect was only significant in HEK293T cells ( Figures 3C, D ).

To corroborate our results with more physiological triggers of NF-κB signaling, without transfecting IKKβ, we more closely investigated CD81KO HEK293T and HeLa cells upon induction of NF-κB via PMA ( Figure 4 ). PMA mimics the second messenger lipid diacylglycerol (DAG) which activates protein kinase C (PKC). Some PKC isoforms require the second messenger Ca²⁺, to become fully activated (Reyland, 2009). Hence, cells transfected with the NF-κB reporter were additionally treated with Ionomycin to increase cytosolic Ca²⁺ levels. Ionomycin treatment did not increase NF-κB activity and did not induce NF-κB activity when administered alone ( Figures 4A, B ). However, in line with our previous results, NF-κB reporter activity in HEK293T ( Figure 4A ) and HeLa cells ( Figure 4B ) was strongly suppressed by CD81. Importantly, this phenotype was fully recapitulated in the absence of transfecting any overexpression plasmid in CD81KO cells, as we also observed suppression of PMA-induced NF-κB activity by endogenous CD81 ( Figure 4C ). Furthermore, when using a complete independent readout for NF-κB signaling and in the absence of any plasmid transfection upon KO of endogenous CD81 in HEK293T cells, we also found that levels of TNFα mRNA were increased at different time points post PMA stimulation, when cells were depleted for CD81 ( Figure 4D ). In conclusion, endogenous as well as overexpressed CD81 can suppress NF-κB activity induced by PMA of IKKβ in different cell lines.

To explore potential mechanisms of increased NF-κB signaling, we assessed the interaction of the NF-κB signaling components p50 and p65 via a split Kasubira-green assay and found that the interaction was increased in CD81KO vs control HEK293T cells ( Figure 5A ). Furthermore, moving to hepatoma cells, we found overall increased basal levels of p65 and phosphorylated p65, even though independently of TNFα ( Figure 5B ). To unambiguously directly analyze endogenous NF-κB signaling in Huh7.5 cells, nuclear translocation of p65, the subunit of the NF-κB transcription factor family, was followed upon treatment with PMA or TNFα ( Figure 5C ). PMA treatment induced p65 translocation after approximately 30 min which was strongly increased in CD81KO hepatoma cells ( Figure 5C ). Quantification of the nuclear intensity of the p65 staining revealed that while it reached a plateau in control cells at 30 min after treatment, it further increased in CD81KO Huh7.5 from 60 min post treatment on ( Figure 5D ). A similar trend was observed for TNFα-treated samples ( Figures 5C, D ). In conclusion, the cumulated results employing various NF-κB inducers, cell lines, overexpression and knockout conditions as well as different NF-κB readouts based on reporters and endogenous signals suggests a suppression of NF-κB activation by CD81.