NHBE cells and A549 cells were obtained from Lonza (Basel, Switzerland) and the American Type Culture Collection (Manassas, VA), respectively. MDCK cells were obtained from the Centers for Disease Control and Prevention (CDC)’s Division of Scientific Resources. Citrated whole blood was collected from healthy volunteers under a protocol approved by the CDC’s institutional review board, and PBMCs were isolated by centrifugation through Ficoll-Hypaque solution (Lymphoprep; Axis Shield, Oslo, Norway).

Influenza viruses used in this study include the laboratory H1N1 virus A/WSN/33, A/Puerto Rico/8/34 (PR8 [H1N1]), and H3N2 virus A/X-31; seasonal H1N1 virus A/Brisbane/59/2007 (A/Bris/59/07) and H3N2 virus A/Brisbane/10/2007; pandemic H1N1 viruses A/California/08/2009 and A/Mexico/4108/2009; and influenza B virus strain B/Brisbane/60/2008. All viruses were propagated for 2 d in the allantoic cavity of 10-d old embryonated chicken eggs. Pooled allantoic fluid was clarified by centrifugation, aliquoted, and stored at −80°C until use.

Female C57BL/6 mice, 6–8 wk old, were purchased from The Jackson Laboratory (Bar Harbor, ME).

Ultrapure Escherichia coli LPS (0111:B4), polyinosinic-polycytidylic acid [Poly(I:C)], and R848 and CpGA (ODN 2336) DNA (all from InvivoGen, San Diego, CA) were used. RIG-I inhibitor (epigallocatechin gallate) was from Sigma.

Human PBMCs were cultured in RPMI 1640 containing 10% FBS, 100 U/ml penicillin, 100 μg/ml streptomycin solution, and 2 mM l-glutamine (RPMI complete medium). A549 cells, a human epithelial cell line derived from lung adenocarcinoma (American Type Culture Collection), were grown in complete DMEM. For in vitro infection, viruses were diluted in medium containing 0.3% BSA. Cells were washed twice with medium only, and virus was added, at a volume of 100 μl at a multiplicity of infection (MOI) of 1.0, to each well. After 1 h of incubation with the virus, cells were washed with PBS and cultured in complete medium for FACS staining or in medium with 0.3% BSA and 1 μg/ml N-p-tosyl-l-phenylalanine chloromethyl ketone-treated trypsin (Sigma-Aldrich) for the virustiter assay. For the virus-inhibition assay in A549 cells, they were pre-incubated with UV-inactivated supernatants from purified monocytes infected overnight with A/Bris/59/07 virus or treated with PBS (mock infection). In some experiments, A549 cells were pretreated with mock supernatant plus rIFN-α and IFN-β (PBL InterferonSource, Piscataway, NJ) or A/Bris/59/07 supernatant with anti-IFN-α (clone 2 and 13) and anti-IFN-β (clone 3 and 15)-neutralizing Abs (PBL InterferonSource). For virus inactivation by UV irradiation, viruses or supernatants from virus-infected cultures were placed on ice and irradiated using a 254-nm UV stratalinker (Stratagene, La Jolla, CA) for 30 min. For heat inactivation, virus was incubated in a 56°C water bath for 40 min. To assess the impact of cytokines and chemokines on influenza-induced monocyte differentiation, NHBE cells or PBMCs were infected with A/Bris/59/07 overnight; supernatants were collected and UV inactivated. Fresh human PBMCs were then primed with the UV-inactivated culture supernatants for 4–6 h, followed by infection with A/Bris/59/07 (MOI = 1) overnight.

Human PBMCs were cultured with Poly(I:C) (10 μg/ml), LPS (1 μg/ml), R848 (5 μg/ml), and CpGA (10 μg/ml) for 16 h. Cells were then analyzed by flow cytometry to assess the frequencies of various cell types.

PBMCs were isolated from citrated blood samples. Cells were resuspended in PBS containing 10% FBS and stained for flow cytometry. The following Abs were used for analysis of DC subsets in PBMCs: CD3–PE–Cy7, CD14-Alexa Fluor 700, CD19–PE–Cy7, CD56–PE–Cy7, CD123–PerCP–Cy5.5, CD11c–allophycocyanin, and HLA-DR–allophycocyanin–Cy7 (BD Bioscience). For detection of the expression of α−2,6–linked sialic acids and α-2,3–linked sialic acids, lectin Sambucus nigra agglutinin (SNA) and Maackia amurensis agglutinin (MAA I) (both FITC conjugated) (EY Laboratories) were used. For analysis of monocyte, macrophage, and neutrophil infiltration in mouse lung tissue, CD11c-allophycocyanin, CD45-Alexa Fluor 700, CD11b-Pacific Blue, B220–PE–Cy7, F4/80-PE, Ly6C–allophycocyanin–Cy7, Gr-1–allophycocyanin, and Ly6G-FITC (BD Bioscience) were used. For analysis of DC infiltration in mouse lung tissue, CD11c-allophycocyanin, CD45-Alexa Fluor 700, CD11b-Pacific Blue, B220–PE–Cy7, I-A/E-PE, and CD103–PerCP–Cy5.5 (all from BD Bioscience) were used. The following Abs were used for analysis of cell activation: HLA-DR–allophycocyanin–Cy7, CD40-PE, and CD86-FITC (all from BD Bioscience). Samples were analyzed using an LSRII flow cytometer (BD Biosciences), and the cytometry data were analyzed using FlowJo software (Tree Star).

Human and mouse IFN-α and IFN-β in culture supernatant and mouse lung homogenates were quantified by ELISA kits (PBL Biomedical Laboratories, Piscataway, NJ). A customized panel of 12 inflammatory cytokines and chemokines (IP-10, TNF-α, IL-6, IL-8, RANTES, VEGF, IL-1α, IL-12p70, MCP-1, MIP-1β, IL-4, and IFN-γ) was used to measure cytokines with a Bio-Plex suspension array system (Bio-Rad, Hercules, CA).

A549 cells were washed with chilled PBS and then lysed in 200 μl ice-cold lysis buffer (50 Mm Tris-Cl [pH 8], 150 mM NaCl, 1% v/v Triton X-100, 2 mM EDTA, 1 mM PMSF, 20 μM leupeptin containing aprotinin 0.15 μg/ml) for 20 min at 4°C. Equal quantities of solubilized protein were resolved by 10% SDS-PAGE, blotted to nitrocellulose membrane, and probed with either a mouse monoclonal anti-influenza A NP Ab or an anti–β-actin Ab (Sigma). Proteins were visualized with SuperSignal West Pico chemiluminescent substrate (Pierce, Rockford, IL).

Confluent monolayers of MDCK cells in six-well plates were washed with DMEM and infected with serial 10-fold dilutions of cell culture supernatant for 1 h at 37°C in 5% CO₂. Cells were washed with DMEM and overlaid with 1.6% SeaKem LE agarose (Lonza) mixed 1:1 with 2× L15 medium (Lonza) containing 4 mM HEPES, 2 mM l-glutamine, 5 μg/ml gentamicin, 1.5 mg/ml sodium bicarbonate, and 1 μg/ml N-p-tosyl-l-phenylalanine chloromethyl ketone-treated trypsin (Sigma-Aldrich). Virus plaques were stained with 0.3% crystal violet solution (BD, Sparks, MD) after 72 h of incubation at 37°C in 5% CO₂ and counted.

FACS-sorted mDCs (5 × 10⁴) from mock- or A/Bris/59/07 virus-infected PBMCs were cultured together with CFSE-labeled naive CD4⁺CD45RA⁺ CD45RO⁻ or CD8⁺CD45RA⁺CD27⁺ T cells from a different donor (1 × 10⁵) in 200 μl RPMI complete medium in 96-well round-bottom plates. After 5 d, cells were collected and analyzed directly for cell proliferation.

Total RNA was isolated from sorted mDCs using the RNeasy Mini Kit (QIAGEN, Carlsbad, CA), and cDNA was generated using the Superscript First-Strand Synthesis System for RT-PCR using random hexamer primers (Invitrogen, Carlsbad, CA), according to the manufacturer’s protocol. Resulting cDNA was diluted 1:5, and 2 μl was used in a SYBR Green (SA Bioscience, Frederick, MD)-based real-time PCR reaction using a Mx3000 real-time PCR instrument (Stratagene, Cedar Creek, TX). GAPDH or β-actin was used to normalize the Ct values. Primer sets used for these studies are listed in Table I.

Female C57BL/6 mice were anesthetized with 2,2,2-tribromoethanol in tert-amyl alcohol (Avertin; Sigma-Aldrich) before intranasal inoculation with 100 50% mouse infectious dose (MID₅₀) of PR8 virus or PBS as a control. Lung tissues were collected on days 1, 3, 5, 7, and 10 postinfection (p.i.) to characterize cellular infiltrates by flow cytometry. In some animals, monocytes were depleted by clodronate-containing liposomes (5 mg/ml, 50 μl/each mouse; Encapsula NanoSciences, Nashville, TN) administered i.v. daily for 6 consecutive days, starting 18 h before PR8 infection. Control mice received an equivalent volume of PBS liposome. Cellular infiltrates in the lung were characterized on day 5 p.i. by flow cytometry. The animals were monitored for 14 d p.i. to assess morbidity. To determine lung viral titers, lungs were collected on days 3 or 5 p.i., each lung was homogenized in 1 ml cold PBS, and clarified homogenates were titrated in eggs to determine virus infectivity, starting at 1:10 dilution (limit of detection, 10^1.5 50% egg infective dose/ml). To assess the infection of monocytes in vivo, lung tissue was collected on day 5 p.i.; a single-cell suspension was prepared and stained for mDC markers, together with anti-NS1 Ab, to assess the presence of intracellular NS1 (an indication of virus infection). Animal research was conducted under the guidance of the CDC’s Institutional Animal Care and Use Committee in an animal facility accredited by the Association for Assessment and Accreditation of Laboratory Animal Care International.

All data were analyzed using Prism software (GraphPad Software). Data are expressed as mean ± SD, and the significances of differences were determined by the unpaired Student t test; p values < 0.05 were considered significant.

To examine the effect of influenza virus infection and TLR ligation, human PBMCs were infected with IAV or stimulated with Poly(I: C) (TLR3 ligand), LPS (TLR4 ligand), R848 (TLR7/8 ligand), and CpGA (TLR9 ligand) and analyzed by flow cytometry. As shown in Fig. 1A, 16 h p.i. with A/Bris/59/07 (MOI = 1), the percentage of HLA-DR⁺ lineage (CD3/CD19/CD56/CD14) marker-negative (Lin⁻) cells in total PBMCs were markedly increased. However, stimulation with ligands of TLRs 3, 4, 7/8, and 9 failed to increase the percentage of the HLA-DR⁺Lin⁻ population (Fig. 1A). The Lin⁻ and HLA-DR⁺ cells were further characterized for the expression of CD11c and CD123. mDCs were defined as CD11c⁺ CD123^low/−, and pDCs were defined as CD11c⁻CD123⁺. As shown in Fig. 1B, although influenza virus stimulation induced a >10-fold expansion of HLA-DR⁺Lin⁻CD11c⁺CD123^low/− mDCs, it had no such impact on pDC precursors. Similar data were obtained using PBMCs from multiple donors (Supplemental Fig. 1). DCs can originate either from common progenitor cells in the bone marrow or blood monocytes in the periphery (16, 17). Accumulating evidence suggests that during infection or inflammation, DCs differentiate more readily from monocytes (9, 10). Therefore, we examined whether the expanded mDCs were derived from monocytes. Monocytes were defined as CD14⁺HLA-DR⁺ cells within the entire PBMC population. Upon influenza virus infection but not LPS stimulation, monocytes in PBMCs differentiated into CD11c⁺CD123^low/− mDCs (Fig. 1C). To rule out the role of other cell types in PBMCs on the rapid differentiation of monocytes into mDCs, monocytes were purified from PBMCs using CD14 MicroBeads (Miltenyi Biotec, Auburn, CA) and infected with A/Bris/59/07. As shown in Fig. 1D, purified monocyte cultures differentiated into mDCs in vitro. Infection of cells with IAV can occur through the binding of viral surface hemagglutinin to sialylated glycans on the appropriate host cell (18, 19). We then assessed the expression of influenza virus receptor on the surface of monocytes. Monocytes strongly bound to lectin SNA, which primarily detects influenza receptor α-2,6–linked sialic acids, but not to MAA I, which primarily detects α-2,3–linked sialic acids (Fig. 1E). These findings suggest that monocytes expressed receptors used by human influenza viruses for entry into cells, which is consistent with a previous publication (20). Indeed, 16 h p.i., ~20–30% of monocyte-derived mDCs expressed IAV protein NS1, as detected by flow cytometry (Fig. 1F). Collectively, our results suggest that blood-circulating monocytes can be infected by influenza virus, because NS1 can only be found in infected cells and then differentiated into mDCs.

To determine whether rapid monocyte differentiation could also be induced by other strains of influenza viruses, we infected PBMCs with A/X-31 (H3N2), A/WSN/33 (H1N1), A/Brisbane/10/2007 (H3N2), A/Mexico/4108/2009, and A/California/08/2009 and found that these different IAV strains are capable of inducing rapid monocyte differentiation into mDCs (Fig. 2A). This ability to induce mDCs is not restricted to A viruses; type B influenza virus (B/Brisbane/60/2008) has the same effect on monocyte differentiation (Fig. 2B). These results suggest that monocyte differentiation into mDCs is a common mechanism in response to influenza virus infection. To examine whether viral replication is needed for the induction of mDCs, we infected PBMCs with live virus or heat- or UV-inactivated viruses. As shown in Fig. 2C, live virus infection was required for monocyte differentiation. Both UV- and heat-inactivated virus failed to induce monocyte differentiation. In vitro differentiation of DCs from monocytes using GM-CSF and IL-4 requires 5–7 d (8). In contrast, almost 100% of influenza virus-induced monocytes in PBMCs differentiated into mDCs by 16 h p.i. (Fig. 2D). Supernatants from A/Bris/59/07-activated PBMCs or primary NHBE cells failed to induce differentiation of fresh monocytes (Supplemental Fig. 2). These data suggest that influenza virus-induced monocyte differentiation is live virus dependent and is not triggered by soluble mediators secreted by activated cells.

As professional APCs, the typical DC function is to present Ags and trigger adaptive immunity. We next investigated whether influenza virus-induced mDC act as APCs. Activated DCs upregulate the expression of MHC class II and costimulatory molecules (CD86 and CD40). As shown in Fig. 3A, the expression of CD40, CD86, and MHC class II was substantially upregulated on mDCs activated by TLR ligands. However, there was little or no upregulation of these molecules on virus-induced mDCs (Fig. 3A). Next, we purified the mDCs following A/Bris/59/07 infection or LPS stimulation and determined the expression of proinflammatory cytokines. Although LPS-stimulated mDCs had increased expression of proinflammatory cytokine genes, including IL-12p35, IL-12p40, TNF-α, and IL-6, A/Bris/59/07 infection failed to induce the expression of these genes assessed by real-time RT-PCR using primers listed in Table I (Fig. 3B). Consistent with the gene-expression profile, Bio-Plex assay of cell culture supernatants of virus-induced mDCs demonstrated lower levels of proinflammatory mediators compared with LPS-activated mDCs (Fig. 3C). Interestingly, mDCs differentiated from circulating blood monocytes upon influenza virus infection secreted large amounts of MCP-1 and IP-10, which are strong chemoattractants for monocytes (Fig. 3D). Next, we evaluated whether the differentiated mDCs were able to stimulate T cells in an allogeneic T cell-proliferation assay. Naive CD45RA⁺CD45RO⁻ CD4 T cells and CD45RA⁺CD27⁺ CD8 T cells from donors were sorted, labeled with CFSE, and cultured together with blood mDCs sorted from mock- or virus-stimulated PBMCs from a different donor. After 5 d of coculture, as demonstrated by CFSE dilution, the uninfected blood mDCs triggered substantial CD4 and CD8 T cell proliferation. However, virus-induced mDCs could not induce a proliferative T cell response (Fig. 3E). Collectively, these results indicate that influenza virus-induced mDCs are poor in Ag presentation but could actively recruit more monocytes to the site of infection through production of MCP-1 and IP-10.

Because influenza virus-induced mDCs failed to function as conventional APCs as a result of their inability to upregulate costimulatory molecules and activate the allogeneic T cell proliferative response, we examined their ability to contribute to the antiviral response through the secretion of type I IFN. PBMCs were stimulated with LPS or A/Bris/59/07 for 16 h. mDCs were isolated for RNA extraction, and IFN-α/β mRNA expression was detected by real-time PCR. A significant increase in IFN-α/β mRNA expression was detected in virus-activated mDCs compared with mock-infected or LPS-activated mDCs (Fig. 4A). Furthermore, both IFN-α and IFN-β were detected in the culture supernatants of virus-activated mDCs, but neither was detected in the culture supernatant of mock-infected or LPS-activated mDCs (Fig. 4B). Type I IFN exerts antiviral effects through the induction of antiviral proteins encoded by ISGs. A/Bris/59/07 infection, but not LPS, induced a potent induction of the IFN-stimulated antiviral genes tetherin, viperin, and IFITM3 in mDCs (Fig. 4C). In addition, the cytosolic RNA sensors, RIG-I and MDA5, were upregulated in virus-differentiated mDCs (Fig. 4C). Release of type I IFN and the expression of ISGs and RIG-I/MDA5 by virus-differentiated mDCs suggest a possible role for mDCs in the antiviral response. Using an in vitro infection model, we tested whether the differentiated mDCs mediate a direct antiviral response. Purified monocytes were mock infected or infected with A/Bris/59/07 overnight. The culture supernatants were collected and inactivated with UV to prevent the replication of any virus present. A549 cells were preincubated with UV-inactivated culture supernatants for 30 min, followed by infection with A/Bris/59/07 virus. Viral replication was assessed by NP expression in the cell lysates, as well as the viral titer in the culture supernatant. Pre-incubation of A549 cells with supernatant from virus-induced mDCs resulted in a marked decrease in NP protein expression (Fig. 4D), as well as a 60% reduction in virus titer in the culture supernatant (Fig. 4E). Furthermore, the addition of 100 ng/ml of recombinant type I IFN decreased the virus titer in the culture supernatant. Similarly, the addition of anti–IFN-α and anti–IFN-β Abs significantly increased the virus titer (Fig. 4E). In summary, influenza virus-induced mDCs exhibited a direct antiviral activity by producing type I IFN and upregulating the expression of ISGs/RIG-I/MDA5.

To validate our in vitro findings with human PBMCs, we used an in vivo mouse model of influenza infection. Groups of mice were inoculated intranasally with 100 MID₅₀ of PR8 virus. At different time points p.i., lung cell suspensions were prepared for flow cytometry to characterize infiltrating cells. Monocytes were defined as CD11b^int, Gr-1^int, F4/80^low, Ly-6C^high cells (21). Two main populations of DCs were identified in the lung based on the expression of CD11c and B220. The main subset is CD11c⁺ B220⁻ myeloid (conventional) DCs with a minor population of the CD11c^intB220⁺ subset defined as pDCs. mDCs were further divided into two main subpopulations on the basis of CD11b and CD103 expression; CD11c⁺MHCII⁺CD11b⁺ cells localize to the submucosa and the lung parenchyma, whereas CD11c⁺MHCII⁺ CD103⁺ DCs preferentially localize to airway mucosa (22, 23). As shown in Fig. 5A, the total number of monocytes in the lung increased significantly, peaking between 4 and 7 d p.i. As expected, influenza virus infection also promoted a significant increase in the number of DCs in the lung, with peak levels occurring on day 7 p.i., concomitant with the peak of monocytes (Fig. 5B). Although mDCs, especially CD11b⁺ mDCs, were the most numerous following influenza infection, pDCs and CD103⁺ mDCs were minor populations. As is the case with human monocytes, mDCs in mice were infected; 13–15% of mDCs expressed IAV protein NS1 (Fig. 5C). Furthermore, on day 5 p.i., the CD11b⁺ mDCs expressed type I IFN mRNA (Fig. 5D) and secreted type I IFN in the culture supernatant (Fig. 5E). Taken together, our results demonstrate that, in vivo, IAV infection indeed recruited a substantial number of monocytes to the lung, with a corresponding increase in the number of type I IFN-producing mDCs.

To determine the source and antiviral function of virus-induced mDCs in vivo, we selectively depleted monocytes by treating mice with clodronate liposomes. Intravenous injection of clodronate liposome was shown to substantially reduce Gr-1^high monocytes, but not neutrophils and AMs (24), up to 24 h posttreatment. The depletion of monocytes was confirmed by the significant reduction in monocytes in lungs at 5 d p.i. by FACS staining (Fig. 6A). Interestingly, at day 5 p.i., the number of mDCs in the lung tissue of clodronate liposome-treated mice was also substantially reduced (Fig. 6A). Fifty percent of mice treated with clodronate liposome succumbed to infection (Fig. 6B). Furthermore, there was a significant reduction in type I IFN in lungs of monocyte-depleted mice (Fig. 6C). The virus titer in lungs was increased by 12-fold in monocyte-depleted mice on day 5 p.i. (Fig. 6D). These data indicate that monocyte depletion significantly reduced the number of virus-induced mDCs in lungs upon influenza virus infection. Furthermore, following influenza virus infection, monocyte depletion was associated with reduced type I IFN secretion and increased virus replication in the lungs and mortality.