SUBVERSION OF PULMONARY DENDRITIC CELL FUNCTION BY PARAMYXOVIRUS INFECTIONS¹

RSV and hMPV preparations

RSV A2 was grown in HEp-2 cells (American Type Culture Collection, Manassas, VA.) and purified by polyethylene glycol precipitation, followed by centrifugation on 35 to 65% discontinuous sucrose gradients as described elsewhere (25). The hMPV strain CAN97-83 was obtained from the Centers for Disease Control (CDC), Atlanta, GA, with permission from Dr. Guy Boivin at the Research Center in Infectious Diseases, Regional Virology Laboratory, Laval University, Quebec City, Canada. Virus was propagated and titrated in LLC-MK2 cells (ATCC, Manassas, VA) in MEM (without serum) containing 1.0 μg trypsin/ml (Worthington, Lakewood NJ), as described elsewhere (26).

Infection of mice and treatment protocol

Female, 8–10-week-old BALB/c mice were purchased from Harlan (Houston, TX) and were housed under pathogen-free conditions in the animal research facility of the University of Texas Medical Branch (UTMB), Galveston, TX, in accordance with the National Institutes of Health and UTMB institutional guidelines for animal care. Under light anesthesia, mice infected intranasally (i.n.) with 50μl of RSV or hMPV diluted in PBS (final administered dose: 10⁷ PFU) (27). Control mice were inoculated with the same volume of PBS (referred to hereinafter as mock infection). To monitor the progression of disease, daily determination of body weight was assessed.

Preparation of lung and lymph node cells

Lungs and lung draining lymph nodes (LN) were harvested at different time points up to 14 days after hMPV, RSV or mock infection. At the indicated time points (see Results) mice were anesthetized with an intraperitoneal injection of ketamine and xylazine before the thoracic cavity was opened. Mice were exanguinated and the trachea opened by incision of the cricothyroid membrane. Prior to harvest, lungs were gently perfused via the right ventricle with 5 ml of PBS containing 2 mM EDTA to remove blood cells from the pulmonary circulation. They were then cut into small pieces and incubated with collagenase A (0.5 mg/ml PBS; Sigma-Aldrich) and type IV bovine pancreatic DNase (20 μg/ml PBS; Sigma) for 1 h at 37°C. After digestion, lung cells were dispersed by shearing through a 20-gauge needle, followed by filtration through a nylon screen cell strainer (100 μm). Single cell suspensions were washed and contaminating erythrocytes were lysed using ACK lysis buffer (Biosource). Lung draining LN were consistently collected and digested to obtain single cell suspensions as mentioned above for lung tissue.

Flow cytometry

Lung and LN cells were incubated with anti-FcγRII/FcγRIII mAb (24G2; BD Biosciences, San Diego, CA). After washing, cells were stained with the following anti-mouse antibodies: anti-CD11c, anti-I-A/I-E (MHC-II), anti-CD11b, anti-CD49b/pan-NK (DX5), anti-Ly6C (Gr-1), anti-CD45R/B220, anti-CD103 (all from BD-Pharmingen, San Diego, CA) and anti-mPDCA1 (Miltenyi Biotec). In a separate set of experiments, lung cells were stained with anti-CD11c, anti-MHC-II in combination with anti-CD40, anti-CD80, anti-CD86, anti-programmed death-1 ligand (PD-L1) and anti-PD-L2 (BD-Pharmingen, San Diego, CA). Samples were stained at 4°C in PBS with 1% FBS and analyzed with a FACS Canto flow cytometer equipped with BD FACSDiva software (both from Becton Dickinson Immunocytometry Systems, San Jose, CA). Analysis was performed using FlowJo software (Tree Star V7.2.2, Ashland, OR).

Cell sorting

Lung cells were first enriched by MACS for specific CD11c⁺ and mPDCA-1⁺ populations (Automacs, Miltenyi Biotec), followed by FACS sorting isolation for cDC and pDC, respectively. CD11c⁺ cells were stained with Cy7-conjugated CD11c and FITC-conjugated MHC-II antibodies and the mPDCA1⁺ cells were stained with FITC-conjugated CD11c, PerCP-conjugated B220, and PE-conjugated Gr-1 (RB6-8C5) antibodies (BD-Pharmingen). CD11c^hiMHC-II^hi (cDC), and CD11c^intB220⁺Gr1⁺ cells (pDC) were sorted using FACS Aria instrument (Becton Dickinson Immunocytometry Systems, San Jose, CA). Routine post-sorting analysis was performed to ensure that that the purity was > 97%. DC were cultured in RPMI-1640 supplemented with 10% FCS, 2 mmol/L L-glutamine, 1 mM sodium pyruvate, and HEPES (complete medium).

Cytokine production

Lung cDC or pDC isolated from uninfected mice were infected in vitro with RSV or hMPV at multiplicity of infection (MOI) of 3 for 24h in 96-well plates (10⁵ cells/well) in a total volume of 200 μl. Cell-free supernatant was collected and tested for multiple cytokines using the Bio-Plex Mouse Cytokine 23-Plex panel (Bio-Rad Laboratories, Hercules, CA), according to the manufacture’s instructions. The panel included the following cytokines: IL-1α, IL-1β, IL-2, IL-3, IL-4, IL-5, IL-6, IL-9, IL-10, IL-12 p40, IL-12 p70, IL-13, IL-17, TNF-α, GM-CSF, IFN-γ, G-CSF, KC, CCL3 (MIP-1α), CCL4 (MIP-1β), CCL11 (Eotaxin), CCL2 (MCP-1) and CCL5 (RANTES). IFN-α and IFN-β were measured by ELISA (PBL Biomedical Laboratories, Piscataway, NJ). The IFN-α ELISA recognizes the αA, α1, α4, α5, α6, and α9 isoforms. In a separate set of experiments, pDC were isolated from mock-, hMPV-, and RSV-infected mice after 3 days of infection. 10⁵ cells/well were stimulated with 1 μg/ml of stimulatory oligodeoxinucleotide (ODN) with mouse-specific CpGs. We used ODN 1826 sequence (5′-tccatgacgttcctgacgtt-3′, Invivogen, San Diego, CA).

T cell Proliferation

For T cell proliferation assays, FACS sorted cDC from virus-infected or uninfected mice (10⁴ cells/well) were cultured for 3 days with CD4+ T cells from DO11.10 mice (10⁵ cells/well) in 96-well round bottom microtiter plates in complete medium. cDC were loaded with 1 μM of OVA peptide (323–339) 1h prior to co-culture with T cells. Cells were pulsed with 1 μCi/well of [³H]-thymidine for the last 18 h. Cells were harvested onto filters using a cell harvester (Brandel, Gaithersburg, MD) and incorporation of [³H]-thymidine to DNA was determined using a Wallac 1450 Microbeta/Trilux liquid scintillation counter (Perkin Elmer, Boston, MA).

Statistical analysis

Statistical analysis was performed using the InStat 3.05 biostatistics package (GraphPad, San Diego, CA) using a one way ANOVA to ascertain differences between groups, followed by a Tukey-Kramer test to correct for multiple comparisons. Unless otherwise indicated, results are expressed as mean ± SEM.

Trafficking of pDC, cDC and IKDC after hMPV or RSV infection

Lung DC are the primary immunological sentinels that can efficiently sense invading pathogens by a set of pattern recognition receptors (14,15). For an effective immune response to occur, DC must be able to sample the peripheral microenvironment and migrate through the afferent lymph to the lymph nodes (LN) where they prime naïve T cells (28). Herein, we determined the recruitment of the pDC, cDC and IKDC to the lung and LN after hMPV infection and compared to that induced by RSV. Mice were infected i.n. with hMPV, RSV or mock inoculum and assessed daily for body weight loss to evaluate the progression of the disease. At different time points, mice were sacrificed to analayze the number of DC tissue in lung and LN. As shown in Fig. 1A, hMPV-infected mice showed a biphasic body weigh loss at days 1 (modest) and 7 and total recovery by day 11. On the other hand, maximum body weight loss in RSV-infected mice occurred at day 2 after infection with gradual recovered by day 7. Recruitment of pDC was evaluated by identifying cells expressing CD11c^int/B220⁺/Gr-1⁺ (Fig. 1B). Following hMPV infection the number of pDC rapidly increased from 1 ± 0.1 × 10⁵ cells/lung (in mock-infected mice) to 4.7 ± 0.8 × 10⁵ cells/lung by day 1 and reaching a maximum of 8.4 ± 1.6 × 10⁵ cells/lung on day 8 (Fig. 1C). By day 14, number of lung pDC returned to levels similar to those of mock-infected mice. On the other hand, in RSV-infected mice, pDC peaked in the lung at day 3 after infection (8.2 ± 1.2 × 10⁵ cells/lung) and returned to basal levels by day 8. In regard to the trafficking of pDC to LN, we generally observed an increase in size and total number of cells in LN from infected mice as early as day 1 after infection. Number of pDC in LN was similar for both hMPV and RSV infections, although pDC were recruited in slightly higher numbers in hMPV-infected mice (Fig. 1D). pDC were rapidly recruited to LN by day 1 and peaked by day 6 (3.9 ± 0.3 × 10⁴) after hMPV infection and started to decline by day 8. Similar results were observed when pDC were identified by their expression of mPDCA1 and B220 (data not shown). Overall, the data show that in this murine model both hMPV and RSV induced a distinct kinetics of pDC recruitment to the lung that appeared overall to parallel the body weight curve. Pattern of pDC recruitment to the draining LN was similar for both viruses.

Recruitment of cDC to the lung and LN during hMPV or RSV infection was examined by the expression of CD11c^hi and MHC-II^hi (Fig. 2A). As shown in Fig. 2B, number of cDC increased from 1.6 ± 0.2 × 10⁵ cells/per lung on day 1 to a maximum of 32.8 ± 7.9 × 10⁵ cells/lung on day 10 after infection with hMPV. Thereafter, the number of lung pDC declined slowly, yet remained significantly elevated until day 18 (12.2 ± 1.3 × 10⁵ cells/lung) compared to mock-infected mice (2.2 ± 0.1 × 10⁵ cells/lung). Similar recruitment pattern was observed when mice were infected with RSV but lung cDC peaked earlier since a maximum of 32.3 ± 1.7 × 10⁵ cells/lung were observed at day 8 after infection. Trafficking of cDC to mediastinal LN was almost identical in hMPV or RSV infections, with an increasing time-dependent number of cells and a peak at day 6 after infection (Fig. 2C).

IKDC were identified by expression of CD11c^int/B220⁺/DX5⁺. Figure 3A shows the morphology of sorted lung IKDC stained by Wright-Giemsa. As shown in Fig. 3B, in hMPV-infected mice IKDC recruitment resembled that of pDC, i.e. as early as day 1 and with a first peak at day 2 (3.4 ± 0.4 × 10⁵ cells/lung) and a second peak at day 8 after infection (4.0 ± 0.1 × 10⁵ cells/lung). By day 10, the numbers of IKDC have returned to similar levels as those observed in mock-infected mice (1.5 ± 0.4 × 10⁵ cells/lung). On the other hand, in RSV-infected animals, IKDC peaked at day 3 (4.7 ± 0.7 × 10⁵ cells/lung) and returned to control levels by day 8 after infection (Fig. 3B). Recruitment of IKDC to LN was similar in both hMPV and RSV infections. As shown in Fig. 3C, IKDC started to be recruited by day 2 with a peak of 4.2 × 10⁴ cells at day 6 after infection, to rapidly decline to 0.7 × 10⁴ cells by day 8 after infection. Thus, IKDC trafficking after hMPV and RSV infection resembles that of pDC. Overall, these data indicate that IKDC is the smallest DC population recruited to the airways upon paramyxovirus infection, followed by pDC and being the cDC the predominant DC subset recruited to the lung.

Production of cytokines by ex-vivo infected lung DC

We have previously shown that hMPV and RSV infections induce a distinct profile of cytokines in human peripheral blood pDC and monocyte-derived DC (29). Since the response of lung DC to these paramyxoviruses has not been investigated, we next examined the ability of lung pDC and cDC to produce cytokines upon hMPV and RSV infection. We isolated lung pDC (CD11c^int/B220⁺/Gr-1⁺/mPDCA1⁺) and cDC (CD11c^hi/MHC-II^hi) from naïve mice (Figs. 4A and 5A) by cell sorting. Isolation of sufficient numbers of lung IKDC from naïve mice to carry out this set of experiments was not feasible. Cells were infected for 24 h and cell-free supernatants were tested for the presence of various cytokines including IFN-α and IFN-β. A highly purified pDC population was isolated from lungs after cell-sorting (Fig. 4A). Exposure of pDC to hMPV resulted in production of significant amount of IFN-α, while very low to undetectable levels were observed upon RSV infection (Fig. 4B). In addition, hMPV but not RSV was able to induce TNF-α and CCL5 production, although their concentrations did not reach statistical significance. Production of IL-12p40 was induced by both viruses, with hMPV being a stronger inducer than RSV. No significant difference in CCL3 and CCL4 production was observed after infection with either hMPV or RSV. Overall, RSV failed to induce cytokine production by lung pDC, with the exception of CCL3 and CCL4. Concentrations of IL-1α, IL-1β, IL-2, IL-3, IL-4, IL-5, IL-6, IL-9, IL-10, IL-12 p70, IL-13, IL-17, GM-CSF, IFN-γ, G-CSF, KC, CCL11, and CCL2 were at the lower limit of detection in control or virus-infected lung pDC (data not shown). These results collectively indicate that pDC produce a limited repertoire of cytokines in response to viral infection and that hMPV is a stronger inducer of cytokines compared to RSV.

When we determined the response of cDC to both hMPV and RSV, we observed that hMPV but not RSV was able to induce significant levels of IFN-β (Fig. 5B). Both viruses were able to induce significant amounts of IL-1α, IL-6, IL-10, KC and CCL11. Other cytokines including IL-12p40, IL-12p70, TNF-α, and chemokines CCL3, CCL4, and CCL5 were also similarly induced by hMPV and RSV but did not reach statistical significance when compared with uninfected cells. Levels of IFN-α were below the lower limit of detection in control or viral-infected cDC (data not shown). These data indicate that, with the exception of IFN-β production, hMPV induce a similar profile of cytokines than RSV in pulmonary cDC.

hMPV and RSV inhibit CpG-induced cytokine production by lung pDC

Evidence is accumulating that viral pathogens can use DC as a conduit for subverting the immune response and establishing infection in the host (30). We and others have previously shown that both RSV and hMPV have the ability to inhibit the production of cytokines such as IFN-α by human pDC in response to TLR9 agonists (29,31). However, it is unknown whether hMPV and RSV are capable of inhibiting cytokine production in lung pDC after infection in vivo. Therefore, BALB/c mice were infected with hMPV, RSV or mock for 3 days (when sizable numbers of pDC are recruited at the site of infection, see Fig. 1B). Lung pDC were isolated by cell sorting as described in Material and Methods. Cells (10⁵ pDC/well) were stimulated with 1 μg/ml of CpG-ODN 1826 and cell-free supernatant was collected after 24 h. As shown in Fig. 6, infection with hMPV or RSV significantly inhibited the capacity of lung pDC to produce IFN-α, IL-6, TNF-α, CCL3, CCL4, and CCL5 (by RSV infection) in response to CpG-ODN. We also observed a modest reduction in the production of CCL11, CCL5 (by hMPV infection) and CCL2 and a marginal change in IL-12p70 by any of the two infections in vivo. No change was observed in the production of IL-12p40 in any of the three cultures. These data demonstrate that both hMPV and RSV are able to interfere with the ability of lung pDC to respond to a secondary stimulus for the production of IFN-α and other cytokines.

Impaired antigen presenting capacity by lung cDC after viral infection

We have previously shown that RSV and hMPV are able to inhibit the ability of human monocyte-derived DC (moDC) to stimulate T cell proliferation in vitro (29). To investigate the effect of hMPV and RSV on the antigen presenting capacity of pulmonary cDC, we isolated lung cDC from infected and mock-infected mice at different time points after infection. cDC isolated from lungs at 1, 2, 3, and 6 weeks after hMPV, RSV or mock infection were loaded with OVA peptide and co-cultured with purified CD4⁺ T cells from DO11.10 mice. Co-cultures of pulmonary cDC from mock-infected mice with DO11.10 T cells resulted in a robust T-cell proliferation (Fig. 7). At the 1 week time point, lung cDC from hMPV- or RSV-infected mice were more efficient in stimulating CD4⁺ T cell proliferation than cDC isolated from mock mice. However, at week 2, we observed that infection with RSV significantly decreased the ability of cDC to stimulate CD4⁺ T cell proliferation by ~50%. cDC from hMPV-infected were also deficient in the induction of OVA specific T cell proliferation but in a lower extent (~20 %). At week 3 after infection, both viruses were able to decrease in a similar degree (~70%) the cDC stimulatory capacity when compared to cDC from mock-infected mice. By week 6 after infection, cDC from both RSV- and hMPV-infected mice were able to stimulate T cell proliferation as efficient as those from mock mice. When expression of RSV or hMPV antigens was assessed in lung cDC at 2 or 3 weeks after infection, RT-PCR assays indicated that cDC isolated from infected mice did not express viral antigen (data not shown). These findings indicate RSV and hMPV infection transiently impairs the ability of pulmonary cDC to present antigen to T cells.

Surface molecule expression by lung cDC after hMPV and RSV infection

The observed inhibition of T cell proliferation by lung cDC isolated from infected mice may be the result of altered expression of APC costimulatory molecules. Positive and negative costimulatory molecules individually or collectively regulate T cell activation thresholds. To examine the possibility that viral infection altered the surface molecule expression, lung cDC (CD11c^hi/MHC II^hi) were isolated from hMPV-, RSV- and mock-infected mice and were stained with anti-CD40, anti-CD80, anti-CD86, anti-PD-L1, and anti-PD-L2 antibodies. As shown in Fig. 8, expression of CD80 gradually increased over a three week period after RSV or hMPV infection, compared to mock-infrected mice. Similarly, PD-L1 expression increased progressively over the time of infection. Eight weeks after infection, expression of both CD80 and PD-L1 by cDC was comparable to that observed in mock-infected mice. Expression of CD86 was slightly increased by week 2 after RSV infection but not hMPV infection compared to cDC from mock-infecetd mice. CD40 and PD-L2 were not affected by viral infection. These findings indicate that both RSV and hMPV are able to induce activation of pulmonary cDC.

In addition to the surface molecules described above, we examined whether RSV or hMPV infection altered the phenotype of lung cDC subsets based on CD103 expression. This marker associates with two major cDC populations in mice, CD103⁺/CD11b^low and CD103⁻/CD11b^hi (32), which differ in their ability to prime CD4⁺ and CD8⁺ T cells (33), production of proinflammatory cytokines and generation of Foxp3-mediated regulatory function of naïve T cells (34). CD103 (αE) is the α-chain of the α_Eβ₇ integrin, which has been reported to be essential for the adhesion of human and mice intestinal lymphocytes to epithelial cells through the interactions with E-cadherin (35). CD103⁺ DC in the lung express high amounts of CD11c, MHC-II, and low levels of CD11b (32). It is currently unknown whether respiratory viral infections alter the phenotype of cDC based on expression of CD103. Analysis of cDC CD103⁺ population was performed by three-color flow cytometry analysis in total lung cells from virus- or mock-infected mice. Gated CD11c^hi/MHC-II^hi cells were further analyzed for the expression of CD103 (Fig. 9A, upper panels). Consistent with a previous a report, we found that the lung cDC (CD11c^hi/MHC-II^hi) compartment in uninfected mice is composed by a mixed populations of CD103⁺ and CD103⁻ cells (32). Moreover, CD11b expression was inversely correlated to CD103 expression since we observed CD103⁺/CD11b⁻ and CD103⁻/CD11b⁺ cells in the cDC subset (data not shown). However, we observed that the percentage of CD103⁺ cDC decreased from ~50% to ~20% after 7 days of infection, as shown in Fig. 9A (lower panels). Percentage of CD103⁺ cDC was consistently decreased in infected mice at week 2 and week 3 but returned to normal levels at week 8 after infection (Fig. 9B).