In line with the literature (42) and our sequencing data ( Supplementary Figure 1A ), the fluorometric analysis of intracellular Ca²⁺ influx failed to demonstrate activation of the plasma membrane-located P2RX7 in Cloudman (CM) mock-or MVA-PK1L-Ova infected cells upon Bz-ATP specific stimulus ( Figure 1A ). Although recent New Generation Sequencing analysis (14) has shown that the expression of the P2X7 receptor was upregulated after MVA infection, qRT-PCR analysis of infected CM cells failed to demonstrate an upregulation of expression but indicated a stable constitutive expression after infection ( Supplementary Figure 1B ). For the matter of simplification, we distinguish between active and inactive P2RX7, referring to the BALB/c P2RX7 or DBA/C57BL/6 P2RX7 (less sensitive to ATP stimuli (42)), respectively. To exclude that the lack of intracellular Ca²⁺ increase was due to faulty loading of the fluorescent indicator FURA-2-AM, we additionally stimulated the cells with ionomycin, which is a receptor-independent trigger for a maximal increase of intracellular Ca²⁺. Even higher Bz-ATP stimuli could not increase intracellular Ca²⁺ concentrations ( Supplementary Figure 1C ). Interestingly, when stimulating the CM cells with ATP, known to activate additional purinergic receptors besides P2RX7, we observed a gain in intracellular Ca²⁺ amounts, suggesting the activity of other receptors of this or the P2Y receptor family (43) ( Supplementary Figure 1D ). Next, we were interested in studying the role of active P2RX7 and transfected our CM cells with the fully functional P2RX7 expressed by BALB/c mice (42). We were able to demonstrate its activity after transfection, by an increase in the concentration of intracellular Ca²⁺ upon Bz-ATP stimulus ( Figure 1B ). This response was reversed when the reconstituted cells were treated with the P2RX7-specific competitive inhibitor A740003 at 20µM. Transfection with the empty vector control, similar to CM WT cells, did not alter the amount of intracellular Ca²⁺ upon Bz-ATP treatment. Toxicity of A740003 was excluded ( Supplementary Figure 1E ). Furthermore, the receptor activity represented by an increased Ca²⁺ influx in the P2RX7-transfected cells appeared to be significantly higher at earlier time points during MVA infection after 4 h.p.i. ( Supplementary Figure 1F ), suggesting that the infection itself modulates the activity of the receptor. However, calcium levels did not change after 20h MVA infection when compared to uninfected cells indicating that the receptor activity is only transiently increased after infection ( Figure 1C ). Interestingly, we did not observe P2RX7-specific pore function in P2RX7 transfected cells ( Supplementary Figure 1G ) (44). The release of extracellular particles, which is reported to be partly P2RX7 dependent (7), was enhanced when P2RX7 transfected cells were infected with MVA, supporting the regulation of P2RX7-specific functions during MVA infection ( Figures 1D, E ). Overall, we confirmed the reconstitution and functionality of P2RX7 in CM cells after transfection.

Recent studies have shown that innate triggers derived from infected feeder cells are relevant for the activation of T cells by antigen-presenting cells (33). We were interested in investigating whether the presence of a functional P2X7 receptor in feeder cells may have an impact on the antigen uptake and presentation capacity of bone marrow-derived dendritic cells (BMDCs) for activation of CD8⁺ T cells. We demonstrate that using MVA-OVA-infected feeder cells bearing the active P2X7 receptor led to a significantly higher CD8⁺ T cell activation as determined by IFNү production in B8R- specific T cells or by TNFα production in either B8R- or OVA-specific T cell lines when co-cultured with uninfected dendritic cells as cross-presenting APC ( Figure 2A ). These data are in line with the P2RX7-dependent release of these cytokines in mice (45). The frequency of cross-presenting BMDCs with SIINFEKL/H2-K^b complexes on the cell surface, as well as the amount of these peptide/MHCI complexes per cell, was significantly increased ( Figure 2B left, middle). Interestingly, the presence of the active P2RX7 in CM cells led to the increase of MHCII surface expression in co-cultured BMDCs ( Figure 2B right). The expression of other maturation markers, such as CD40 or CD86 on co-cultured BMDCs was not affected (data not shown). Furthermore, pre-treatment of CM P2RX7 cells with A740003 before co-culture with BMDCs led to a significantly reduced CD8⁺ T cell activation ( Supplementary Figure 2A ) and SIINFEKL/H2-K^b expression ( Supplementary Figure 2B ) which was comparable to CM WT or CM pcDNA3 cells. To corroborate the specific function of P2RX7, we used HEK293 as feeder cells expressing active human P2RX7 ( Figure 2D ). Indeed, the co-incubation of BMDCs with MVA-infected HEK293 hP2RX7 feeder cells resulted in increased SIINFEKL/H2-K^b expression ( Figure 2C ). In sum, we demonstrated that the presence of functional P2RX7 in feeder cells aids in improving antigen cross-presentation upon MVA infection.

We first hypothesized that the increased SIINFEKL/H2-K^b surface expression on antigen-presenting cells and the improved CD8⁺ T cell activation that we found when BMDCs were co-cultured with P2RX7 feeder cells, might be due to an increased amount of viral antigens. We analyzed the expression of early antigens, such as B8R, a native MVA antigen, or Ova, expressed under the control of the early MVA promoter PK1L, or the late viral antigen A19L in CM P2RX7 or CM pcDNA3 transfected cells and compared it to CM WT cells. Expression of these antigens was initially lower in CM P2RX7 cells compared to CM WT or the empty vector control CM pcDNA3. However, at 24hpi B8R, Ova and A19L mRNA fold change in CM P2RX7 cells was significantly higher when compared to CM WT or CM pcDNA3 cells ( Figure 3A ). We concluded that the expression kinetics of viral genes in CM P2RX7 was delayed, although mRNA expression at later time points in these cells was significantly higher. This was further confirmed when analyzing the replication capacity of MVA in the different CM feeder cell lines. We also observed a higher residual viral titer at 0hpi in CM P2RX7 ( Figure 3B ), suggesting a possible role of the P2X7 receptor for viral entry, as previously stated (12). Since MVA has lost its ability to replicate in most mammalian cells (19), we wanted to exclude the possibility that the presence of this receptor was affecting viral replication behavior and, as a consequence, promote antigen cross-presentation. However, at 24hpi viral particle amounts in CM P2RX7 cells were comparable to CM WT or CM pcDNA3 cells ( Figure 3B right). Similarly, the synthesis of OVA protein in CM P2RX7 cells was initially reduced at 8hpi but comparable at 20hpi ( Figure 3C ), indicating that expression of viral genes and synthesis of corresponding proteins might be delayed in CM P2RX7 transfected cells.

The above data exclude that altered protein amounts in feeder cells as the source of antigenic uptake by antigen-presenting cells contributed to the improved antigen presentation by BMDCs co-cultured with infected CM P2RX7 cells. It has been postulated that the antigen has to be released but needs to be still cell-associated for the antigen-presenting cells to phagocytose and process it for cross-presentation to CD8⁺ T cells (46–48). We confirmed the activation of the extrinsic apoptosis pathway by cleavage of Caspase 8 ( Figure 3C ). The exposure of phosphatidylserine residues on the cell surface is important for the activation of phagocytosis by APCs and has been linked to P2RX7 activation (49–51). We demonstrate that the presence of the functional P2X7 receptor led to the increased surface expression of phosphatidylserine residues in cells having a permeable cell membrane ( Figure 3D right) when mock-infected or infected with MVA for 6h. In contrast, MVA-infected CM P2RX7 with intact cell membrane, hence alive cells, displayed reduced phosphatidylserine levels as compared to CM WT or CM pcDNA3 cells after 20h MVA infection. Overall, we speculate that expression of the functional P2X7 receptor in CM feeder cells does not modulate the antigen-presentation capacity by APCs by increasing the total amount of antigenic protein in feeder cells, but rather allows for increased viral mRNA levels in infected cells at later time points as well as by altering apoptotic pathways. This depicts the importance of further analyses on the RNA level and the possible altered localization of P2RX7 affecting other cellular pathways.

Extracellular particles are known to contain crucial regulatory molecules for cell-to-cell communication (11). Similarly, supernatant from P2RX7 transfected cells differs from control cells (37). Here we hypothesized that both the extracellular particle fraction (EP-fraction), as well as the supernatant fraction (sup-fraction) from CM P2RX7 cells, are responsible for the enhanced antigen cross-presentation we observed. We first isolated the EP-fraction and the supernatant fraction as depicted in Figure 4A and determined both, mRNA and protein content. OVA protein amounts were comparable in wildtype and transfected CM cells after overnight infection in both, the EP- and the sup-fraction ( Figure 4B ). Interestingly, mRNA levels of the MVA early antigen B8R, but not Ova or A19L expression, were significantly higher in the EP-fraction at 0 hours post-infection leading to the hypothesis that the lower mRNA expression we observed in the cell extracts ( Figure 3A ) might be due to the release of RNA in extracellular particles ( Figure 4D ). This finding highlights the importance of focusing on the role of RNA in the regulation of MVA antigen presentation. We also showed that CM P2RX7 transfected cells allowed the secretion of significantly higher amounts of inflammatory cytokines pre- as well as post-infection ( Figure 4C ), demonstrating that the presence of functional P2RX7 modulates other signaling pathways as well. A detailed overview of the secreted cytokines can be seen in Supplementary Figure 3A . To further investigate the function of these culture sub-fractions from P2RX7 transfected cells, we delivered EP- and sup-fractions to the co-culture of uninfected CM cells and BMDCs to monitor subsequent CD8⁺ T cell activation and expression of SIINFEKL/H2-K^b complexes on dendritic cells. The EP-fraction from CM P2RX7 was not able to induce stronger CD8⁺ T cell IFNү or TNFα production as compared to the EP-fraction derived from pcDNA3 transfected CM cells. However, the sup-fraction led to increased IFNү and TNFα production when added to the co-culture of CM and BMDCs ( Figure 4F ). Interestingly, both fractions from CM P2RX7 cells were able to induce a significantly higher SIINFEKL/H2-K^b expression as compared to the empty vector control-derived fractions ( Figure 4G ). To establish whether these fractions could initiate similar CD8⁺ T cell activation as the infection of CM P2RX7 cells, these were added to MVA-infected CM WT cells ( Supplementary Figures 3B–E ). Even though both fractions seem to contribute to the increased expression of SIINFEKL/H2-K^b on BMDCs, the addition of EPs or supernatants from infected P2RX7 CM cells could not further increase CD8⁺ T cell activation or SIINFEKL/H2-K^b expression on BMDCs significantly. In addition, we filtered supernatants from infected feeder cells to remove any cell components larger than 0.2µM, such as apoptotic bodies (but leaving exosomes and microvesicles in the fraction) and added this filtered supernatant (fil sup) to either uninfected CM or MVA-infected cells ( Figure 4H left, middle). We found a significant contribution of the filtered supernatant from CM P2RX7 cells for the activation of dendritic cells and subsequent presentation of the SIINFEKL peptide on its MHCI. The total amount of cross-presenting BMDCs (frequency) with SIINFEKL/H2-Kb complexes at the cell surface as well as the amount of SIINFEKL/H2-Kb complexes per BMDC (MFI) was significantly increased ( Figure 4H left, middle). Notably, MHC I expression (total H2-K^b) of the antigen-presenting cells was not altered upon the addition of this filtered supernatant ( Figure 4H right). To sum this up, we demonstrate that the cellular fraction of P2RX7 transfected cells as well as other subcellular fractions are critically involved in enhancing antigen cross-presentation.

As a plasma membrane receptor, P2RX7 is associated with the activation of the canonical inflammasome pathway (52). Recent studies, however, show that P2RX7 can also localize to and function intracellularly on mitochondrial structures (53). We confirmed the presence of intracellular P2X7 receptors in our feeder cells. These were significantly increased in P2RX7 transfected CM cells ( Figure 5A ). Mitochondria and energy metabolism are impacted during viral infection and since the presence of P2RX7 on mitochondrial surfaces of HEK293 hP2RX7 and N13 microglial cells has been shown recently by others (53, 54), we determined the mitochondrial activity in our feeder cells. As expected, maximal respiration and spare capacity were significantly increased in CM cells bearing the active P2X7 receptor as compared to CM WT or empty vector-transfected cells. Basal respiration was only increased in CM P2RX7 cells when compared to CM pcDNA3 cells. Also after infection, the maximal respiration and spare capacity were significantly upregulated in CM P2RX7 cells ( Figure 5B ). In addition, the extracellular acidification rate (ECAR), an indicator for glycolysis processes (55), was significantly higher in both, mock-infected and MVA-infected CM P2RX7 cells, as compared to CM WT or CM pcDNA3 cells ( Figure 5C ). Interestingly, intracellular ATP concentrations in the uninfected CM P2RX7 cells were comparable to CM WT cells but slightly lower than in CM pcDNA3 cells. After infection, however, intracellular ATP was significantly increased in CM P2RX7 cells only ( Figure 5D ), while extracellular ATP inversely correlated after infection showing a significant decrease for CM P2RX7 cells only ( Figure 5E ). In summary, we report that presence of P2RX7 significantly alters the energy metabolism of the cells by increasing the availability of mitochondrial ATP after MVA infection.

Since the ECAR measured in this assay is only a quantitative measurement of the total amount of acid (H⁺) produced during both the tricarboxylic acid cycle and glycolysis, further assays may help to determine more specific alterations in the glycolytic metabolism (55). This study did not include an in-depth analysis of different extracellular particle fractions such as apoptotic bodies, microvesicles and exosomes. Further fractionation and subsequent characterization will allow to determine the exact content of these particles. Further work should address the role of DBA P2RX7 in CM cells. Even though the plasma membrane receptor is not functional in cells with a DBA background due to the P451L mutation (42), further studies may try to characterize other functions of this receptor during MVA infection in vitro and in vivo. Up to now, animal models such as appropriate P2RX7 knockout mice are lacking to investigate the role of the receptor in vivo. Since ATP may play a role during the co-culture of immune cells (85), future studies should assess its role in feeder cells, APCs and during co-culture. In addition, understanding how the altered mitochondrial metabolism affects EP release and composition as well as cross-presentation on the molecular level would be important for future studies. Due to the limitations of our in vitro murine model for the cross-presentation of MVA antigens, future research should include human cells.

For isolation of bone marrow female 12-to 16-week adult C57BL/6N mice were purchased from Janvier and were allowed to acclimate for a minimum of one week in the in-house animal facility. For weekly T cell stimulation, the spleen of adult C57BL/6N mice was used. Animals were maintained at the Zentrale Einrichtung für Tierversuchsanstalt (ZETT) at the University of Düsseldorf under specific pathogen-free conditions. Experimental procedures have been approved by the regional authorities (North Rhine-Westphalia State Environment Agency - LUA NRW, Germany) and the animal use committee at the University of Düsseldorf (Reg. No O119/11).

Recombinant MVA were generated by homologous recombination as previously described (33, 86). All stock preparations of MVA used in this study were diluted to a concentration of 1x10⁹ viral particles/mL and maintained at -80°C. Viral aliquots were thawed in a water bath, sonicated for one minute, briefly vortexed and spun down for usage. Freeze/thawed aliquots were not used more than three times.

Unless differently stated, cells were harvested, pelleted in a falcon and infected with MVA (MOI1, unless otherwise specified) for one to two hours at 37°C, 5% CO₂ with intermittent shaking every 15min. Cells were washed twice before incubation with other cells for cross-presentation experiments. For the remaining experiments cells were seeded and incubated immediately. Since harvesting at different time points was required for expression kinetics analyses and titration experiments, cells were seeded and allowed to adhere before infection. Infection was then performed directly on the plate with intermittent shaking and washing after one hour of incubation at 37°C and 5% CO₂. For infection in 96-well plates (Mitochondrial function assay and intracellular ATP determination assay) virus was added to each well of the plate, shaken every 15min for 1h and subsequently incubated for the remaining time frame at 37°C and 5% CO_2.

1x10⁶ cells were either mock- or MVA (MVA-PK1L-OVA) infected at an MOI of 1 for either 4h or 20h. For Ca²⁺ measurements infected cells were placed in a falcon for loading with 4µM FURA-2 AM (Sigma-Aldrich) in saline solution (12.5mM NaCl, 0.5mM KCl, 0.1mM MgSO₄, 2mM HEPES, 0.55mM D-glucose, 0.5mM NaHCO₃ (all Sigma-Aldrich)) supplemented with 0.5mM CaCl₂ (pH 7.4, Merck) and 250µM sulfinpyrazone (Sigma-Aldrich) at 37°C for 20min. Cells were then washed, resuspended in saline solution and stimulated with the indicated concentrations of Bz-ATP (Sigma-Aldrich) and 1µM ionomycin (Invitrogen) for the recording of intracellular Ca²⁺ release. For detection of pore opening at the cell membrane, cells were loaded with 2µL ethidium bromide (EtBr (Sigma-Aldrich)) and stimulated with 200µM Bz-ATP and 100µM digitonin (Sigma-Aldrich). Measurements were done in a thermostat quartz cuvette using a Perkin-Elmer KS50 rotating and heating system at a wavelength of 340/380nm (excitation) and 505nm (emission) for intracellular Ca²⁺ release and at a wavelength of 360nm (excitation) and 580nm (emission) for EtBr- pore opening assay.

Cloudman S91 cells (ATCC CCL-53.1) were seeded at a density of 1.5x10⁵ cells per well in a 6-well plate and transfected with either 3µG pcDNA3 control or P2RX7 encoding plasmid DNA using Lipofectamine reagent (Invitrogen) according to the manufacturer’s instructions. Briefly, plasmid DNA was dissolved in medium, incubated with 1µL PlusReagent for 5min and then 3µL Lipofectamine was added and further incubated for 30min at room temperature. The Lipofectamine-DNA mixture was then added dropwise to the cells and cells were selected for geneticin (0.2mg/mL) resistance two days post-transfection. Transfection efficacy was confirmed by Western Blot analysis of P2RX7 synthesis and by fluorometric analysis measuring the P2RX7-dependent intracellular Ca²⁺ increase.

Cells were grown on a round cover dish placed in a 6-well plate at a density of 5x10⁵ cells per well. Cells were stained with 2µM PKH-26 (Sigma-Aldrich) and 2µM Quinacrine (Sigma-Aldrich) for 10min at 37°C and 5% CO₂ in saline saccharose solution (30mM saccharose, 0.1mM K₂HPO₄, 0.1mM MgSO₄, 0.5mM D-glucose, 0.2mM HEPES (all Sigma-Aldrich)) supplemented with CaCl₂ (pH7.4). Cells were placed in a holder device for round cover glasses and stimulated with 200µM Bz-ATP. Images were acquired in 6-second intervals for approximately 15min. Images were taken at 60x magnification of the Olympus Fluoview FV3000 (Olympus). Data visualization was achieved using OMERO software (Open microscopy imaging).

Bone marrow was obtained from 12-to 16-week-old C57BL/6N and 5x10⁶ bone marrow cells were seeded with 10% GM-CSF (obtained from supernatant of B16 cells expressing GM-CSF, originally kindly provided by Georg Häcker, Freiburg) in RPMI-medium (Gibco) containing 10% heat-inactivated FBS and 50µM 2-mercaptoethanol (M2 Medium) in 10cm Petri-dishes. On day three fresh M2 Medium and GM-CSF was added to the primary culture and on day six 10mL medium was replaced with fresh M2 Medium containing GM-CSF. BMDCs were used on day seven for all experiments.

CD8⁺T cell lines were generated as described recently (33). For weekly T cell stimulation, both EL4 cells (ATCC TIB-39) and naïve splenocytes from C57BL/6N were irradiated with 100Gy or 30Gy, respectively. EL4 cells were loaded with 1µg/mL B8R-peptide (TSYKFESV; immunodominant peptide derived from the B8 protein from vaccinia virus) or Ova-peptide (SIINFEKL; derived from ovalbumin) and then co-incubated with splenocytes, CD8⁺ specific T cells and M2 Medium containing 5% TCGF (T-cell growth factor). Both peptides are H2-K^b-restricted.

Cloudman S91 murine melanoma (CM) cells (MHC I haplotype H2-d) were used as feeder cells for antigen cross-presentation assays. A total of 2x10⁶ cells were either mock- or MVA-PK1L-OVA (MOI1) infected for 20h, washed and subsequently incubated with psoralen (1µg/mL) (Sigma-Aldrich) for 15min at 37°C and 5% CO₂ and treated with UV-A light (PUVA) for further 15min. Cells were harvested, transferred to a falcon and washed with medium. CM feeder cells were co-incubated with uninfected BMDCs, which were previously generated from bone marrow of C57BL/6N mice (MHC I haplotype H2-K^b) at a ratio of 1:1 in a 6 cm dish for 18h. The next day, the co-culture of CM and BMDCs was harvested, washed in M2 Medium and resuspended in M2 medium in a final volume of 1mL. One part of the co-culture suspension (200µL) was immediately stained for the surface expression of peptide/MHCI complexes (SIINFEKL peptide within H2-K^b) on BMDCs, while 100µL of the CM-BMDCs co-culture (containing 2x10⁵ BMDCs as antigen-presenting cells) was further incubated with 2x10⁵ B8R- or Ova- specific CD8⁺ T cells in the presence of 1µg/mL Brefeldin A (Sigma Aldrich) for 4h at 37°C and 5% CO₂. Further analysis of CD8⁺ T-cell activation is described below.

To determine the antigen presentation capacity of dendritic cells in the cross-presentation setting, peptide-specific T cell lines were used as a read out system. After 4h incubation (see above cross-presentation assay), cells were washed with PBS and dead cells were excluded by staining with Fixable viability dye eFluor 506 (Invitrogen) (1:600) for 20min on ice. Cells were washed with FACS buffer (PBS supplemented with 1% BSA and 0.02% sodium azide) and then stained using anti-mouse CD8α eFluor 450 (eBioscience) (1:300) for 20min on ice. Subsequently, cells were permeabilized with BD Cytofix (BD Biosciences) for 15min on ice and then stained with Anti-mouse IFNy APC (Invitrogen) (1:400) and Anti-mouse TNFα PE-Cyanine7 (Invitrogen) (1:300) in 1:10 diluted BD Perm/Wash for 30min on ice. Cells were washed twice and resuspended in 1% PFA for subsequent analysis using the FACS Canto II device (BD Biosciences).

Antigen processing and presentation capacity was also assessed by measuring the MHCI/peptide complex formation as SIINFEKL/H2-K^b expression on the surface of the dendritic cells. For this co-cultured BMDCs (see above cross-presentation assay) were washed with PBS and dead cells were stained with Fixable viability dye eFluor 660 (Invitrogen) (1:2000) for 20min on ice. Fc-receptors were blocked using anti-mouse CD16/CD32 (eBioscience) (1:200). After Fc-blocking, surface staining was performed for 30min on ice using anti-mouse CD11c PE (Invitrogen), anti-mouse H2-Kb FITC (Biolegend) and anti-mouse SIINFEKL/H2-K^b PE-Cyanine 7 (eBioscience) (all 1:300 in FACS buffer). Cells were washed twice and resuspended in 1% PFA for subsequent analysis using FACS Canto II. Alternatively, cells were stained with Fixable viability dye eFluor 506 (Invitrogen) (1:600) for 20min on ice, followed by the Fc-blocking step and surface staining with CD11c APC-Cyanine 7 (BD Pharmingen) and MHCII PE (all 1:300 in FACS buffer).

For kinetic analysis of viral gene expression (0h to 24hpi), 2x10⁶ cells were infected with MVA-PK1L-Ova (MOI1) for 1h at 4°C, resulting in the virus attachment to the cell surface. After washing cells were harvested at the indicated time points, spun down and the pellet was resuspended for total RNA isolation as described in the manufacturer´s protocol (RNeasy Mini Kit (Qiagen)). Briefly, cells were lysed using RLT buffer containing 10µL 2-ß-mercaptoethanol and mixed with one volume of 70% ethanol for subsequent isolation using the RNeasy Mini spin column. cDNA was then transcribed using the Revert Aid H minus first strand cDNA synthesis (Thermo Fisher Scientific) according to the manufacturer’s instructions and used as a template for subsequent quantitative PCR reaction with PowerUp SYBR Green Master Mix (Applied Biosciences). Expression of viral B8R, Ova, A19L and cellular P2rx7 genes was normalized to expression of 18S-rRNA housekeeping gene and ΔΔCT was calculated by further comparison of ΔCT values with the 0h time point of CM wildtype cells. Primer sequences are listed in Supplementary Table 1 .

In order to determine the replication capacity of MVA, 1x10⁶ CM cells (WT, pcDNA3- or P2RX7- transfected) were infected with MVA-p7.5-GFP at MOI5 for 0h (1h at RT), washed and further incubated until harvested at 4h, 8h or 24hpi. Collected samples were vortexed and subjected to three rounds of freeze-thaw-sonication cycles to release viral particles. Viral suspensions were then used to prepare serial dilutions that were plated on 96-well plates containing MVA-permissive DF-1 cells (ATCC CRL-12203) (80% confluent). Fluorescent signal and cytopathic effect was monitored for seven days post-infection to determine the 50% endpoint titer of viral particles per milliliter by using the Spearman-Karber method to calculate the tissue culture infectious dose 50 (TCID₅₀).

For extraction of proteins, 2x10⁶ cells were infected (see above “infection of cells”) and harvested at the indicated time points. Collected cells were spun down by centrifugation, washed with PBS and resuspended in RIPA buffer (Thermo Fisher Scientific) containing HALT Protease & Phosphatase Inhibitor cocktail (Thermo Fisher Scientific) (1:100). After three rounds of freeze-thaw-sonication cycles, supernatants were harvested after a single centrifugation step at full speed for five minutes at 4°C. Protein content was quantified using the Pierce BCA Protein Assay Kit (Thermo Fisher Scientific). SDS-PAGE and blotting on nitrocellulose membranes was performed as described elsewhere (87). Membranes were incubated with Anti-Ovalbumin (Rockland) (1:20 000); Anti-Cleaved caspase-8 (Cell signaling) (1:1000); Anti-P2RX7 (Sigma-Aldrich) (1:200) and Anti-ß- Actin (Sigma-Aldrich) (1:50 000). Relative quantification of specific proteins was done by calculating ratio of the protein of interest with the ß-Actin loading control using the ImageJ analysis tool (US National Institutes of Health, Bethesda, USA).

Surface staining of phosphatidylserine residues on MVA-infected cells was done according to Apotracker-Green protocol (Biolegend) at either 6hpi or 20hpi. Briefly, 2x10⁵ cells were washed with FACS buffer and incubated in 400nM Apotracker-Green staining solution for 20min at room temperature. Cells were subsequently stained with fixable viability dye eFluor 660 (1:2000) for 20min on ice, washed and immediately analyzed by FACS. Cells were either gated for APC-negative (non-permissive for viability dye) and FITC-positive (Apotracker Green-positive) populations, designating early apoptotic cells or gated for APC- and FITC-double-positive populations, indicating late apoptotic cells.

For extracellular particle isolation the protocol was adapted according to Pegoraro and colleagues (8). Four T75 flasks (approximately 8x10⁶ cells/flask) were seeded with CM cells one day before infection to obtain 90% confluency. Before infection, one flask per cell line was counted in order to calculate the respective MOI. Cells were allowed to rest for 30min at room temperature and after washing 3mL medium was added in each flask. MVA-PK1L-Ova (MOI 1.5) was added and flasks were shaken every 15min for one hour (at 4°C for RNA isolation). After one hour cells were washed and harvested (0h value) or further incubated for a total of 20h at 37°C and 5% CO₂. For harvesting, medium was discarded, cells were washed with PBS and 3mL saline solution supplemented with 0.05mM CaCl₂ was added. Cells were stimulated with 200µM Bz-ATP for 30min at 37°C. Thereafter, the supernatant was aspirated and centrifuged at 300g for five minutes at 4°C to remove cell debris. The cleared supernatant was harvested, aliquoted in Eppendorf tubes and centrifuged at 20.000 g for one hour at 4°C. The supernatant was discarded and the remaining extracellular particle fraction (EP-fraction) was either used for quantitative RNA analyses (resuspended in RLT buffer with 2-ß-mercaptoethanol), western blot analyses (resuspended in RIPA buffer with HALT Protease & Phosphatase Inhibitor cocktail) or for cross-presentation assays (resuspended in PBS). For cross-presentation assays, EP fractions were additionally PUVA treated prior to the last centrifugation step, as described above.

For isolation of supernatants, 2x10⁶ cells per condition tested were used. Cells were either MVA-PK1L-Ova (MOI1) or mock-infected for the indicated time (8h or 20h for western blot analysis; 20h for Legendplex and cross-presentation assays), harvested and supernatants (sup-fraction) were collected after centrifugation at 300g for five minutes. For indicated experiments, supernatant fractions were further passed through a 0.2µM size pore filter (fil sup-fraction) to be used for cross-presentation experiments. All supernatant fractions (sup- or fil sup-fractions) used for cross-presentation assays were additionally PUVA treated as described above.

The release of cytokines/chemokines was analyzed using the Legendplex MU anti-virus response panel (Biolegend). Briefly, 2x10⁶ cells were either MVA-PK1L-Ova (MOI1) or mock-infected for 20h. After harvesting the cell suspensions, supernatants were collected after centrifugation at 300g for 5min and processed according to the manufacturer’s instructions. Data was analyzed using the Biolegend LEGENDplex Data Analysis Software (Biolegend).

Infection of CM feeder cells was performed as described above for cross-presentation assays. On day two, CM WT feeder cells (either mock- or MVA-PK1L-OVA infected, MOI1) were co-incubated with BMDCs and, additionally, pulsed with either EP-, sup- or fil sup-fractions. These fractions were isolated from either infected CM pcDNA3 (transfected cells with inactive P2RX7) or infected CM P2RX7 cells (transfected cells with active P2RX7) after 20hpi as described above. On day 3, cross-presentation assays were continued as described above.

To assess the expression of P2RX7, 2x10⁵ CM cells (WT, pcDNA3- or P2RX7-transfected) were mock- or MVA-PK1L-OVA (MOI1) infected. After 20hpi, cells were stained with fixable viability dye eFluor 660 (1:2000) for 20min on ice, permeabilized with BD Cytofix for 15min on ice and then stained with anti-P2RX7 (1:200) for one hour on ice to quantify the intracellular presence of P2RX7. Cells were further incubated with anti-mouse-IgG-PE (Jackson laboratories) (1:200) secondary antibody for 30min on ice, washed and immediately used for FACS analysis by FACS Canto II.

The day prior to infection, 2x10⁴ cells were seeded in a Seahorse XF96 Cell culture Microplate (Agilent Technologies). Cells were allowed to adhere for 1h at room temperature and were further incubated at 37°C at 5% CO₂ overnight. The next day cells were infected with MVA-PK1L-Ova (MOI5, 6h). Mitochondrial function was assessed using the Seahorse XF Cell Mito Stress test (Agilent technologies) according to the manufacturer’s instructions. Compounds have been used at the concentration of 15µM for Oligomycin, 5µM for FCCP and 5µM for Rot/AA.

The day prior to infection, 5x10⁴ cells were seeded in a 96 flat well chimney base plate and incubated overnight at 37°C and 5% CO₂. Cells were infected with MVA-PK1L-Ova for 6h (MOI5) before intracellular ATP concentrations were determined using the Luminescent ATP detection assay kit (Abcam) as described in the supplier´s protocol. Briefly, cells were lysed and ATP was stabilized by a detergent during a shaking step. After the addition of the substrate solution, prompted luminescence was measured and compared to ATP standard samples using a Spark plate reader (Tecan).

1x10⁶ CM cells (WT, pcDNA3- or P2RX7-transfected either MVA-PK1L-Ova or mock-infected, MOI5) were seeded in 1mL in a 6-well plate. The supernatant was harvested after 6hpi. For each condition, 50µL supernatant was incubated with 50µL of FirezymeB Diluent buffer (Firezyme). Samples as well as an ATP standard (Sigma- Aldrich) were compared in a standard curve at serial dilutions run by the Luminometer Victor 3 1420 Multiwell counter (Perkin Elmer) with automated addition of 100µL Enliten Luciferase/Luciferin reagent (Promega) to detect emitted luminescence.

Details on statistical analyses are integrated in figure legends. When indicated, data was normalized to untreated or WT control cells. Unpaired two-tailed student’s t-test was used to calculate statistical significances using Prism 8 (GraphPad Software). Extracellular particles ( Figure 1C ) were quantified by counting three adjacent frames of each replicate upon Bz-ATP stimulus and normalized to cell numbers (determined by quinacrine staining) per frame. Graphical data represent mean values with error bars indicating SD or SEM with P-values of ≤ 0.05 (*), ≤ 0.01 (**), ≤ 0.001 (***) and ≤ 0.0001 (****) indicating significant differences between groups.