Aging Impairs Alveolar Macrophage Phagocytosis and Increases Influenza-Induced Mortality in Mice

Older people exhibit increased morbidity and mortality in response to respiratory viral infections (1, 2). As the number of older adults increases, the incidence of morbid complications from respiratory infections also grows. Indeed, during the years 1988–2002, people >65 y of age exhibit a 20% increase in hospitalization rates for community acquired pneumonia (3). Hence, discerning how aging modifies immunity to influenza virus on a molecular basis is critical to developing novel therapies and treatments that will protect older adults from respiratory infection.

Experimental studies have documented several age-associated alterations during influenza viral infection that include reduced antiviral CD8⁺ T cell responses, decreased NK cell function, altered PG production, increased expansion of regulatory T cells, and reduced inflammasome activity (4–8). Despite these reports, there remains controversy as to the impact of aging on the susceptibility to influenza viral infections in murine models. On one hand, prior reports indicate that aged mice (age range 18–26 mo) exhibit higher mortality, morbidity, and lower viral clearance than do young mice during influenza viral lung infection (4, 6, 7, 9). On the other hand, a recent study documented that aged (24–26 mo old) mice infected with influenza virus exhibit significantly lower mortality and higher viral clearance than do young mice (10). Additionally, the critical cellular mechanisms by which aging alters the host response to influenza viral lung infection remains unclear.

In this study, we used dose titrating aliquots of influenza virus to evaluate how aging affects mortality and lung damage during influenza viral infection. We identified alveolar macrophages (AM), key tissue resident macrophages for lung homeostasis (11), as critical for mortality during influenza viral lung infection with aging. The transcriptional signatures of AM from young and advanced aged mice were substantially different: we found that 3545 genes were significantly altered with aging and that cell cycle pathways were markedly downregulated. Prior to and during infection, advanced aged mice exhibited 2-fold lower concentrations of AM. Functionally, AM from advanced aged mice were impaired in scavenging apoptotic neutrophils, displayed selective downregulation of a key scavenging receptor, CD204, and exhibited defects in limiting lung damage. Therefore, our study has found that aging impairs the intrinsic function of AM to limit lung damage during influenza viral lung infection.

C57BL/6 mice of 2–4 mo (representative weight, 26 ± 2 g [SD]), 16 mo (36 ± 3.7 g), and 22–24 mo of age (33 ± 2.5 g) were obtained from the National Institute of Aging rodent facility. Mice were infected with purified human influenza virus A/Puerto Rico/8/34 (H1N1) (PR8) (Advanced Biotechnologies, Eldersburg, MD) as previously reported (12). Mice were anesthetized with isoflurane and instilled intranasally (i.n.) with 50 μl of PBS containing the indicated dose of PR8 virus or 50 μl of PBS control. Following infection, mice were monitored daily for changes in weight, clinical scores, and mortality. Clinical scores were determined by ruffled fur, activity, hunched back, and mortality (Table I), specifically mice were euthanized when 30% of their original weight was lost, which was recorded as death in survival experiments. No animals were used in the study when they displayed evidence of infection or other illnesses prior to PR8 infection. Both the Yale University and University of Michigan Institutional Animal Care and Use Committees approved the use of animals in this study. Prior to viral infection, all mice were kept in pathogen-free conditions.

Tissues were harvested from PBS-instilled or PR8-infected mice (n = 3–5 per time point for each experiment) at 1, 3, 6, and 9 d postinfection (DPI). Each time point was repeated at least three times independently. After euthanasia, the bronchoalveolar lavage (BAL) was obtained by washing the lungs twice with 1 ml of cold sterile 1× PBS. To harvest lungs, the chest cavity was opened and the lungs were removed and flash frozen in liquid nitrogen.

Lung injury was assessed by measuring total protein and lactate dehydrogenase (LDH) levels in the BAL fluid. Total protein in BAL was measured using the BCA protein assay kit (Thermo Fisher Scientific, Waltham, MA) following the manufacturer’s instructions. LDH activity was measured with an LDH assay kit (Roche, New York, NY) according to the manufacturer’s instructions. Myeloperoxidase (MPO) levels in the BAL were measured with kits from (eBioscience, San Diego, CA), according to the manufacturer’s instructions.

Lung viral titers were determined by plaque assay using Madin–Darby canine kidney cells (provided by Dr. A. Iwasaki, Yale School of Medicine) as previously reported (12). Madin–Darby canine kidney cells were cultured in six-well plates overnight until the cell monolayer reached 100% confluence. Lung supernatants were thawed at room temperature and diluted to different dilutions (10¹–10³) with 1% BSA in 1× PBS. Cells were washed twice with 1× PBS followed by 1 h incubation with 200 μl of lung supernatants at 37°C with frequent shaking. Cells were washed twice with 1× PBS to remove excess virus and were overlaid with 2 ml of agarose mixture containing 50% 2× MEM, 17.35% double distilled H₂O, 1% DEAE-dextran, 7.5% NaHCO₃, 0.35% acetylated trypsin (1 mg/ml), and 2% agarose. Plates were incubated at 37°C for 48 h. The agarose was removed from the plates and the plates stained with 0.1% crystal violet. Virus plaques were counted and the PFU per milliliter were calculated using the formula: number of plaques × dilution factor × 5. All dilutions of each sample were run in duplicate.

RNA from whole-tissue samples was extracted as described below. During reverse transcription, RNA was transcribed with a PAN influenza primer (5′-TCTAACCGAGGTCGAAACGTA-3′). RT-PCR was run as described below using the influenza-specific primers (5′-AAGACCAATCCTGTCACCTCTGA-3′, 5′-CCTGACGTCGCATCTGCGAAAC-3′). For other genes measured by RT-PCR the following primers were used: c-Maf, 5′-ACTGAACCGCAGCTGCGCGGGGTCAG-3′, 5′-CTTCTCGTATTTCTCCTTGTAGGCGTCC-3′; and MafB, 5′-TCCACCTCTTGCTACGTGTG-3′, 5′-CGTTAGTTGCCAATGTGTGG-3′.

For each group, data were normalized to β-actin expression (5′-CCGCCCTAGGCACCAGGGTG-3′, 5′-CCGCCCTAGGCACCAGGGTG-3′) and then compared among the groups. Data were analyzed using the Bio-Rad Laboratories (Hercules CA) built-in system.

To obtain single-cell suspensions, lungs were harvested from animals, minced, and then digested with 1 mg/ml collagenase D (Roche) and 100 U/ml DNAse (DN25; Sigma-Aldrich, St. Louis, MO) in PBS without calcium and magnesium for 45 min at 37°C. After digestion, lung tissue was disrupted into single-cell suspension by passage through a 100-μm sieve (Fisherbrand; Thermo Fisher Scientific). To remove RBCs, cells were spun for 5 min at 1400 rpm and RBCs were lysed by resuspension of the pellets for 3 min on ice in RBC lysis buffer (BioLegend, San Diego, CA). Lysed cells were spun and resuspended with 1:50 FcγR blocker (BioLegend) in HBSS containing 1% BSA and 2 mM EDTA. Cells were then incubated on ice for 10 min and stained with the indicated fluorescent Abs for 30 min on ice. Cells in the BAL were collected, spun down, and blocked 1:50 with FcγR blocker on ice for 10 min followed by staining with the indicated fluorescent Abs. For all infected samples, cells were fixed with fixation buffer (BD Biosciences, San Diego, CA) for 15 min on ice after staining with indicated fluorescent Abs. AM were characterized according to their forward and side scatter profiles, and by high surface expression of Siglec-F, F4/80, and CD11c as previously reported (13, 14). The Siglec-F Ab was purchased from BD Biosciences. All other Abs were purchased from eBioscience. Flow cytometry data were acquired on an LSR II (BD Biosciences) and analyzed with FlowJo (Ashland, OR) software.

To assess the AM proliferation rate, single-cell suspensions were prepared and stained as described above followed by intracellular Ki67 (allophycocyanin) (eBioscience) staining according to the manufacturer’s protocol.

For sorting of AM, single-cell suspensions were prepared as above from lung homogenates and stained with CD45 (allophycocyanin-Cy7), CD11c (FITC), F4/80 (Pacific Blue) and Siglec-F (PE). Cells were sorted in PBS containing 1% BSA and 2 mM EDTA on a FACSAria (BD Biosciences) and collected into HBSS supplemented with 1% BSA and 2 mM EDTA for adoptive transfer or TRIzol (Invitrogen, Carlsbad, CA) for RNA extraction (see below).

FACS-sorted AM were centrifuged at 5000 rpm for 5 min, resuspended in PBS, and 3 × 10⁵ cells were transferred into recipient mice via i.n. route. Mice were subsequently infected with 10⁴ PFU of PR8 i.n. 1 d following adoptive transfer, and lungs were harvested on 3 DPI as described above.

Alexa Fluor 488 beads (Invitrogen) were resuspended in PBS and sonicated at room temperature for 30 min prior to administration into recipient mice. After i.n. instillation of 1 × 10⁷ beads in 40 μl of PBS, mice were allowed 1 h to recover with food and water before harvesting the lungs.

Lung tissues were prepared and cells were obtained and stained as described above. Fluorescent data were recorded by an LSR II flow cytometer or Beckman MoFlo Astrios and analyzed using FlowJo software.

Neutrophils were purified from the bone marrow of young mice using EasySep mouse neutrophil enrichment kits from Stemcell Technologies (Vancouver, BC, Canada) as specified in their protocol. The neutrophils were diluted to a final concentration of 1 × 10⁶/ml and rendered apoptotic by incubation in RPMI 1640 with 0.5% FBS in 5% CO₂ overnight at 37°C. Following overnight incubation, Molecular Probes Vybrant cell-labeling stain DiI (designated as Dil) was added in a volume of 5 μl per every 1 × 10⁶ cells to the neutrophils in the incubator for 20 min. The neutrophils were then centrifuged at 1420 × g for 5 min a 4°C. The supernatant was then discarded and the remaining neutrophils were resuspended in sterile 1× PBS at a concentration of 3 × 10⁶ neutrophils per 50 μl. From this suspension, 3 × 10⁶ neutrophils were immediately instilled i.n. into mice that had been anesthetized by isoflurane. The mice were allowed to regain consciousness and rest for 2 h before they were euthanized and BAL was harvested from the lungs. BAL was stained as described above, except Siglec- F (PE) was substituted for Siglec-F (Alexa Fluor 647), and fluorescent data were recorded and analyzed using FlowJo software.

Neutrophil apoptosis and purity assessment were performed with samples of neutrophils before and after staining using the above-mentioned DiI cell stain. The aliquots of neutrophils were diluted to a concentration of 1 × 10⁶ neutrophils/ml for further analysis to determine the level of apoptosis and extent of staining with DiI cell stain, annexin V (BV421), CD11c (FITC), Ly6G (Alexa Fluor 647), and Thermo Fisher Live/Dead fixable dye in near infrared. The enriched samples of neutrophils were ∼80–82% pure for each test, and of the cell membrane–stained neutrophils, 81.1–89.2% were positively stained with DiI. To determine the extent of apoptosis, the flow plots of annexin V were compared with those of the Live/Dead. In experiment 1, 40% of the neutrophils were apoptotic, 50% were dead, and 10% were healthy. In experiment 2, 91.1% of the neutrophils were apoptotic, 7.3% were dead, and 1.62% were healthy.

Apoptosis in AM or in neutrophils was measured by annexin V staining (BioLegend) according to the manufacturer’s protocol.

BAL from healthy and infected mice was obtained as described above. After centrifugation, cell pellets were suspended with 1× RBC lysis buffer (BD Biosciences) and incubated at room temperature for 3 min. Cells were washed with PBS twice followed by resuspension with 100 μl of 1× PBS. Cells were counted using an automated hematology analyzer. Cells (1 × 10⁵) were then spun down using a cytospin machine (Thermo Scientific) for 8 min at room temperature. Slides were allowed to air dry before H&E staining with a Hemacolor staining kit (EMD Millipore, Darmstadt, Germany) following the manufacturer’s protocol. After staining, slides were allowed to air dry overnight before mounting with Vectashield (Vector Laboratories, Burlingame, CA) and coverslip. Slides were then analyzed with a microscope and neutrophils were identified and counted based on their nuclear appearance.

To deplete AM, ready-made clodronate liposomes and control liposomes (FormuMax Scientific, Sunnyvale, CA) were administered to mice using the manufacturer’s recommended dose via i.n. instillation 1 d before and 1 d after PR8 infection. On 6 DPI, clodronate-treated and control mice were euthanized and lung samples were harvested. For survival experiments, mice were monitored on a daily basis as described above.

RNA was extracted from the upper left lung using an RNA extraction kit (Qiagen, Bergisch Gladbach, Germany) and was used as a template to generate cDNA with a cDNA synthesis kit (Clontech Laboratories, Mountain View, CA). To extract RNA from purified AM, AM were placed into 200 μl of TRIzol followed by RNA extraction according to the manufacturer’s recommendations. RNA was then transcribed into cDNA as described above. Real-time PCR was run using the CFX96 Touch real-time PCR detection system (Bio-Rad Laboratories) using the following cycles: 1) 95°C for 10 s; 2) 94°C for 10 s, 60°C for 30 s, 72°C for 20 s (repeat 39 times); 3) 95°C for 10 s, 65°C for 5 s, and 95°C for 5 s. Results were analyzed using the Bio-Rad Laboratories built-in software or the Δ cycle threshold method. For microarrays, RNA quantity was determined by NanoDrop at 260 nm and RNA integrity was assessed by a 2200 TapeStation system with RNA screen tape (Agilent Technologies, Santa Clara, CA). Labeling was performed using the Agilent Technologies Low Input Quick Amp Labeling kit. In brief, the first strand cDNA synthesis was performed using an oligo(dT) primer containing a T7 RNA polymerase promoter site. The cDNA was used as a template to generate Cy3-labeled cRNA that was used for hybridization. After purification and fragmentation, aliquots of each sample were hybridized to SurePrint G3 Mouse Gene Expression v2 8×60K microarrays (Agilent Technologies). After hybridization, each array was sequentially washed and scanned by an Agilent Technologies microarray scanner. Arrays were visually inspected individually for hybridization defects and quality control procedures were applied. Intensity information from captured array images and the annotation information from the microarray experiments were determined using Agilent Feature Extraction 12.0.0 software. Probes with annotations for “accession” were extracted, and interquartile normalization was applied to normalize the gene expression signals by BRB-ArrayTools v4.5.0 (http://brb.nci.nih.gov/BRB-ArrayTools/). In case of redundant probes, we took the one with the highest interquartile range from the samples representing the gene expression levels. To assess genes impacted by aging, differentially expressed genes were identified using significant analysis of microarrays (http://www-stat.stanford.edu/∼tibs/SAM). A false discovery rate (FDR) of 5% was set as the threshold for significance. Data were visualized by generating heat maps with Java TreeView. The microarray data have been deposited at the National Center for Biotechnology Information (http://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?token=yvctemmwrtyttgj&acc=GSE84901) under accession number GSE84901.

Gene set enrichment analysis was performed using the MetaCore software suite (http://www.genego.com/metacore.php). The gene list containing all the genes that were differentially expressed between young and advanced aged macrophages (with statistical significance by significant analysis of microarrays test, FDR < 5%) was analyzed using the Pathway Maps tool in the MetaCore suite, which maps the listed genes to defined signaling pathways that have been experimentally validated and are widely accepted. All reported enriched pathways are listed according to −log (p values) and with an FDR < 0.05%.

A Mann–Whitney U test was used to calculate statistical significance for all comparisons besides survival analysis. The log-rank (Mantel–Cox) test was used to calculate survival statistical significance. Statistical significance was calculated using Prism 7 (GraphPad Software). A p value <0.05 was considered significant (*p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001). Error bars are presented as mean with SEM. Data shown are representative of at least three independent experiments unless otherwise stated.

To study how aging impacts the clinical response to influenza viral lung infection, we first assessed the morbidity and mortality of influenza viral infection (A/PR/8/34, H1N1) in advanced aged (22–24 mo), aged (16–18 mo), and young mice (2–4 mo). One week after influenza viral lung infection (dose, 1 × 10⁴ PFU), 80% of advanced aged mice, 50% of aged mice, and 25% of young mice died (Fig. 1A). Further comparisons between young and advanced aged mice revealed higher morbidity (Fig. 1B, 1C), increased lung damage (Fig. 1D, 1E), impaired viral clearance (Fig. 1F), and increased histological evidence of inflammation (Fig. 1G) than did young infected mice. However, there was no difference in viral titers in serum or evidence of extrapulmonary organ damage between the age groups (e.g., kidney and liver; data not shown). This finding suggests that advanced aged mice do not die of systemic organ failure during influenza viral infection in this model, but instead succumb to morbid complications of influenza viral lung infection.

To further test the susceptibility of advanced aged mice to influenza infection, we infected advanced aged mice with dose ranges of influenza virus (i.e., 1 × 10² to 1 × 10⁶ PFU). We found that at 1 × 10³ PFU, advanced aged mice exhibited a similar mortality as did young mice infected with 1 × 10⁴ PFU (compare Fig. 1A and 1H). No mortality was observed at 1 × 10² PFU dose in advanced aged mice, whereas mortality could not be significantly increased at 1 × 10⁶ PFU dose in advanced aged mice as compared with the 1 × 10⁴ PFU-infected group (Fig. 1H). Overall, these data demonstrate that aging increases mortality and morbidity to influenza viral infection.

As AM exhibit antiviral properties that are critical for host defense to RNA viruses (e.g., influenza virus and respiratory syncytial viruses [RSV]) (15–17), we assessed the role of AM in host defense to influenza virus in both young (2–4 mo) and advanced aged (22–24 mo) C57BL/6 mice.

Clodronate liposomes have been employed in the lung to deplete myeloid cells, which include AM (17), during RNA viral infections (15, 17). Therefore, we administered clodronate liposomes to young and advanced aged mice during influenza viral infection to test the dependency of AM for mortality with aging. We specifically employed influenza doses that induced <50% mortality in young mice (i.e., 1 × 10⁴ PFU) and in advanced aged mice (i.e., 1 × 10³ PFU), based on our dose titration experiments (Fig. 1A, 1H). In either case, clodronate liposomes significantly enhanced mortality (Fig. 2A, 2B). These data demonstrate that AM are critical for host defense in both young and advanced aged mice during influenza viral infection.

We next quantified AM (defined as F4/80⁺ CD11c^hi, Siglec-F⁺ cells) in young and advanced aged mice before infection via flow cytometry. We found a 40–50% lower concentration of AM in the lungs of advanced aged mice than in young mice (Fig. 2C). A similar effect was found in advanced aged and young BALB/c mice (Fig. 2D), indicating that with aging, AM concentration decreases and also that these findings are not restricted to one murine strain. (Note that C57BL/6 mice were used for the remainder of this study).

Although after influenza viral lung infection, the numbers of AM decreased in both advanced aged and young mice log-fold; for advanced aged mice this reduction was more pronounced than for young mice infected with the influenza virus (Fig. 2E). When we assessed the degree of apoptosis in the aged cohorts prior to and after influenza infection, we found no significant alteration in the degree of apoptosis between the groups either before or after infection (Fig. 2F), indicating that increased apoptosis is not likely a mechanism for reduced numbers of AM with aging before or after influenza infection.

To investigate the global impact of age on AM, we next investigated the broad gene transcription profile of advanced aged and young AM via microarray. We determined that the gene signatures of FACS-purified AM from young and advanced aged mice were vastly different. The difference extended to 3545 genes that were significantly different between the age groups (Fig. 3A, Supplemental Table I). Among these altered genes, bioinformatics analyses revealed that pathways involved in the cell cycle were most downregulated with aging (Fig. 3B, 3C, Supplemental Table II). Some of these pathways included metaphase checkpoint, initiation of mitosis, spindle assembly, and chromosome separation. Pathways that were upregulated with aging included inflammatory pathways involved with substance P, PGE₂, macrophage inhibitory factor, oxidative burst and IL-8, and vascular endothelial growth factor signaling (Supplemental Table III). Thus, at baseline aging has a large impact on the transcriptome of AM and leads to specific downregulation in pathways involved in cell cycling but upregulation of certain inflammatory pathways.

As AM self-renew within the lung with minimal contribution from circulating monocytes (14, 18), we hypothesized that the reduced transcriptional signature of proliferation of AM with aging is due to impaired cell turnover. In support of this, we found that aged AM exhibited lower proliferation than did AM from young mice, as demonstrated by Ki67 staining (Fig. 3D). We next measured the gene expression of the repressors of self-renewal: MafB and c-Maf (19, 20). We found that AM from advanced aged mice exhibited higher gene expression of c-Maf but not of Mafb as compared with AM from young mice (Fig. 3E, 3F). Overall, these data indicate that impaired cell proliferation likely contributes to reduced AM numbers with aging.

We examined the functional impact of aging on AM responses during influenza viral infection via adoptive transfer. Specifically, we used an established approach to adoptively transfer AM into the lungs of mice (16) to examine whether adoptive transfer of AM from young mice into advanced aged mice could modify lung damage to influenza viral infection. First, we i.n. transferred 3 × 10⁵ FACS-purified AM from young C57BL/6 CD45.1⁺ donor mice into noninfected, advanced aged C57BL/6 CD45.2⁺ mice. The transferred cells were detected in the BAL 4 d after transfer, indicating that the aged lung permitted engraftment of AM (Fig. 4A). We then purified AM from young and from advanced aged C57BL/6 CD45.2⁺ mice and i.n. transferred them into young C57BL/6 CD45.1⁺ recipients. Four days after transfer, we found similar numbers of young and advanced aged AM in the lungs of young recipient mice (Fig. 4B), indicating that aging did not impair the cell-intrinsic ability of AM to engraft into the lung.

We next transferred 3 × 10⁵ AM from either young donor mice or advanced aged donor mice into advanced aged recipient mice. One day after transfer, mice were infected with 1 × 10⁴ PFU of influenza virus. Three days after infection, we found that advanced aged mice engrafted with young AM had less lung damage than did advanced aged mice engrafted with age-matched AM (Fig. 4C). The viral titers in these advanced aged engrafted mice were comparable to those of advanced aged mice that received AM from either young or advanced aged donors (Fig. 4D), indicating that defects with aging in AM did not impair viral control in this model. Interestingly, advanced aged mice that were infected with influenza virus but did not receive adoptive AM transfer exhibited a similar degree of lung damage as advanced aged mice that received adoptive transfer of AM purified from advanced aged mice (Fig. 4C). Overall, these data show that aging impairs the intrinsic ability of AM to limit lung damage during influenza viral infection.

To define the cell-intrinsic mechanism of aging on AM, we examined the effect of aging on AM phagocytosis in vivo. As clearance of apoptotic neutrophils has been shown to contribute to inflammation resolution (21), we examined the impact of aging on the ability of AM to phagocytose apoptotic neutrophils in vivo. For this purpose, we instilled apoptotic neutrophils i.n. into young and advanced aged mice and found that AM in advanced aged mice exhibited a lower ability to phagacytose apoptotic neutrophils than did young mice (Fig. 5A, 5B). We also found that AM in advanced aged mice exhibited a significantly lower ability to bind and internalized florescent particles than AM in young mice (Fig. 5C), which is consistent with the results of a prior in vitro study (22).

When we enumerated neutrophil influx into the lungs of young and advanced aged mice during influenza viral infection, we found a similar peak of neutrophil influx between the groups (Fig. 5D), similar to what has been previously reported in young infected mice (23). However, neutrophils were retained within the lungs of the advanced aged mice but not in the young infected mice (Fig. 5D). Additionally, by the end of influenza infection advanced aged mice exhibited 2-fold higher MPO levels, a marker of neutrophil activation, than did young mice (Fig. 5E). Finally, we measured MPO levels in the BAL of clodronate-treated advanced aged mice infected with 1 × 10³ PFU of virus, a dose that we found accelerates mortality during influenza viral infection (Fig. 2B). As compared with control treated and infected advanced aged mice, clodronate-treated and infected advanced aged mice exhibited a significant, 2-fold increase in MPO levels (Fig. 5F). These data link AM to increased neutrophil activation within the lungs during influenza infection with aging. Overall, these data demonstrate that with aging AM exhibit a defect phagocytosing apoptotic neutrophils. The data also show that during influenza viral infection with aging, neutrophils are retained in the lung, which could contribute to neutrophil activation and lung damage.

Owing to the phagocytosis defect in AM with aging, we next determined the cellular pathway by which aging impairs scavenging of debris. For this purpose, we assessed the expression of scavenging receptors on AM from young and advanced aged mice. Under basal conditions, we found that AM from advanced aged mice exhibit a significant reduction in the scavenging receptor CD204 (also known as macrophage scavenging receptor-A) (24) compared with AM from young mice (Fig. 6A). However, expressions of CD44, CD36, Axl, and CD206 were not different between the groups (Fig. 6B–E), indicating that aging selectively impairs the expression of a key phagocytic receptor CD204 on AM.

Our study has demonstrated that aging increases morbidity and mortality during influenza viral lung infection in mice. These results are compatible with clinical studies that show that as people age they exhibit increasing morbidity and mortality during influenza viral infection (1–3). Our results are also consistent with two prior studies in aged mice (6, 7) but sharply contrast with a recent study that found in a small cohort of mice (n = 5 per group) that mice aged >24 mo exhibited a significant longer survival during influenza viral lung infection than did young mice (10). In our study, we employed dose-titrating aliquots of virus, and demonstrated that advanced aged mice exhibit an increased susceptibility to influenza viral lung infection than that in young mice. Thus, our study resolves a recent controversy concerning the impact of aging on the susceptibility to influenza lung infection in mice (4, 6, 7, 9, 10).

Although prior studies have identified that signaling pathways (e.g., inflammasomes) (6) and altered inflammatory mediators (e.g., PGD₂) contribute to age-enhanced mortality during influenza viral lung infection (7), the critical cells that enhance mortality to influenza viral lung infection with aging remain to be elucidated. To identify the critical cell for age-enhanced mortality during influenza viral infection, we identified AM as critical for mortality with aging. We documented that with aging AM exhibit an impaired ability to limit lung damage during influenza viral infection. AM are known to maintain lung homeostasis by clearing debris (11, 25). Prior in vitro studies have found that with aging, AM exhibit reduced inflammatory responses to either Streptococcus pneumoniae or RSV (26, 27). Our study has defined the impact of aging on AM function in vivo. Specifically, we show, via adoptive transfer, that aging impairs the ability of AM to limit lung damage without altering viral clearance during influenza viral infection.

To identify a cellular mechanism by which aging impairs the ability of AM to limit lung damage during influenza viral infection, we documented that aging impairs the ability of AM to bind and internalize apoptotic neutrophils and found that during influenza viral infection, advanced aged mice display more neutrophil retention than do young infected mice. As excessive neutrophil numbers increase lung damage during influenza viral infection (23), our study indicates that the retention of neutrophils in the lung with aging contributes to increase lung damage. We also determined that aging selectively downregulates CD204, a key scavenging receptor that has been found to be critical for binding and internalization of apoptotic cells in vitro (28). CD204 is also important for limiting noninfectious lung damage (29), host defense to systemic herpes viral infection (30), and for bacterial pneumonia (31). Why aging selectively downregulates this receptor is not clear from our study. Possibilities include that the basal inflammatory milieu is increased in the aging lung and that this increased inflammatory response may downregulate CD204 (24), given that aging and senescence are associated with an increased inflammation generally (32). Clearly, future studies are required to examine these possibilities in more detail.

AM are not the only immune cells within the lung that exhibit phagocytic functions. Dendritic cells and recruited monocytes also act as phagocytes. However, AM orchestrate the recruitment of inflammatory monocytes during RSV infection via the production of type I IFNs (16). AM may function in a similar fashion in response to influenza viral infection. Also, the impaired viral control that we noted with aging is probably not due to a decreased ability of aged AM to clear virus, as adoptive transfer of young donor AM into advanced aged mice did not enhance viral clearance as compared with advanced aged mice that received advanced aged donor AM. Moreover, defects in AM phagocytic function with aging may have secondary effects on the clearance of apoptotic cells or other debris during respiratory viral infections. It remains to be seen in future studies how aging impacts both viral clearance and the influx of monocytes into the lung during respiratory viral infection.

We found that aging leads to a reduction in both the number and the proliferative capacity of AM prior to infection. Our gene expression studies revealed that aging downregulates several pathways involved in cell cycle regulation, and it upregulates a repressor of self-renewal, C-Maf. Emerging evidence has now established that AM, similar to other tissue-resident macrophages, self-renew in situ rather than being replenished from infiltrating monocytes (14, 18). Our results imply that with aging the self-renewal capacity of AM decreases, leading to a reduced population size. The reduced population size along with the intrinsic alteration in phagocytosis that our study reveals could both contribute to enhanced mortality with aging during influenza viral infection.

In conclusion, our study has revealed a novel mechanism by which aging impairs phagocytosis by AM of apoptotic neutrophils to contribute to an increased morbidity during influenza viral infection. Future therapeutics aimed at improving the function of AM may improve outcomes in older people infected with respiratory viruses.

We thank Dr. Bethany Moore and Dr. Jeffrey Curtis (both University of Michigan) for their careful critique of the manuscript.

The authors have no financial conflicts of interest.