Influenza Virus Overcomes Cellular Blocks To Productively Replicate, Impacting Macrophage Function

ABSTRACT

Whether influenza virus replication in macrophages is productive or abortive has been a topic of debate. Utilizing a panel of 28 distinct human, avian, and swine influenza viruses, we found that only a small subset can overcome cellular blocks to productively replicate in murine and primary human macrophages. Murine macrophages have two cellular blocks. The first block is during viral entry, where virions with relatively acid-stable hemagglutinin (HA) proteins are rendered incapable of pH-induced triggering for membrane fusion, resulting in lysosomal degradation. The second block is downstream of viral replication but upstream of late protein synthesis. In contrast, primary human macrophages only have one cellular block that occurs after late protein synthesis. To determine the impact of abortive replication at different stages of the viral life cycle or productive replication on macrophage function, we assessed cytotoxicity, nitric oxide or reactive oxygen species production, and phagocytosis. Intriguingly, productive viral replication decreased phagocytosis of IgG-opsonized bioparticles and Fc receptor CD16 and CD32 surface levels, a function, to our knowledge, never before reported for an RNA virus. These data suggest that replication in macrophages affects cellular function and plays an important role in pathogenesis during infection in vivo.

IMPORTANCE Macrophages are a critical first line of defense against respiratory pathogens. Thus, understanding how viruses evade or exploit macrophage function will provide greater insight into viral pathogenicity and antiviral responses. We previously showed that only a subset of highly pathogenic avian (HPAI) H5N1 influenza virus strains could productively replicate in murine macrophages through a hemagglutinin (HA)-mediated mechanism. These studies expand upon this work and demonstrate that productive replication is not specific to unique HPAI H5N1 viruses; an H1N1 strain (A/WSN/33) can also replicate in macrophages. Importantly, we identify two cellular blocks limiting replication that can be overcome by an avian-like pH of activation for nuclear entry and a yet-to-be-identified mechanism(s) to overcome a postnuclear entry block. Overcoming these blocks reduces the cell's ability to phagocytose IgG-opsonized bioparticles by decreasing Fc receptor surface levels, a mechanism previously thought to occur during bacterial and DNA viral infections.

INTRODUCTION

Macrophages are one of the first lines of defense against infection. They are poised to secrete large amounts of cytokines, orchestrate the adaptive immune system, and clear infected and dying cells to aid in recovery (1). Further, alveolar macrophages are essential in preventing respiratory failure after infection (2) and for preventing bacterial superinfections during influenza infection (3). However, such responses must be tightly regulated, as excessive cytokine levels contribute to immunopathology and disease severity during infection (4–6).

Whether influenza viruses productively replicate in macrophages is controversial. Several reports state that influenza virus replication is abortive in macrophages (7, 8), while others report productive replication (9–12). These reports varied in the type of macrophages used as well as influenza strains, which could account for the differences in replication. Recently, our group used a panel of H1 to H16 influenza viruses to show that only a subset of highly pathogenic avian (HPAI) H5N1 viruses, those associated with high morbidity and mortality in humans, could productively replicate in primary murine alveolar macrophages and the murine macrophage cell line Raw264.7 (13). We defined productive replication as the production of infectious progeny that could spread to neighboring cells (13). Using a representative abortive viral strain, pandemic H1N1 A/California/04/09 (CA/09) virus, we showed that replication was aborted early in the viral life cycle: prevention of efficient nuclear entry of the vRNPs (13). This block could be overcome by the HPAI H5 hemagglutinin (HA) protein (13).

These studies expand on our initial findings and demonstrate that the A/WSN/33 (WSN) H1N1 virus, like HPAI H5N1 strains, can overcome two cellular blocks to productively replicate in murine macrophages. Using a panel of human and avian influenza viruses, we found that the majority of human strains are unable to exit the endosome prior to accumulation in the lysosome-associated membrane protein 1 (LAMP1)-positive lysosomes due to a pH of activation incompatible with the macrophage intracellular environment. In contrast, most avian strains have a compatible pH of activation and can accumulate in the nucleus but fail to translate, highlighting a second cellular block exists. Interestingly, all viral strains tested in primary human macrophages are able to enter the nucleus, but the majority fails to productively replicate. However, late translation occurred, suggesting that the postnuclear entry block varies between murine and human cells. Finally, we demonstrate that productive replication decreased macrophage phagocytic ability by reducing Fc receptor surface levels, a function never before attributed to RNA viruses. These studies further elucidate the mechanism of influenza replication in macrophages and, importantly, that influenza replication in macrophages affects macrophage function by decreasing phagocytosis.

RESULTS

WSN virus replicates in Raw264.7 macrophages through an HA/NA-dependent mechanism.

We previously demonstrated that a subset of HPAI H5N1 influenza viruses could overcome a cellular block in nuclear entry to productively replicate in Raw264.7 macrophages and primary murine alveolar macrophages through an HA-dependent mechanism (13). By expanding the panel of human, swine, and avian viruses tested for replication in Raw264.7 macrophages (Table 1), as defined by production of infectious progeny that could spread to neighboring cells, we found that the H1N1 A/WSN/33 (WSN) virus could also enter the nucleus (Fig. 1A) and productively replicate in Raw264.7 macrophages, similar to VN/1203 HPAI H5N1 virus (Fig. 1B).

TABLE 1.

Summary of influenza virus replication in Raw264.7 macrophages

Virus	Subtype	Replication	Receptor	Block in nuclear NP entryc	NS1 productionc
A/WSN/33	H1N1	Yes	2,6b	No (++++)	++++
A/CA/04/09	H1N1	Noa	2,6b	Yes (−)a	−a
A/New Caledonia/20/99	H1N1	No	2,6b	Yes (−)	−
A/Puerto Rico/8/34	H1N1	No	2,3b	Yes (−)	ND
A/Tennessee/F1052/10	H1N1	No	2,6b	Yes (−)	ND
A/Brisbane/10/07	H3N2	Noa	2,6b	No (++)	ND
A/Tennessee/F4008/12	H3N2	No	2,6b	Yes (−)	ND
A/HK/68	H3N2	No	2,3/2,6b	No (++)	++
A/Wuhan/359/95	H3N2	No	2,6b	No (++)	+
A/Swine/NE/2/92	H1N1	No	2,6b	Yes (+)	ND
A/Swine/NC/0668/11	H3N2	No	2,3b	Yes (+)	ND
A/Swine/TX/4119-2/98	H3N2	No	2,6b	No (+++)	ND
A/GreyTeal/Australia/2/79	H4N4	No	2,3	No (++)	ND
A/HK/486/97	H5N1	Noa	2,3	ND	ND
A/HK/213/03	H5N1	Noa	2,3	No (ND)	ND
A/VN/1203/04	H5N1	Yesa	2,3	No (++++)	ND
A/HK/483/97	H5N1	Yesa	2,3	No (ND)a	Yesa
A/Duck/Bangladesh/19097/13	H5N1	No	2,3	No (++++)	ND
A/Mallard/Alberta/85/76	H5N2	Noa	2,3	No (++++)	−
A/Duck/Potsdam/2216-4/84	H5N6	Noa	2,3	No (++++)	−
A/Turkey/WI/1/68	H5N9	No	2,3	No (+++)	+++
A/Teal/HK/W312/97	H6N1	Noa	2,3	No (++++)	++
A/Anhui/01/13	H7N9	No	2,3	ND	ND
A/WSN/33 + CA/09 HA	H1N1	Yes	2,6	No (++++)	ND
A/WSN/33 + CA/09 HA/NA	H1N1	No	2,6	No (++)	ND
A/CA/09 + HK/486/97 HA	7 + 1 H5N1	No	2,3	ND	ND
A/CA/09 + VN/1203/04 HA	7 + 1 H5N1	Yesa	2,3	No (ND)	ND
A/CA/09 + HK/483/97 HA	7 + 1 H5N1	Yesa	2,3	No (ND)	ND

^aAlso reported in Cline et al. (13).

^bDetermined by the receptor binding assay described in the Materials and Methods.

^cPresented as the percentage of staining normalized to the percentage of nuclear staining in MDCK cells (−, <10%; +, 10 to 25%; ++, 26 to 50%; +++, 51 to 75%; ++++, >75%). ND, not determined.

FIG 1.

A/WSN/33 replicates in Raw264.7 macrophages and requires HA and NA. MDCK cells and Raw264.7 macrophages were infected with CA/09, WSN, or VN/1203 (A and B) or reverse genetics viruses (C and D) (MOI of 2 [A and C] or 0.01 [B and D]). Cells were fixed at 4 hpi and stained for nucleoprotein (NP; green). Nuclei were stained with DAPI (blue). (A and C) Representative images are shown. The percentage of nuclear NP-positive cells was calculated, and means ± standard deviations are shown in the bottom left corner (A), or the percentage of nuclear NP in Raw264.7 macrophages standardized to the percentage of nuclear NP in MDCK cells to account for differences in infection (percentage of MDCK cells) is shown (C). (B and D) Viral replication was measured by TCID₅₀ analysis using MDCK cells in the viral supernatant. Results are reported as the viral titer at 2 hpi (residual virus after infection) subtracted from the viral titer at 24 hpi. Data are averages from 2 to 5 independent experiments performed at least in duplicate. Error bars indicate standard deviations (SD). The bar is 50 μm.

To determine if the WSN HA protein was involved, we wanted to rescue CA/09 virus expressing WSN HA. However, we cannot perform this experiment given the current moratorium on influenza gain-of-function research. Therefore, we performed loss-of-function experiments replacing the WSN HA, neuraminidase (NA), or HA and NA genes with that from CA/09 virus and rescued the viruses by reverse genetics (rgWSN-CA/09 HA, rgWSN-CA/09 NA, and rgWSN-CA/09 HA/NA, respectively). MDCK cells and Raw264.7 macrophages were then infected at a multiplicity of infection (MOI) of 0.01, supernatants were collected at 2 and 24 h postinfection (hpi), and viral titers were measured by 50% tissue culture infectious dose (TCID₅₀) analysis. The rgWSN-CA/09 HA and rgWSN-CA/09 NA viruses had similar nuclear localization and replicated to levels similar to those of the WSN and rgWSN parental viruses in Raw264.7 macrophages (Fig. 1C and D). In contrast, the rgWSN-CA/09 HA/NA virus did not productively replicate in Raw264.7 macrophages (Fig. 1D), suggesting that both the HA and NA genes are required for WSN replication in Raw264.7 macrophages. This is not surprising given NA's role in release (14) and the required balance between the HA and NA activities that can be critical for efficient replication (51).

Productive replication requires pH of activation-mediated exit of the endosome prior to degradation in lysosomes.

HA regulates viral entry and fusion by binding to specific sialic acid motifs on the cell surface, allowing entry into endosomes, where it then is triggered at a specific pH to undergo irreversible conformational changes that cause membrane fusion prior to release of the genome and polymerase complexes into the host cell cytoplasm (15–20). To achieve fusion, a specific acidic pH of the endosome compartment must be reached (known as the pH of fusion or pH of activation). The HA activation pH is distinct to each influenza virus strain (17). Previous studies suggested that the pH of the endosome-lysosome pathway is more neutral in a macrophage than an epithelial cell (21–26). Using a pH-sensitive endocytic dye (pHrodo 10-kDa dextran) which increases in fluorescence intensity as the pH drops, we confirmed that the pH of endocytic compartments in MDCK cells was lower over time, ranging from 6.3 to 4.7, compared to Raw264.7 macrophages (Fig. 2A). Thus, we hypothesized that viruses requiring a more acidic pH of activation would be unable to exit the endosome before being targeted and degraded in lysosomes in macrophages. To test this, we measured the pH of activation of CA/09, VN1203, and WSN viruses using a syncytium assay. As previously reported, CA/09 virus fuses at pH 5.2 to 5.5 (Fig. 2B) (27, 28), while VN/1203 and WSN viruses fuse at pH 5.9 to 6.0 (Fig. 2B) (29), suggesting that in Raw264.7 macrophages, CA/09 would not encounter an optimal acidic environment with a pH of 5.0 to 5.5 until the lysosomal compartment, where the virus can encounter activated proteases that degrade the virus.

FIG 2.

(A) HA activation pH values for WSN reverse genetics viruses. Raw264.7 macrophages and MDCK cells were incubated in the presence of 10 μg/ml pHrodo green 10-kDa dextran at 37°C for 30 min. Cells were washed and fluorescence was read on a plate reader. The pH of endosome compartments was calculated compared to a standard curve generated from cells incubated in the presence of pH-adjusted buffers. (B) Influenza pH of WSN activation mutants containing CA/09 genes and parental viruses were rescued by reverse genetics. To determine the pH of activation, Vero cells were infected with the indicated viruses. After 6 hpi (VN/1203 virus) or 16 hpi (WSN and CA/09 viruses), cells were incubated in the presence of TPCK trypsin followed by incubation with pH-adjusted PBS. The pH of activation was reported as the highest pH at which fusion occurs (indicated by arrows).

To test this hypothesis, MDCK and Raw264.7 macrophages were synchronously infected with CA/09 or WSN virus. Cells were fixed at different time points and stained for nucleoprotein (NP) and the late endosome/lysosome marker lysosome-associated membrane protein 1 (LAMP1). Colocalization was determined by defining individual 3-dimensional NP surfaces using IMARIS software, although the amount of NP staining found in the nucleus of MDCK cells and WSN-infected Raw264.7 cells often made accounting for individual NP particles difficult. We then compared the fluorescence intensity of the LAMP1 channel to the fluorescence intensity of cells that were only stained with the secondary antibody. As shown in Fig. 3A, CA/09 NP particles were mainly located in the nucleus or the cytoplasm at 4 hpi in MDCK cells, and there was no difference between CA/09 and WSN LAMP1 colocalization (Fig. 3B). In contrast, the NP of CA/09 virus was colocalized with LAMP1 significantly more than WSN in Raw264.7 macrophages (Fig. 3C). From these data we conclude that CA/09 virus accumulates in lysosomes in Raw264.7 macrophages, where we hypothesize that the virus is degraded before efficient nuclear entry and subsequent replication can occur. Due to current standard operating procedures for bringing fixed samples out of our animal biosafety level 3 facility, we were unsuccessful at obtaining LAMP1 staining for VN/1203 samples.

FIG 3.

CA/09 colocalizes with LAMP1-positive compartments in Raw264.7 macrophages. MDCK cells (A) and Raw264.7 macrophages (B) were plated on coverslips. Cells were infected the next day (MOI of 2) on ice for 1 h. Cells were washed and warm medium was added (time point 0). Cells were fixed at the indicated time points and stained for NP (green) and LAMP1 (red). Nuclei were stained with DAPI. The percentage of NP colocalized with LAMP1 was determined in MDCK cells (C) and Raw264.7 macrophages (D) by using the mean fluorescence intensity of the LAMP1 channel at NP surfaces compared to the secondary-only control using Imaris software. Data are averages from 2 independent experiments (B and C). Panels A and B are representative images. Error bars indicate SD, and an asterisk represents significance (P < 0.005) as measured by two-tailed t test. The bar is 15 μm.

Finally, does the pH of activation overcome the block in nuclear entry, allowing productive replication in Raw264.7 macrophages? To test this, we determined the pH of activation of the WSN and CA/09 reverse genetics viruses and modified the pH of activation through site-directed mutagenesis. WSN, rgWSN, rgWSN-CA/09 HA, and rgWSN-CA/09 NA viruses all efficiently entered the nucleus and replicated in Raw264.7 cells and had a pH of activation of 5.9 (Fig. 2B). However, the rgWSN-CA/09 HA/NA virus, which showed decreased NP nuclear entry and replication at the limit of detection, had a pH of activation of 5.4, which is identical to that of the CA/09 and rgCA/09 viruses (Fig. 2B). We also rescued a VN/1203 virus with a mutation in the HA gene (HA2-K58I) and determined the pH of activation of this virus to be 5.5 (Fig. 4A) (29). The HA2-K58I virus showed a slight yet significant decrease in nuclear NP localization compared to the parental virus (65% versus 95%) (Fig. 4B). However, it still replicated in Raw264.7 macrophages to the same level as the parental VN/1203 virus at 24 hpi (Fig. 4C), indicating that either there is a threshold of NP nuclear entry needed to regulate replication in Raw264.7 macrophages or that there is a second cellular block downstream of nuclear entry.

FIG 4.

(A) HA activation pH value for VN/1203 HA2-K58I. Influenza pH of activation mutant rgVN/1203 HA2-K58I (HA2-K58I) and parental virus was rescued by reverse genetics. To determine the pH of activation, Vero cells were infected with the indicated viruses. After 6 hpi, cells were incubated in the presence of TPCK trypsin, followed by incubation of pH-adjusted PBS. pH of activation was reported as the highest pH at which fusion occurs (indicated by arrows). (B) Raw264.7 macrophages and MDCK cells were infected with the indicated viruses and fixed and stained for NP protein 4 hpi. The percentage of nuclear NP-positive cells was calculated and standardized to the percentage of nuclear NP in MDCK cells to account for differences in infection (percentage of MDCK cells). Viral replication in Raw264.7 macrophages was measured by TCID₅₀ analysis in the viral supernatant. (C) Results are reported as the viral titer at 2 hpi (residual virus after infection) and was subtracted from the viral titer at 24 hpi. Results are averages from 2 independent experiments. *, P < 0.05.

Raw264.7 macrophages have a second cellular block.

Although many of the avian influenza viruses tested were efficient for nuclear entry, they did not replicate in Raw264.7 macrophages (Table 1), suggesting that viruses must overcome two blocks to productively replicate in macrophages. Ramos et al. showed that altering influenza receptor usage of VN/1203 from α2,3 usage to α2,6 usage increases the virus's M gene RNA levels during infection of human dendritic cells (30). Thus, we asked if receptor usage impacted productive replication (Table 1). We saw no trend for productive replication and preferential binding to either α2,3- or α2,6-linked sialic acids. However, we did find that production of the nonstructural protein 1 (NS1) does not occur in Raw264.7 macrophages (13). To determine if influenza viruses can enter the nucleus but do not replicate NS1 protein, we infected MDCK and Raw264.7 macrophages (MOI of 2) with a panel of viruses for 20 h and stained for NS1 by immunofluorescence microscopy (Fig. 5 and Table 1). Most viruses tested that showed efficient nuclear entry of NP protein had only a small percentage of NS1⁺ cells. Although the percentage of NS1⁺ cells varied between the different influenza viruses tested, this suggests that the second block in influenza replication in Raw264.7 macrophages occurs prior to new protein translation. Importantly, this second block is also found in primary human macrophages.

FIG 5.

Viral strains that reach the nucleus are blocked at protein translation. Raw264.7 macrophages and MDCK cells were infected with the indicated viruses (MOI of 2) and incubated for 4 h (for NP staining) or 20 h (for NS1 staining). Cells were stained for NP and NS1 (green), and the percent positive (± standard deviations) cells was calculated (bottom right corner). Nuclei were stained with DAPI (blue). The bar is 50 μm. Representative images of 3 independent experiments are shown.

Influenza virus replication in primary human macrophages.

Through the use of primary human macrophages, we took a subset of our panel of viruses from Table 1 and determined viral replication and nuclear NP accumulation. Viruses that replicated in Raw264.7 macrophages also productively replicated in primary human macrophages (Fig. 6A and Table 2). However, all influenza viruses tested showed high levels of nuclear NP accumulation in primary human macrophages, even viruses that could not accumulate in the nucleus of murine macrophages, such as CA/09 virus (Fig. 6B and Table 2). Therefore, we have two phenotypes of influenza virus in macrophages: those that are blocked postnuclear entry and those that can productively replicate.

FIG 6.

Influenza virus replication in primary human macrophages. Primary human macrophages were generated after isolating monocytes from blood as described in Materials and Methods. (A) Cells were infected and viral replication was determined at 24, 48, and 72 hpi. (B) NP staining (green). Nuclei were visualized using DAPI (blue) at 4 hpi. Images in panel B are representative images. The percentage of nuclear NP-positive cells was calculated, and means ± standard deviations are shown in the bottom left corner. The bar is 50 μm. Data are averages from at least 3 independent experiments performed in at least duplicate from at least 2 donors. Error bars indicate SD.

TABLE 2.

Summary of influenza virus replication in primary human macrophages

Virus	Subtype	Replication	Receptor	Block in nuclear NP entryb
A/WSN/33	H1N1	Yes	2,6a	No (++++)
A/CA/04/09	H1N1	No	2,6a	No (+++)
A/New Caledonia/20/99	H1N1	No	2,6a	No (++++)
A/Brisbane/10/07	H3N2	No	2,6a	No (++++)
A/Tennessee/F4008/12	H3N2	No	2,6a	No (++++)
A/HK/68	H3N2	No	2,3/2,6a	No (++)
A/Wuhan/359/95	H3N2	No	2,6a	No (++)
A/Swine/NC/0668/11	H3N2	No	2,3a	No (++++)
A/Swine/TX/4119-2/98	H3N2	No	2,6a	No (++++)
A/GreyTeal/Australia/2/79	H4N4	No	2,3	No (++++)
A/HK/486/97	H5N1	No	2,3	ND
A/VN/1203/04	H5N1	Yes	2,3	No (ND)
A/HK/483/97	H5N1	Yes	2,3	No (ND)
A/Mallard/Alberta/85/76	H5N2	No	2,3	No (++++)
A/Duck/Potsdam/2216-4/84	H5N6	No	2,3	No (++++)
A/Turkey/WI/1/68	H5N9	No	2,3	No (++++)
A/Teal/HK/W312/97	H6N1	No	2,3	No (++++)
A/Anhui/01/13	H7N9	No	2,3	ND

^aDetermined by the receptor binding assay described in Materials and Methods.

^bPresented as the percent of staining normalized to the percent of nuclear staining in MDCK cells (−, <10%; +, 10 to 25%; ++, 26 to 50%; +++, 51 to 75%; ++++, >75%). ND, not determined.

We asked if the block in influenza replication in primary human cells was the same as the second block in Raw264.7 macrophages by measuring the levels of vRNA, cRNA, and mRNA in CA/09- and WSN-infected cells as described previously (13). Briefly, A549 human lung carcinoma cells and primary human macrophages were infected with either CA/09 or WSN (MOI of 5) on ice for 1 h, total RNA was isolated from cells at 30 min, 4 h, and 12 h postinfection, and the levels of vRNA, cRNA, and mRNA of the NP gene were amplified by real-time reverse transcription-PCR (RT-PCR). As shown in Fig. 7A and previously (13), all three RNA species for CA/09 virus could be readily detected in A549 cells. In contrast to Raw264.7 macrophages, where very little CA/09 RNA is detected (13), all three RNA species could also be detected in primary human macrophages, and vRNA levels were comparable to WSN vRNA levels (Fig. 7B). Surprisingly, CA/09 cRNA and mRNA levels were higher than those in WSN-infected cells in spite of the lack of productive replication (Fig. 7B). Therefore, unlike in Raw264.7 macrophages, RNA synthesis occurs in CA/09 virus-infected human macrophages, indicating that the postnuclear entry block differs between human and Raw264.7 macrophages.

FIG 7.

CA/09 infection in primary human macrophages produces RNA synthesis and protein translation. A549 cells (A) and primary human macrophages (B and C) were infected with CA/09 or WSN. (A and B) Total RNA was extracted and vRNA, cRNA, and mRNA species were determine as described in Materials and Methods. Data are standardized to GAPDH levels and input levels at 30 min postinfection. WSN levels for the 12-h time point were set to 1. (C) Cells were fixed 20 h postinfection and stained for NS1 protein (green). Nuclei were stained with DAPI (blue). The percent NS1-positive cells was calculated and is indicated in the panels. The bar is 50 μm. Data are averages from 2 independent experiments from 2 different donors.

Finally, we asked if translation was blocked in CA/09-infected human macrophages by performing immunofluorescence staining for NS1 protein 20 hpi. Unlike Raw264.7 macrophages (Fig. 5 and reference 13), over 50% of CA/09-infected cells stained positive for NS1 (Fig. 7C). Therefore, the block in CA/09 replication occurs after protein translation. However, we did note that the NS1 staining patterns varied between the two viruses. NS1 was only observed in the nucleus of CA/09-infected cells, while WSN-infected cells had NS1 in the nucleus and the cytoplasm. These data suggest that protein trafficking is affected in CA/09-infected cells, which could lead to the block in replication, but further studies are needed to define this cellular block.

Influenza virus replication affects macrophage function.

Macrophages infected with H5N1 virus have increased cytokine production (12, 31) and decreased cell death compared to H1N1-infected macrophages (32). However, no studies to date have examined the impact of productive or abortive infection at different stages of the viral life cycle on macrophage function. Therefore, we infected Raw264.7 macrophages and primary human macrophages with CA/09, WSN, and VN/1203 viruses at an MOI of 1 and collected cell supernatants 24 hpi to examine the impact on cytotoxicity and nitrite production. Although levels were above those of the uninfected controls, none of the viruses induced significantly different levels of cytotoxicity in Raw264.7 macrophages (Fig. 8A). In contrast, there was no increase in cytotoxicity in primary human macrophages (Fig. 8A). Additionally, we found no significant differences in nitrite levels (Fig. 8B) or production of reactive oxygen species (ROS) in either Raw264.7 or primary human macrophages between influenza viruses that could or could not replicate (Fig. 8C). ROS production was monitored by measuring increases in the fluorescence intensity of the fluorophore DCF, which only fluoresces after becoming oxidized by ROS. Ethanol-treated cells, which served as a positive control, showed a significant increase in ROS production compared to uninfected cells (Fig. 8C).

FIG 8.

Influenza virus replication in macrophages affects macrophage phagocytosis. Raw264.7 macrophages and primary human macrophages were infected with the indicated viruses (MOI of 1). At 24 hpi, the amount of cytotoxicity (A), nitrite levels (B), and ROS production (C) were determined as described in Materials and Methods. T:0, time point 0; EtOH, ethanol. (D) Phagocytosis of macrophages was measured by the percentage of FITC-positive cells (indicating IgG-opsonized pHrodo-S. aureus bioparticle uptake) measured by flow cytometry. FcγRIII/FcγRII cell surface levels in Raw264.7 macrophages (E and G) and primary human macrophages (F and H) were measured by flow cytometry at 24 hpi. Std., standardized. Data in panels D, E, F, G, and H are presented as the percentage of uninfected cells, which was set at 100%. Error bars indicate SD. *, P < 0.05; **, P < 0.005.

Finally, we asked if productive or abortive viral replication impacted phagocytic ability, as phagocytosis of dead cells and pathogens is an essential macrophage function (1). To quantify phagocytosis, Raw264.7 macrophages and primary human macrophages were infected with CA/09 or WSN viruses (MOI of 1) and then incubated with IgG-opsonized pHrodo-Staphylococcus aureus bioparticles to monitor phagocytosis 24 hpi. These particles are pH sensitive and only fluoresce after encountering an acidic pH. Briefly, bioparticles were added to the previously infected or mock-infected cells on ice and shifted to 37°C for 30 min before determining the percentage of FITC-positive macrophages by flow cytometry. WSN virus, which can replicate in macrophages, significantly decreases the percentage of cells that phagocytose the bioparticles compared to uninfected cells (Fig. 8D). In contrast, CA/09 virus did not alter macrophage phagocytosis ability compared to uninfected cells (Fig. 8D). We then asked if Fc receptor cell surface levels were altered in influenza-infected macrophages. Uninfected and CA/09- or WSN-infected macrophages were fixed and stained for cell surface levels of the FcγRIII and FcγRII receptors CD16 and CD32, respectively, 24 hpi. The percentage of FcγRIII/FcγRII-positive cells was decreased in WSN-infected macrophages compared to uninfected cells in both Raw264.7 and primary macrophages (Fig. 8E and F). Additionally, the mean fluorescence intensity (MFI) of FcγRIII/FcγRII staining of positive cells was decreased during WSN infection, indicating decreased surface levels of FcγRIII/FcγRII (Fig. 8G and H). Similar results were obtained with the HPAI H5N1 viruses HK/483 and VN/1203 that productively replicate in macrophages (Fig. 8E and G). In contrast, infection with CA/09 virus or the HPAI H5N1 viruses HK/483 and VN/1203, which do not productively replicate, does not decrease the percentage of or the cell surface levels of FcγRIII/FcγRII compared to those of uninfected cells (Fig. 8E to H). These data demonstrate that productive influenza virus replication decreases macrophage phagocytosis, presumably by decreasing FcR cell surface levels.

DISCUSSION

In this study, we expand on our previous findings that only a subset of HPAI H5 influenza viruses can replicate in murine macrophages (13) and demonstrate that WSN also undergoes productive replication. In the murine macrophage cell line Raw264.7, viruses must overcome two cellular blocks in order to productively replicate, with the first requiring a specific pH of activation in order to exit the endosome prior to lysosomal degradation. The second cellular block appears to be downstream of nuclear exit and pretranslation (Fig. 9), although future work is needed to define the precise mechanism. Overcoming this second block may also be important for productive replication in primary human blood-derived macrophages. All strains tested could get to the nucleus, suggesting first that the pH of activation-mediated block is specific to murine macrophages. It will be interesting to determine if this block also occurs in nonhuman mammals and birds. Finally, productive replication was associated with decreased surface Fc receptor levels and phagocytosis of opsonized beads, highlighting the impact that this could have on viral pathogenesis.

FIG 9.

Model for influenza virus replication in Raw264.7 and primary human macrophages. (1) In Raw264.7 macrophages, influenza viruses enter the cell (13) and either degrade in the lysosome or enter the nucleus. (2) A second block occurs prior to protein translation. (3) Select viruses can productively replicate. All viruses enter the nucleus in human macrophages, but most viruses encounter a block after nuclear entry despite productive translation of viral protein.

Expressing the HPAI H5 HA on nonreplicating CA/09 virus allows it to overcome both cellular blocks (13). Unfortunately, we were unable to ask if CA/09-WSN HA virus would increase the pH of activation to that of parental WSN, supporting productive viral replication, or if both the HA and NA genes would be needed. Expressing only the CA/09 HA or the CA/09 NA genes on WSN virus was insufficient to decrease pH of activation and inhibit productive replication. We had to express both CA/09 HA and NA on WSN to have the pH of activation become more acidic, like that of CA/09 virus, decreasing nuclear NP entry and productive replication. These data show that the WSN HA and NA genes are necessary for efficient nuclear entry in Raw264.7 macrophages, highlighting an understudied area of research: how NA contributes to influenza virus-mediated membrane fusion. Su et al. previously reported that NA contributes to HA fusion by measuring fusion levels after transfecting cells with either HA alone or HA and NA genes and assessing cell-cell fusion (33). Higher levels of fusion were reported when NA was transfected along with HA than with HA alone (33). While NA-promoted HA activation has been linked to NA enzymatic activity for H5N1 influenza virus (34), its mechanism remains unknown.

In complementary experiments, we introduced a point mutation in the VN/1203 HA that changed the pH of activation to be more acidic (29), significantly decreasing the percentage of nuclear NP. However, this virus replicated to wild-type VN/1203 levels. Further investigation identified many viruses, predominantly avian strains, which could enter the nucleus of Raw264.7 macrophages but did not productively replicate, leading to the identification of the second cellular block. Most of these viruses had very low percentages of cells that expressed NS1 protein levels, suggesting a block pretranslation. Intriguingly, primary human blood-derived macrophages only have this second cellular block. Quantifying the RNA species levels and NS1 protein production between CA/09 (nonreplicating) and WSN (replicating) viruses in primary human cells demonstrated that RNA replication and NS1 protein expression were not blocked in CA/09-infected primary human cells, yet they failed to produce infectious progeny virus. Londrigan et al. demonstrated that newly synthesized HA and NA proteins are incorporated into the plasma membrane of mouse peritoneal exudate cell macrophages infected with the H1N1 reverse genetics virus rgA/PR8/34-A/Brazil/11/78 HA, but this virus does not productively replicate (35). Future studies will determine where the cellular block occurs, if it can be overcome by individual viral genes, and importantly, if other nonhuman mammalian or avian macrophages have one or two cellular blocks.

Finally, we asked if replication, whether productive or abortive early or late in the viral life cycle, impacted macrophage function. Mok et al. reported that infection of macrophages with an H5N1 virus showed decreased cell death compared to H1N1 infection by measuring nuclear condensation or nuclear fragmentation by DAPI staining (36). However, we found no difference in cytotoxicity, as measured by lactate dehydrogenase (LDH) levels, between replicating and nonreplicating viruses in either Raw264.7 or primary cells. Similarly, we also found that nitric oxide and ROS production were unchanged after influenza infection.

We did, however, find that WSN infection significantly decreased phagocytosis of IgG-opsonized bioparticles as well as cell surface receptor levels of the Fc receptors CD16 and CD32. This is the first report to our knowledge of an RNA virus infection decreasing the level of an Fc receptor. Previous reports demonstrated that bacterial infections downregulate Fc receptors, leading to reduced phagocytosis ability. Infection of peripheral blood mononuclear cells (PBMCs) with Escherichia coli reduced phagocytic ability of the PBMCs 2 dpi, with both mRNA levels and mean fluorescence intensity (MFI) of CD16 and CD32 protein levels decreased (37). How WSN decreases Fc receptors is a topic of future studies. Influenza proteins in the cellular membrane may mask Fc receptor epitopes, which has been observed during the transfection of the Ebola virus glycoprotein and staining for major histocompatibility complex class I (MHC-I) and CD29 on the cell surface by flow cytometry (38). Future studies will first determine if cell surface WSN proteins mask Fc receptor epitopes or if WSN infection downregulates Fc cell surface levels and, if applicable, how WSN downregulates Fc cell surface levels. WSN infection may decrease Fc receptors at the transcriptional level, such as that for E. coli. Alternatively, Fc receptors are endocytosed where they can be degraded or recycled (39–41). Future studies will also determine if WSN infection of macrophages alters Fc receptor total protein levels, if other Fc receptors are affected, and whether decreased phagocytic ability affects the host response during infection in vivo, such as clearance of bacteria during influenza infection.

In summary, we demonstrate that certain influenza strains overcome two blocks in Raw264.7 macrophages and one block in primary human macrophages from PBMCs to productively replicate. The first block in Raw264.7 macrophages is mediated by the viral strain's pH of activation, and the second block occurs prior to protein translation. In human macrophages, the block is after protein translation. Replication in either macrophage type decreased the phagocytic ability of the cell by decreasing cell surface levels of Fc receptors. Future studies will further define the blocks in replication, how specific strains overcome these blocks, and how replication-competent viruses decrease Fc receptor surface levels. Our studies highlight the complicated interplay between macrophages and influenza virus and may explain some of the confusions in the field. It is likely that different viral strains as well as distinct macrophage types will yield different results. For example, Yu et al. showed that H1N1 and H5N1 productively replicate in human alveolar macrophages (12), while van Riel et al. did not find productive replication but used different viral strains for the infections (8). van Riel et al. observed replication of an H3N2 strain in monocyte-derived macrophages, whereas we did not observe replication in the 4 H3N2 strains that we tested, suggesting strain-specific variations (8). Much more work is needed to understand the impact of this interplay on viral pathogenesis.

MATERIALS AND METHODS

Laboratory facilities.

All experiments using HPAI H5N1 and H7N9 viruses were conducted in a biosafety level 3 enhanced-containment laboratory (52). Investigators were required to wear appropriate respirator equipment (RACAL Health and Safety, Inc., Frederick, MD). All other viruses were used under enhanced biosafety level 2 conditions by vaccinated personnel.

Cells.

Madin-Darby canine kidney (MDCK) (American Type Culture Collection [ATCC] CCL-34; Manassas, VA) and human lung A549 (ATCC CCL-185) cells were cultured in minimum essential medium (MEM; Corning) supplemented with 1 mM sodium pyruvate (Gibco), GlutaMAX-I (Gibco), and 10% fetal bovine serum (FBS; Benchmark). The Raw264.7 murine macrophage cell line (ATCC TIB-71) was cultured in RPMI 1640 medium (Gibco) supplemented with 1 mM sodium pyruvate, GlutaMAX-I, and 10% FBS. African green monkey kidney Vero (ATCC CCL-81) cells were cultured in Dulbecco's modified eagles medium (DMEM; Gibco) supplemented with 1% penicillin-streptomycin (Gibco) and 5% FBS (HyClone). All cells were cultured at 37°C under 5% CO₂.

Primary human macrophage isolation and culture.

Primary human macrophages were cultured from monocytes by adherence in the presence of human serum for 7 days prior to infection (8, 42). Peripheral blood mononuclear cells (PBMCs) were isolated by layering whole blood from healthy donors on Ficoll-Paque Plus (GE Healthcare) and centrifuging at 1,000 × g for 20 min at room temperature. PBMCs were pelleted by centrifugation at 320 × g for 10 min at 4°C. Plasma above the PBMC layer was removed and saved for culturing of macrophages. PMBCs were collected and washed once with PBS and pelleted again by centrifugation. Lymphocytes were removed by pelleting through a 50% Percoll cushion (GE Healthcare) by centrifugation at 730 × g for 20 min at room temperature. Lymphocytes pellet to the bottom while the remaining cells are at the medium-Percoll interface, which contains monocytes. Monocytes were washed twice in PBS and pelleted by centrifugation. Monocytes were plated in Iscove's DMEM (Corning) supplemented with 1 mM sodium pyruvate, GlutaMAX-I, penicillin-streptomycin (Gibco), and 10% human plasma. Nonadherent cells were washed away after overnight incubation. Cells were incubated for 7 days prior to infection with washing and medium changes every 2 to 3 days (8, 42). Cells were scraped and counted the day of the experiment to determine average cell numbers per well for MOI calculations. Cells from at least 2 different donors were used for experimental repeats.

Viruses.

The influenza H1N1 virus strain A/Tennessee/F1052/10, the H3 A/Swine/NC/0668/11, A/Swine/TX/4119-2/98, and A/Tennessee/F4008/12 viruses, and the H5 viruses A/Hong Kong/483/97, A/Hong Kong/486/97, and A/Duck/Potsdam/2216-4/84 viruses were all propagated in MDCK cells as described previously (43, 44).

Influenza viruses A/Puerto Rico/8/34, A/WSN/33, A/New Caledonia/20/99, A/California/04/09 (CA/09), and A/Swine/NE/2/92 (H1N1); A/Brisbane/10/07, A/Hong Kong/68, and A/Wuhan/359/95 (H3N2); A/Gray Teal/Australia/2/79 (H4N4); A/Vietnam/1203/04, A/Hong Kong/213/03, and A/Duck/Bangladesh/19097/13 (H5N1); A/Mallard/Alberta/85/76 (H5N2); A/Turkey/WI/1/68 (H5N9); A/Teal/Hong Kong/W312/97 (H6N1); and A/Anhui/01/13 (H7N9) were propagated in 10-day-old specific-pathogen-free embryonated chicken eggs at 37°C. Allantoic fluid was harvested, clarified by centrifugation, and stored at −70°C.

Reverse genetics.

The A/WSN/33 virus containing the HA, neuraminidase (NA), or HA and NA genes from A/California/04/09 and corresponding parental A/WSN/33 and A/California/04/09 viruses were generated using the eight-plasmid system as described previously (45). Briefly, all recombinant viruses were generated using pHW2000 plasmids, each containing an individual gene of the corresponding viruses. Plasmids were transfected into cocultured MDCK and 293T cells using Lipofectamine 2000 (Thermo Fisher Scientific). Viruses were propagated in 10-day-old embryonated chicken eggs, and titers were determined by 50% tissue culture infectious dose (TCID₅₀) analysis in MDCK cells.

The reverse genetics (rg) virus rgA/VN/1203/04 HA2-K58I and corresponding parental A/VN/1203/04 viruses have been described previously (29). The HA2-K58I point mutation was introduced into pHW2000-A/VN/1203/04-HA plasmid using a QuikChange site-directed mutagenesis kit (Stratagene) according to the manufacturer's instructions. Viruses were plaque purified on MDCK cells. Virus identity was confirmed by full-genome sequencing at the Hartwell Center for Bioinformatics and Biotechnology at St. Jude Children's Research Hospital.

Receptor binding assay.

Receptor affinity was determined using a solid-phase direct virus binding assay as previously described (46, 47). Briefly, 128 HA units of influenza viruses in 50 μl 0.02 M Tris-buffered saline (TBS) was bound to fetuin-coated plates at 4°C overnight. Plates were blocked in 0.1% bovine serum albumin (BSA) treated with 5% neuraminidase (BSA-NA) in PBS for 1 h at 4°C. Biotinylated glycans (α2,3′SL or α2,6′SL; Glycotech Corporation, Gaithersburg, MD) were added to influenza virus-coated plates at various dilutions in 0.2% Tween 80, 0.1% BSA-NA containing 1 μM oseltamivir carboxylate (Toronto Research Chemicals) in PBS and incubated for 2 h at 4°C. Glycan binding was analyzed using horseradish peroxidase (HRP)-conjugated streptavidin (Invitrogen, Carlsbad, CA) followed by TMP substrate (Sigma, St. Louis, MO). The reaction was stopped with 2N sulfuric acid, and plates were read at 450 nm on a Thermo Labsystems Multiskan Ascent plate reader.

In vitro infections.

MDCK cells, Raw264.7 macrophages, or primary human macrophages were infected at least in duplicate at the indicated multiplicity of infection (MOI) for 1 h at 37°C. Unbound virus was removed, and the cells were washed in PBS and maintained in the appropriate medium containing 0.075% BSA in the presence (non-H5 viruses and low-pathogenicity H5 viruses) or absence (HPAI H5 viruses) of 1 μg/ml of tosylsulfonyl phenylalanyl chloromethyl ketone (TPCK)-treated trypsin (Pierce, Rockford, IL). Cell culture medium was removed at the indicated times and stored at −80°C for the determination of virus titers by TCID₅₀ on MDCK cells.

Immunofluorescence microscopy.

Raw264.7 macrophages, primary human macrophages, or MDCK cells were seeded on 96-well plates or sterile glass coverslips 1 (Raw264.7) or 7 (primary human) days prior to inoculation. Cells were inoculated with medium alone or the indicated viruses (MOI of 2) for 1 h at 4°C and then washed with cold PBS to remove unbound virus. The cells were shifted to 37°C for 30 min or 1, 2, or 4 h and then fixed with 4% paraformaldehyde. To determine nucleoprotein (NP) nuclear localization, cells were permeabilized in 0.5% Triton X-100 in PBS for 10 min, blocked in 10% FBS in PBS for 1 h, and stained for NP (ATCC clone HB-65) overnight. Anti-mouse IgG-Alexa 488 secondary antibody (Invitrogen) was used at 1:200 in 10% FBS in PBS for 1 h. For NP-LAMP1 colocalization, cells were blocked for 1 h in PBS with 10% FBS and 0.5% saponin (Sigma). Staining for NP and LAMP1 (1:1,000; Sigma) was done in 10% FBS with 0.5% saponin overnight. Anti-mouse IgG-Alexa 488 and anti-rabbit IgG-Alexa 555 (Invitrogen) were both used at 1:200 in PBS with 10% FBS and 0.5% saponin for 1 h. All cells were stained for DNA with 4′,6′-diamidino-2-phenylindole (DAPI; 1:1,000; Sigma). All steps were performed at room temperature with the exception of overnight incubations, which were performed at 4°C. Wells were imaged on an Evos microscope (Advanced Microscopy Group) at a magnification of ×40, using identical parameters for each treatment. Nuclei and fluorescein isothiocyanate-positive (FITC⁺) cells were counted using ImageJ software. The percentage of nuclear NP is expressed as the percentage of nuclear NP in macrophages standardized to the percent nuclear NP in MDCK cells from the same experiment: percent nuclear NP⁺ macrophage divided by percent nuclear NP⁺ MDCK. For NP-LAMP1 colocalization experiments, coverslips were mounted in ProLong Gold antifade reagent (Molecular Probes, Eugene, OR), and fluorescence was examined on a Nikon TE2000 E2 microscope equipped with a Nikon C1Si confocal scanhead. Z-stack images were acquired with a Nikon ×60 1.45 Apochromat objective lens using Nikon EZC1 software using identical parameters. NP three-dimensional surfaces were acquired using IMARIS software. Colocalization was determined by the fluorescence intensity of the LAMP1 channel at each three-dimensional surface compared to the fluorescence intensity of the secondary-only antibody intensity.

Intracellular pH.

Raw264.7 macrophages and MDCK cells were plated in black 96-well plates (3 × 10⁴ cells/well). The next day, cells were washed twice with PBS and incubated in the presence of 50 μl of phenol red-free DMEM containing 10 μg/ml pHrodo green dextran conjugate (Thermo Fisher Scientific) for 30 min at 37°C. Cells were washed twice with PBS, and phenol red-free DMEM containing 20 μM HEPES was added to each experimental well. A pH standard curve was generated by washing cells twice with PBS after 30 min of incubation, adding 100 μl of pH-adjusted buffers containing 10 μM nigericin (Invitrogen) and 1 μM monensin (BioLegend), and then incubating at 37°C for 10 min. Fluorescence for all cells was read on a BioTek plate reader using excitation/emission wavelengths of 495 nm/538 nm for 20 min at 1 read per min. To calculate the pH of endosomal compartments for each cell type at each time point, a standard curve was generated using the fluorescence values of the pH calibration wells, and fluorescence of each experimental time point was calculated using this standard curve.

pH of activation assay.

HA activation pH values were measured by syncytium assay as described previously (48). Vero cells were infected at an MOI of 3 for 1 h. Sixteen hours postinfection (hpi), HA-expressing cells were incubated for 15 min with 5 μg/ml TPCK-treated trypsin (Worthington Biochemical, NJ) and treated with pH-adjusted PBS buffers for 5 min at 37°C. Syncytia were then allowed to form in MEM containing 5% FBS for 2 h at 37°C before fixation and staining for microscopy with a Protocol Hema 3 kit (Fisher Scientific). The pH of activation was reported as the highest pH value at which syncytia were observed.

Quantification of vRNA, cRNA, and mRNA levels by real-time RT-PCR.

Total RNA was isolated from primary human macrophages and A549 cells at the indicated time points by TRIzol extraction (Ambion) according to the manufacturer's instructions. cDNAs complementary to the three species of viral RNA were synthesized using primers specific to the viral NP gene containing a nucleotide tag that is unrelated to the viral sequence, as described previously (13, 49), using 200 ng of total RNA and the Superscript III first-strand synthesis system (Invitrogen) according to the manufacturer's instructions.

Real-time PCR was performed using Platinum SYBR green quantitative PCR (qPCR) Supermix-UDG with ROX (Invitrogen) on a CFX96 real-time system (Bio-Rad). cDNA (5 μl) was added to 10 μl Platinum SYBR green qPCR supermix, 1.5 μl forward primer (10 μM), 1.5 μl reverse primer (10 μM), and 2 μl water. All primers have been described previously (13, 49). Human glyceraldehyde-3-phosphate dehydrogenase (GAPDH) levels were also measured on 100 ng RNA using TaqMan GAPDH control reagents (human; Thermo Fisher Scientific). Viral genes were normalized to GAPDH levels, and results are presented as fold change over the value at 30 min.

Cytotoxicity.

Cells were infected at an MOI of 1 as described above and cell supernatants harvested 24 hpi. Cytotoxicity was determined on supernatant (100 μl) using the cytotoxicity detection kit plus (LDH; Roche) according to the manufacturer's instructions. The percent cytotoxicity was calculated by subtracting the value of the low control (medium) and dividing by the difference between the high control (lysed cells) and the low control: percent cytotoxicity = [(experimental value − low control)/(high control − low control)] × 100.

Nitrite assay.

Raw264.7 (2 × 10⁵/well) or primary human (5 × 10⁵/well) macrophages plated 1 or 7 days prior to infection, respectively, in a 24-well plate were infected with the indicated viruses at an MOI of 1 in phenol red-free RPMI. Supernatants were collected 24 hpi and stored at −70°C until use. Cell-free supernatants (50 μl) in a 96-well plate were incubated with 50 μl 1% sulfanilamide (in 5% phosphoric acid) and 50 μl 0.1% N-1-napthylethylenediamine dihydrochloride for 15 min in the dark, and then absorbance was read at 550 nm using on a BioTek spectrophotometer. The concentration of nitrite of each sample was calculated using a standard curve of sodium nitrite.

Reactive oxygen species production.

Raw264.7 (5 × 10⁴/well) or primary human (1 × 10⁵/well) macrophages plated 1 or 7 days prior to infection, respectively, in black 96-well plates (Costar) were incubated with 10 μM the ROS-sensitive fluorophore H2DCFDA (DCF; Invitrogen) for 30 min, followed by inoculation with medium, virus (MOI of 5), or 10 mM ethanol. Fluorescence intensity was measured over time at an excitation wavelength of 485 nm and an emission wavelength of 520 nm on a fluorescent plate reader (BioTek). Data are expressed as the relative fluorescence intensity (RFI) as standardized to the time zero read for that individual well.

Phagocytosis assay.

Phagocytic activity was determined using IgG-opsonized bioparticles as described previously (50). Briefly, macrophages were seeded into 6-well plates 1 (Raw264.7 at 5 × 10⁵ cells/well) or 7 (primary human at 1.5 × 10⁶ cells/well) days prior and then infected with the indicated viruses at an MOI of 1. Twenty-four hpi, cells were moved to ice, and medium was removed and incubated with 60 μl of pHrodo-S. aureus bioparticles (Invitrogen) (preincubated with 1:100 rabbit IgG [Santa Cruz Biotechnology] for 1 h at 37°C) on ice or at 37°C for 30 min. Cells were then kept on ice, washed with ice-cold PBS, scraped, and then fixed in 2% paraformaldehyde for 10 min. FITC-positive cells were measured by flow cytometry (Canto-II). Data are presented as the percentage of FITC-positive cells compared to the percentage of FITC-positive uninfected cells (percentage of uninfected cells).

Determination of FcR cell surface levels.

Raw264.7 macrophages were plated at 5 × 10⁵ cells/well in 6-well plates the day prior to infection and then infected at an MOI of 1. Twenty-four hpi, cells were scraped, fixed in 2% paraformaldehyde for 10 min, washed in PBS, and pelleted by centrifugation before blocking with 10% FBS in PBS for 1 h. Cells were stained with rat anti-mouse CD16/CD32 (1:200 dilution; BD Biosciences) for 1 h at room temperature and then washed twice and stained with a 1:200 dilution of goat anti-rat IgG (H+L), Alexa Fluor 555 conjugate (Invitrogen) for 1 h at room temperature. Cells were washed twice and suspended in PBS plus 5% BSA.

Primary cells (1.5 × 10⁶ plated 7 days prior to infection) infected at an MOI of 1 were fixed 24 hpi in methanol at 4°C for 20 min, followed by 2 washes with PBS. They were then scraped in 10% FBS in PBS and blocked for 1 h at room temperature before staining with mouse anti-human CD16 and CD32 (1:20 dilutions; Abcam) for 1 h at room temperature. After washing twice in PBS, cells were stained with goat anti-mouse IgG (H+L) Alexa Fluor 488 conjugate (1:200, Invitrogen) for 1 h at room temperature and then washed again and resuspended in 5% BSA. Fluorescence was quantitated on a BD FACSCanto II or FACSCalibur flow cytometer, and the percentage of CD16/CD32-positive cells and the mean fluorescence intensity (MFI) were determined using FlowJo software. The percentage of CD16/CD32-positive cells and MFI of CD16/CD32 was standardized to that of uninfected cells.

Statistical analysis.

Statistical significance was determined using a 2-tailed Student t test in Microsoft Excel. All assays were run at least in duplicate, and data are representative of the averages from at least two independent experiments. Error bars represent standard deviations, and statistical significance was defined as having a P value of <0.05.