DNGR‐1 is dispensable for CD8+ T‐cell priming during respiratory syncytial virus infection

During respiratory syncytial virus (RSV) infection CD8+ T cells both assist in viral clearance and contribute to immunopathology. CD8+ T cells recognize viral peptides presented by dendritic cells (DCs), which can directly present viral antigens when infected or, alternatively, “cross‐present” antigens after endocytosis of dead or dying infected cells. Mouse CD8α+ and CD103+ DCs excel at cross‐presentation, in part because they express the receptor DNGR‐1 that detects dead cells by binding to exposed F‐actin and routes internalized cell debris into the cross‐presentation pathway. As RSV causes death in infected epithelial cells, we tested whether cross‐presentation via DNGR‐1 is necessary for CD8+ T‐cell responses to the virus. DNGR‐1‐deficient or wild‐type mice were intranasally inoculated with RSV and the magnitude of RSV‐specific CD8+ T‐cell induction was measured. We found that during live RSV infection, cross‐presentation via DNGR‐1 did not have a major role in the generation of RSV–specific CD8+ T‐cell responses. However, after intranasal immunization with dead cells infected with RSV, a dependence on DNGR‐1 for RSV‐specific CD8+ T‐cell responses was observed, confirming the ascribed role of the receptor. Thus, direct presentation by DCs may be the major pathway initiating CD8+ T‐cell responses to RSV, while DNGR‐1‐dependent cross‐presentation has no detectable role.


Introduction
Respiratory syncytial virus (RSV) is an important respiratory virus of the single-stranded RNA virus family Paramyxoviridae. RSV disease burden is estimated at 64 million cases and up to 160,000 deaths every year worldwide (www.who.int). While RSV disease manifests as a simple common cold in the majority of cases, between 2 and 3% of children develop severe bronchiolitis. Although most of these children recover, they have a greater risk of developing recurrent wheeze and asthma-like symptoms in later childhood [1,2]. RSV infection induces a robust virus-endocytosis or phagocytosis and then diverted into the MHC-I presentation pathway [6]. The ability to cross-present exogenous antigens is particularly well developed in dendritic cells (DCs). DCs are professional antigen presenting cells (APCs) that have the ability to take up antigen in the periphery and then migrate to the lymph node where they can present antigen to naïve T cells. Cross-presentation has presumably evolved to ensure that DCs can present viral antigens to CD8 + T cells and prime cytotoxic T lymphocytes (CTLs) even in situations when they are not themselves infected or where viruses encode inhibitors of endogenous antigen processing.
Several distinct subsets of DCs are found across tissues in both mouse and man but it is unclear how their dynamics change during viral infection. In mouse models of RSV infection, lung CD103 + and CD11b + DCs migrate to the lymph nodes [7,8] while monocyte-derived DCs migrate into the infected lungs [9][10][11]. Lung CD103 + DCs are similar to lymphoid tissue-resident CD8α + DCs in having a superior ability to cross-present antigens to CD8 + T cells [12][13][14]. The two DC subtypes are ontogenetically related and share the ability to avidly phagocytose dead or dying cells and cross-present antigens extracted from the phagocytosed material. This is partly dependent on the C-type lectin receptor, DNGR-1 (Dendritic cell NK Lectin Group Receptor 1) (also known as CLEC9A), which is uniquely expressed by CD103 + and CD8α + DCs and their human equivalents [15][16][17][18][19]. DNGR-1 binds F-actin, which is exposed by dying cells at the point when they lose membrane integrity [20,21]. Upon recognition of F-actin, DNGR-1 signals to prevent phagosomal maturation and this facilitates antigen preservation and subsequent cross-presentation, thereby favoring CD8 + T-cell responses against antigens present in the cellular debris, including ones of viral origin [22][23][24].
RSV infection induces cell death and syncytia formation [1,25] but it is currently not known whether virus-induced cell death influences the initiation and magnitude of the ensuing immune response. Our work aimed to elucidate whether DNGR-1 and, by inference, cell death contributes to CD8 + priming in vivo during RSV infection. We found that DNGR-1-deficient mice mount reduced virus-specific CD8 + T-cell responses to intranasal inoculation with dead RSV-infected cells. However, upon live virus infection, RSV-specific CD8 + T-cell responses are equivalent between DNGR-1 knockout (KO) and wild-type (WT) mice. Thus, cross-presentation of epitopes from dead RSV-infected cells via DNGR-1 can influence the magnitude of CD8 + T-cell responses but is not crucial during RSV infection, suggesting that direct presentation of RSV epitopes by infected DCs may be the main pathway for CD8 + T-cell induction during RSV infection.

DNGR-1 plays no role in viral load, cell infiltration or cytokine production during RSV infection
In order to evaluate whether DNGR-1 KO mice have any defects in innate anti-RSV immunity, WT and DNGR-1 KO mice were intranasally inoculated with 1 × 10 6 focus forming units (FFU) of RSV on day 0. Copies of RSV L gene RNA were detected on day 1, 4 (peak), and 8 post infection (p.i.) using qPCR (Fig. 1A). There was no noticeable difference in viral replication or clearance between WT and DNGR-1 KO mice. In order to assess lung inflammation, the cellular infiltration into the bronchoalveolar lavage (BAL) was investigated. The total number of cells in the BAL was not different between DNGR-1 KO and WT mice (Fig. 1B). Furthermore, in both DNGR-1 KO and WT mice, accumulation of neutrophils in the BAL was detected at day 1 p.i. and lymphocytes were detectable at day 4 and were further increased by day 8 p.i. (Fig. 1B). The frequency of macrophages was not different between DNGR-1 KO and WT mice and no eosinophils were detected in any BAL samples at any time point (data not shown). In addition, early inflammatory mediators (IFN-α, CXCL10, and IL-6) were analyzed in the BAL on day 1 and 4 and all were detected at similar levels in WT and DNGR-1 KO mice (Fig. 1C). Thus, DNGR1 expression does not appear to be essential for the early immune response or viral clearance during RSV infection.

RSV pathology and T-cell infiltration to the lung is not dependent on DNGR-1
In the mouse model, RSV disease severity as measured by weight loss has been linked to a potent response by RSV-specific T cells [1,3,26]. As DNGR-1 has been shown to mediate cross-priming of CTLs during virus infection [23,24], the lack of DNGR-1 might lead to reduced T-cell priming and therefore decreased disease severity. RSV infected WT mice showed weight loss starting on day 5 p.i. with about 15% weight loss on day 6 and recovery on day 8 p.i. (Fig. 2A). Contrary to our hypothesis, DNGR-1 KO mice showed equal weight loss to the WT mice after RSV infection ( Fig. 2A). In addition, total lung cell numbers after RSV infection were equivalent between DNGR-1 KO and WT mice (Fig. 2B) and there was no difference in frequency and absolute number of CD8 + T cells ( Fig. 2C and E). The absolute numbers of CD4 + T cells in the lung did not increase after infection ( Fig. 2C and D) as previously shown [27,28]. Finally, no difference in RSV-specific antibodies was found after primary or secondary RSV infection in DNGR-1 KO mice compared with littermate controls (data not shown). Thus, the absence of DNGR-1 does not affect the influx of T cells into the lungs during RSV infection or the overall pathology.

DNGR-1 deficiency does not affect the induction of RSV-specific CD8 + T-cell responses
We further examined whether the RSV-specific CD8 + T-cell response was affected. First, we quantified the frequency of CD8 + T cells specific for an immunodominant epitope derived from the M2 protein [29,30] by using Alexa Fluor-647 conjugated MHC class I tetramer-M2 82-90 complexes (Fig. 3A). Tetramer + CD8 + T cells were detected in the BAL and lungs at approximately 25% of CD8 + T cells on day 8 post RSV infection (the peak of T-cell However, around 40% of the CD8 + T cells produced IFN-γ after stimulation with M2 82-90 peptide but, importantly, there was no difference between the DNGR-1 KO and WT mice (Fig. 4B). In addition, IFN-γ levels in the BAL were examined at day 8 p.i. and similar IFN-γ production was detected in DNGR-1 KO and WT mice (Fig. 4C). Furthermore, total T-cell responses were quantified by restimulating lung single-cell suspensions with M2 82-90 peptide at various concentrations (5 ng/ml; 0.5 ng/ml; 0.05 ng/ml) or with RSV (MOI of 1) and measuring IFN-γ levels in culture supernatants, which again showed no difference between DNGR-1-deficient and WT mice (Fig. 5A).
The above analysis focused exclusively on the immunodominant epitope M2 82-90 . We performed additional restimulations of lung cells from RSV infected mice with F 85-93 and F 93-106 peptides corresponding to two subdominant epitopes that, like M2 82-90 , are presented by H-2K d [31,32]. Compared with restimulation in medium alone, F peptides induced IFN-γ production from lung cells above background although this response was much weaker than that seen with M2 restimulations. Importantly, there were no differences in the response to either F peptide between DNGR-1 KO and WT mice (Fig. 5B). Taken together, these data indicate that DNGR-1 expression is not critical for the induction of RSV specific CD8 + T cells in BALB/c mice.

Lack of DNGR-1 reduces RSV-specific CD8 + T cells after exposure to dead virus-infected cells
In order to assess whether DNGR-1 can have any impact on the induction of anti-RSV CD8 + T cells, mice were inoculated intranasally with UV-treated, RSV-infected human epithelial cell line 2 (HEp-2 cells). Using this model, we minimized the direct infection of DCs and increased the likelihood that any anti-RSV CD8 + T-cell responses resulted from cross-presentation of the inoculum. There was no weight loss detected in any of the groups after inoculation with infected dead cells (UV-RSV HEp-2; data not shown) and the number of total cells and CD8 + tetramer + cells in the lung were similar in DNGR-1 KO and WT mice ( Fig. 6A and B). However, fewer cells were found extravasating into the airways of DNGR-1 KO mice compared with WT controls as determined by BAL analysis (Fig. 6C). The number of RSV-specific CD8 + T cells in BAL was also reduced (Fig. 6D) and lower levels of IFN-γ in bronchoalveolar lavage fluid (BALF) were found in DNGR-1 KO mice compared with WT mice (Fig. 6E). We conclude that DNGR-1 deficiency can impact on anti-RSV immunity induced by intranasal immunization with dead infected cells. There were no significant differences between the groups using unpaired, two-tailed, unpaired Mann-Whitney test.

Discussion
CD8 + T cells are an important part of the immune response against RSV infection. They are important for viral clearance but at the same time can contribute to immunopathology, at least in mouse models. It is not clear to what extent cross-presentation of viral antigens from dead cells contributes to the CD8 + T-cell response during respiratory infections. As RSV is cytopathic and can cause syncytia formation in epithelial cells, we wanted to investigate whether RSV-triggered cell death is important for the initiation of CD8 + T-cell responses. One way to address this issue is to use mice lacking DNGR-1, which is known to bind to the actin cytoskeleton from damaged cells and ferry associated antigens into the crosspresentation pathway [33]. DNGR-1 + CD103 + DCs are present in the lung and are thought to be one of the main migratory lung DC populations [7,10,11,34,35]. In previous studies, DNGR-1 KO mice were shown to be defective in mounting specific CD8 + T-cell responses to cytopathic lung infection with herpes simplex virus 1 (HSV-1) or skin infection with vaccinia virus [23,24]. Notably, we show a dependence on DNGR-1 for the RSV-specific CD8 + T cells in the airways when mice are inoculated with UVkilled RSV-infected cells. This demonstrates that DNGR-1 can play a role in the anti-RSV response if the lung is exposed to infected dead cell material but, interestingly, we only detect a difference in number of RSV-specific CD8 + T cells in the airways (BAL) and not in the lung tissue. This could be due to a decreased inflammation of the lungs of DNGR-1 KO mice with less infiltration of cells into the airways or that RSV-specific CD8 + T cells in WT mice are more activated and have the ability to migrate into the airways. Independently of these considerations, we found that DNGR-1 is dispensable for the anti-RSV CD8 + T-cell response in the lungs during live virus infection. This is in contrast to what has been found during HSV-1 lung infection [23]. It is possible that the quantity of virally infected dead cell material and/or the host cell for viral replication differ during RSV and HSV infection and therefore the two models differentially depend on cross-presentation via DNGR-1 for mounting CD8 + T-cell responses. Alternatively, or in addition, the rapid decrease of CD103 + DC from the lung early after RSV infection [7] may lead to the contribution of this DC subset, and, consequently, the contribution of DNGR-1, being negligible. It will be interesting to address this point by using BALB/c mice selectively lacking CD103 + and CD8α + DCs.
Overall, our data show that cross-presentation of RSV infected dead cell material via DNGR-1 is not a major pathway for initiating the CD8 + T-cell response during RSV infection. We cannot exclude the possibility that cross-presentation via DNGR-1 might contribute to the priming of a small population of CD8 + T cells specific for epitopes other than the three studied here. It is also still possible that other, DNGR-1independent routes of cross-presentation are important. Nevertheless, the results from this study suggest that DNGR-1 is not a major player in the initiation of RSV immunity and help our understanding of the basic biology behind immunity to this virus.

Mice, virus stocks, and infection
Clec9a gfp/gfp (DNGR1-deficient) mice [22] were kindly provided by C. Reis e Sousa (Cancer Research UK London Research Institute, London, UK). These were backcrossed six generations onto a BALB/c background, then bred and housed in specific-pathogen free conditions according to the UK Home Office guidelines. Clec9a +/+ (wild type) and Clec9a +/gfp (heterozygous) littermates or age and sex-matched WT BALB/c mice purchased from Harlan Laboratories (UK) or Charles River Laboratories (UK) were used as controls. No differences were ever noticed between these control groups and are all therefore referred to as WT mice.  Plaque-purified human RSV (A2 strain from ATCC) was grown to a high titer in HEp-2 cells. Seven-to ten-week old Clec9a gfp/gfp , Clec9a +/gfp , Clec9a +/+ , and WT BALB/c mice were lightly anesthesized with isofluorane and challenged intranasally (i.n.) with a dose of 1 × 10 6 FFU of RSV on day 0.
To prepare UV-killed RSV-infected HEp-2 cells, 80% confluent HEp2 cells were infected with RSV at an MOI of 0.5 in serumfree media (Dulbecco's Modified Eagle medium (DMEM) supplemented with 2 mM L-glutamine). After 2 h incubation at 37°C, serum-containing media (DMEM supplemented with 10% fetal bovine serum and 2 mM L-glutamine) was added. Infected cells were harvested after 24 h, washed and resuspended in sterile PBS and UV-killed using 750 mJ/cm 2 treatment twice in a CX-2000 UV cross-linker (UVP). Cells were further rested for 2 h at room temperature (RT) and mice were then inoculated i.n. with 2 × 10 6 UV-killed RSV-infected HEp-2 cells.

Cell isolation and processing
Mice were culled using a fatal dose of pentobarbital injected intraperitonially (i.p.) according to UK home office guidelines. BAL was collected by flushing the lungs three times with 1 mL of PBS supplemented with 0.5 mM EDTA (Life Technologies). Lung lobes were collected and digested with Collagenase D (1 mg/mL, Roche) and DNase I (30 μg/mL; Sigma-Aldrich) using a gentleMACs cell dissociator (Miltenyi Biotech) according to the manufacturer's protocol and incubated at 37°C for 30 min. Red blood cells were lysed by treating lung cells with ACK lysing buffer and lung cells were passed through 100 μM cell strainers to create single-cell suspensions. Total cell counts were determined using hemocytometer slides and dead cells were excluded by Trypan blue staining (Sigma-Aldrich). To determine the cellular composition in the BAL, cells were transferred onto a microscope slide (Thermo Scientific, UK) using a Shandon Cytospin 3 Centrifuge and slides were stained with hematoxylin and eosin (H&E; Reagena, Gamidor, UK). Cells were categorized as macrophages, lymphocytes, neutrophils, and eosinophils based on their morphology and size under a light microscope.
Biotinylated RSV M2 82-90 peptide monomers (H-2K d / SYIGSINNI) were obtained from the NIH Tetramer Core Facility (Emory University Atlanta, GA, USA) and tetramerization was performed in-house using allophycocyanin-conjugated Streptavidin (Molecular Probes) using a protocol provided by the NIH Tetramer Core Facility. M2 tetramer staining was performed on lung and BAL cells following Fc block steps and prior to surface receptor staining. Cells were incubated with APC-conjugated M2 tetramer in FACS staining buffer for 30 min in the dark at RT. Cells were fixed for 30 min in BD Cytofix/Cytoperm fixation buffer at 4°C, washed and then analyzed on the flow cytometer.
For intracellular detection of IFN-γ, lung cells were stimulated in 96-well plates with 2 μg/mL of the immunodominant RSV M2 82-90 peptide (H-2K d /SYIGSINNI; Synthetic Biomolecules, San Diego, CA, USA) or the irrelevant M 187-195 peptide (H-2D b / NAITNAKII; Synthetic Biomolecules, San Diego, CA, USA) identified to be immunodominant in C57BL/6 (H-2D b ) mice [5] in complete DMEM (cDMEM; supplemented with 10% fetal bovine serum, 2 mM L-glutamine, 100 U/mL penicillin and 100 μg/mL streptomycin). After 1 h incubation, monensin (Golgi Stop, BD) was added. Cells were incubated a further 3 h, washed and incubated with live-dead fixable stain, Fcγ II/III receptor and surface (E) Levels of IFN-γ in the BAL fluid were determined using ELISA. Each symbol represents an individual mouse and data are shown as mean ± SEM of 6-10 mice pooled from two individual experiments except in (B) were data are presented as the mean ± SEM (n = 4-5 mice per group) and are from a single experiment representative of two independent experiments. ***p ࣘ 0.001, *p ࣘ 0.05, unpaired, two-tailed, unpaired Mann-Whitney test.
receptor antibodies as described above. Intracellular cytokine staining was performed using the Cytofix/Cytoperm kit (BD) according to manufacturer's instructions. Briefly, cells were fixed in BD Cytofix/Cytoperm buffer for 30 min at 4°C, washed with FACS staining buffer and then incubated with Fcγ II/III receptor (CD16/CD32) antibody in BD Perm Wash Buffer for 15 min. Cells were then incubated with PE-conjugated anti-IFN-γ antibody (clone XMG1.2, BD biosciences) in Perm Wash buffer for 1 h at 4°C, washed and transferred to FACS staining buffer for analysis. Cells were acquired on the LSR Fortessa flow cytometer (BD) and data were analyzed using Flowjo Software (Version 9.6.4). Cells were gated for nondebris, live cells, and singlets and then analyzed for indicated markers.

Gene Expression
Total RNA was extracted from homogenized lung tissue using TRIzol RNA Isolation reagent in combination with a chloroform separation, followed by washes with isopropanol and 75% ethanol as per the protocol provided (TRIzol Reagent, Ambion, Life Technologies, UK). RNA purity and concentration was determined using the Nanodrop 1000 (Thermo Scientific) and 1-2 μg of total RNA was reverse-transcribed to cDNA using the high capacity RNA-to-DNA kit according to the manufacturer's instructions (Invitrogen, Life Technologies). Real-time quantitative PCR was performed to determine mRNA levels of RSV L gene using primers (Invitrogen) and probes (Eurofins) [36]. Results were normalized to Gapdh (Applied Biosystems) and the exact copy numbers of L gene was calculated using an internal plasmid standard to give an absolute quantification. Analysis was performed using the Quantitect Probe PCR Master Mix (Qiagen) and the 7500 Fast Real-time PCR System (Applied Biosystems).
Levels of IL-6, CXCL10 (IP-10) and IFN-γ protein were measured in the BAL fluid using ELISA kits and following the manufacturer's instructions (all from R&D Systems). IFN-α levels in the BAL were measured by ELISA as previously described [37]. Data was acquired on a SpectraMax Plus plate reader (Molecular Devices) and analyzed using SoftMax software (version 5.2).