Young and aged female BALB/cBy mice were obtained from the National Institute on Aging (NIA) colony (Charles River Laboratories, Wilmington, MA). Mice were housed at the University of Colorado Anschutz Medical Campus vivarium for a minimum of 2 weeks prior to any experimentation. Young mice were 4–5 months of age (similar to humans 25–30 years) and aged mice were 21–22 months of age (similar to humans >65 years) (Flurkey et al., 2007). All animal experiments were performed humanely under a protocol approved by the Animal Care and Use Committee of the University of Colorado Anschutz Medical Campus. Experiments were performed between the hours of 8–10 a.m. to avoid confounding factors related to circadian rhythms. Mice with tumors were removed from the study and mice losing more than 15% of their body weight were humanely euthanized.

S. pneumoniae serotype 3 (ATCC 6303) infection was performed as previously described (Puchta et al., 2014). Briefly, glycerol stock was thawed in tryptic soy broth at 37°C, and incubated statically at 37°C/5% CO₂ until the culture reached mid-log phase (OD 600 nm = 0.45–0.55). Bacteria were washed twice in sterile PBS and resuspended at an appropriate volume to yield approximately 10⁵ colony forming units (CFU) in 50 µl. Bacterial CFU were verified by culturing serial dilutions of the bacterial inoculate on Tryptone Soya Agar plates with 5% sheep blood (Fisher Scientific) at 37°C/5% CO₂ for 18 h. Mice were anesthetized with an intraperitoneal injection of 12.5 mg/kg of ketamine and 1.25 mg/kg of xylazine (Webster Veterinary), and 50 ul of the prepared inoculum or sterile PBS was instilled intranasally. Mice were held vertically for 1 min to assist inoculum draining into the lungs. Blood CFU were determined by plating 10 ul of whole blood on Tryptone Soya Agar plates with 5% sheep blood (Fisher Scientific) at 37°C/5% CO₂ for 18 h.

Respiratory function was measured using unrestrained whole-body barometric plethysmography (Buxco Research Systems) as described (Shults et al., 2015). Briefly, mice were allowed to acclimate in the sealed chamber for 5 min and the parameters of enhanced pause (penh), breathing frequency (breaths per minute) and tidal volume (ml) were recorded for 10 min by the manufacturer’s software (Buxco FinePointe).

Histological analysis and IHC staining of lungs was performed as previously described, with the following modifications (Patel et al., 1999). The left lung lobe was gravity inflated in 10% formalin and 5 µm paraffin-embedded sections were stained using hemoxylin and eosin (H&E) and scored in a blinded fashion by an experimental pathologist. Briefly, a semi-quantitative total injury score was determined for each tissue section by visual assessment of several pathological criteria including: the number and size of large inflammatory cell accumulations (0–6), the quantity of neutrophils in the airways (0–2), the presence of pigmented macrophages (0–1), the presence of peri-vascular inflammation (0–2) and edema (0–2), and the presence of alveolar debris (0–2). For immunohistochemistry, 8 µm sections were de-paraffinized and antigen retrieval was achieved by heating slides in a pressure cooker for 10 min in citrate buffer (Dako). Slides were placed in 3% H₂O₂ for 10 min, washed in PBS +0.1% Tween 20, and incubated for 20 min in 2.5% normal serum (Vector Laboratories). Sections were immersed in primary antibody for 1 h (Ly6G 1:250, BD Biosciences 551,459; S. pneumoniae 1:1,000; Novus Biologicals NB100-64502), washed, incubated with polymer reagent (Vector Laboratories) for 30 min and washed. Ly6G expression was assayed using 3,3′Diaminobenzidine and S. pneumoniae using alkaline phosphatase according to manufacturer’s protocol (Vector Laboratories). Slides were counterstained with Hematoxylin QS (Vector Laboratories), dehydrated in 99% isopropanol, and mounted with VectaMount medium (Vector Laboratories).

RNA isolation and qRT-PCR was performed from snap frozen lung tissue as previously described (Curtis et al., 2019). Briefly, RNA was isolated using the RNeasy Mini kit (Qiagen) and converted to cDNA by reverse transcription using the iScript cDNA Synthesis kit (Bio-Rad). RT-PCR was performed on the QuantStudio 3 Real-Time PCR System (Thermo Fisher) using TaqMan qPCR Master Mix (Applied Biosystems) and TaqMan probes with the FAM reporter: Ly6g (Mm04934123_m1), Cxcl1 (Mm04207460_m1), Ccl2 (Mm00441242_m1), Cxcl2 (Mm00436450_m), Ccl5 (Mm01302427_m) and Cxcr2 (Mm99999117_m) (ThermoFisher). Results were analyzed using the ΔΔCt algorithm (Livak and Schmittgen, 2001). Gapdh (Mm01143545_m1) with the VIC reporter (ThermoFisher) was used as and endogenous control.

Right lung lobes were collected in cold Hank’s buffered salt solution (HBSS; Gibco) containing 2 mg/ml Collagenase D and 0.1 mg/ml DNase I (Sigma-Aldrich) and kept on ice. Samples were dissociated using a GentleMACS machine (Miltenyi Biotec), incubated at 37°C/5% CO₂ for 30 min, filtered through 70 µm nylon mesh filters, washed, and resuspended in Ammonium-Chloride-Potassium (ACK; Gibco) buffer to lyse red blood cells. After a PBS wash, cell pellets were lysed with Cell Lysis Mix and analyzed in duplicate wells for phosphorylated and total p38 MAPK according to manufacturer’s protocol (Invitrogen InstantOne ELISA kit).

Bronchoalveolar lavage (BAL) was performed as previously described, with minor modifications (Shults et al., 2016). Briefly, lungs were aspirated multiple times with 1 ml of PBS using an 18-gauge catheter. Collected BAL fluid was pooled for each animal and red blood cells were lysed with ACK buffer. Viable cells were counted using Trypan blue exclusion. Gr-1^hiLy6G⁺ neutrophils were magnetically separated from total BAL cells using a Myeloid Derived Suppressor Cell Isolation kit (Miltenyi Biotec) according to manufacturer’s protocol.

Real time quantitative PCR (qPCR) was used to quantify bacterial DNA in the lungs of mice as described previously (Earley et al., 2015). Briefly, DNA was extracted from lung using a DNeasy Blood and Tissue Kit (Qiagen), following the manufacturer’s recommended protocol. qPCR for 18s and S. pneumonaie was performed using TaqMan Fast Advanced Mastermix (Applied Biosystems). The TaqMan 18s probe (4319413E, Applied Biosystems) and a custom CspA forward primer (cpsA-348F (5′-GCT​GTT​TTA​GCA​GAT​AGT​GAG​ATC​GA-3′), reverse primer (cpsA-415R (5′-TCC​CAG​TCG​GTG​CTG​TCA-3′) and probe (cpsA-TaqMan FAM 5′-FAM-AATGTTACGCAACTGACGAG-MGBNFQ1-3′) were used as previously described (Park et al., 2010). SYBR Green (Bio-Rad) in conjunction with forward and reverse primers synthesized by Integrated DNA Technologies were used to detect Pan bacteria (F: TCC​TAC​GGG​AGG​CAG​CAG​T, R: GGA​CTA​CCA​GGG​TAT​CTA​ATC​TT), Escherichia coli (F: CAT​GCC​GCG​TGT​ATG​AAG​AA, R: CGG​GTA​ACG​TCA​ATG​AGC​AAA), and Enterobacteriaceae (F: GTGCCAGCMGCCGCGGTAA, R: GCC​TCA​AGG​GCA​CAA​CCT​CCA​AG). Reactions were run on a QuantStudio 3 Real-Time PCR System (Applied Biosystems). Bacteria were quantified as a ratio of target gene/18s.

Fecal bacterial profiles were determined by broad-range amplification and sequence analysis of 16S rRNA genes as previously described (Wheatley et al., 2020). Briefly, DNA was extracted using the QIAamp PowerFecal kit (Qiagen) and amplicons were generated using primers targeting the V3V4 variable region of the 16S rRNA gene. PCR products were normalized using a SequalPrep kit (Invitrogen), purified and concentrated using a DNAClean and Concentrator Kit (Zymo) and quantified using Qubit Fluorometer 2.0 (Invitrogen). Illumina paired-end sequencing was performed on the Miseq platform with versions v2.4 of the Miseq Control Software and of MiSeq Reporter, using a 600-cycle version 3 reagent kit. MicrobiomeAnalyst software (Dhariwal et al., 2017; Chong et al., 2020) was used for data display and analysis. Beta diversity was determined based on the Bray-Curtis Index and significant differences in beta diversity were assessed by PERMANOVA. Standard measures of alpha biodiversity, including richness (the number of OTUs per sample estimated by Chao1), community evenness (the uniformity of OTU distributions estimated by Shannon) and species diversity (Simpson) were estimated, and ANOVA was performed across the four groups for each diversity index. Differences in the relative abundances of individual OTUs between groups were assessed by nonparametric Mann-Whitney/Kruskal–Wallis tests. n = 8–10 mice per group from two identical repeat experiments.

For all non-microbiome related statistical analysis Prism 9 statistical analysis software (GraphPad) was used to perform the analysis. For comparison of two groups, an unpaired two-tailed Student’s t test was used. For comparison of more than two treatment groups a two-way ANOVA with Tukey’s multiple comparisons statistical test was used. A p-value of 0.05 or less was considered significant. Error bars indicate mean ± SEM.

To begin to evaluate the effects of advanced age on S. pneumoniae infection, we infected young and aged mice with Serotype 3 S. pneumoniae intranasally and monitored survival for 7 days after infection (Figure 1A). Aged mice had 100% mortality at day 4 after infection, compared to 50% survival in the young group, mirroring the increased mortality rates observed in humans (Muder, 1998). Whole lung PCR for S. pneumoniae revealed a 10-fold increase (p < 0.05) in bacterial DNA in the lung of aged mice 24 h after infection, relative to the lung of young-infected animals, suggesting an impaired anti-bacterial response in the older group (Figure 1B). Next, we compared lung function in infected young and aged mice by performing whole body plethysmography daily for 3 days after infection. Aged mice had a 25% decrease (p < 0.01) in breath frequency compared to young mice by 48 h after infection (Figure 1C). In addition, there was an age-specific 52% reduction (p < 0.05) in tidal volume at 72 h after infection (Figure 1C). Airway responsiveness, as measured by an increase in Penh, was also elevated at 24 h in the aged mice, although this failed to reach significance. Taken together, these observations confirm that advanced age significantly alters the disease course of S. pneumoniae infection, with greater mortality and impaired lung function.

To determine if the increased mortality and impaired pulmonary function seen in infected aged mice was associated with excessive lung damage, when compared to young mice, lungs were harvested 24 h after infection, sectioned and stained with H&E for evaluation of pulmonary inflammation and edema. Leukocyte recruitment was apparent in and adjacent to the large airway mucosa in both young and aged mice at this timepoint (Figure 2A, blue arrows). However, leukocyte accumulation was more diffuse in the lungs of aged animals, including significantly greater peri-vascular inflammation (p < 0.0001) along with significantly more edema (p < 0.01), relative to younger mice (Figures 2A,C, green arrows). Immunohistochemical staining for the neutrophil marker Ly6G showed an overall increase in neutrophil accumulation in lungs from infected aged mice that was not seen in the lungs of young-infected animals. In young mice, the neutrophils were predominantly adjacent to the airways, while in aged mice these cells accumulated around both the alveoli and near the vasculature (Figure 2B, brown stain, green arrows). In addition, mRNA levels of Ly6g were 4.5-fold higher (p < 0.05) in the lungs of aged-infected mice compared to those of young-infected mice (Figure 2D). The increase in neutrophils paralleled increased activation of p38 MAPK (mitogen-activated protein kinase) in the lung of aged mice indicating increased activation of pro-inflammatory signaling in the lungs (Figure 2E) (Qian et al., 2016). We saw no obvious changes in macrophage numbers in the lungs of infected mice at this timepoint (data not shown). These data reveal an exacerbation of neutrophil-associated inflammation in the lungs of aged-infected mice that is not observed in young mice, which could explain the impaired lung function and increased mortality with age. Relative to younger infected mice, aged mice also had more staining for S. pneumoniae in their lungs (Figure 2B, pink stain), which is in agreement with the data from Figure 1B showing higher levels of S. pneumoniae DNA in the lungs of aged versus young mice. Thus, despite the increased neutrophil infiltration, aged mice are unable to sufficiently control S. pneumoniae within the lung.

Despite the increased number of neutrophils in the lung, we failed to observe a rise in the expression of the pro-inflammatory cytokines, TNFα, IL1β or IL-6, in the lung of aged mice after infection, when compared to the lungs of young-infected mice (data not shown). Therefore, we next evaluated the expression of chemokines in our model to determine if age-related changes in immune cell recruitment could help explain the increased neutrophil infiltrate. To accomplish this, we isolated Ly6G⁺ neutrophils from the bronchoalveolar fluid of S. pneumoniae infected young and aged mice 24 h after infection and examined chemokine expression. Neutrophils from aged-infected mice expressed 4.5-fold higher levels of mRNA for the neutrophil chemokine (C-X-C motif) ligand 1, CXCL1 (KC) compared to those from young-infected mice (p < 0.001) (Figure 3A). CXCL1 has been shown to be critical for the homing of neutrophils to the lung in pneumonia (Paudel et al., 2019), suggesting that chemokine production by neutrophils themselves is likely contributing to additional cellular recruitment to the lung in age. We also observed age-dependent increases in expression of chemokines involved in recruiting other immune cells. (C-C motif) ligand 5 (CCL5; also known as RANTES), which is important for the direct recruitment of multiple cell types, including T cells, eosinophils and basophils, but can also induce indirect recruitment of neutrophils via activation of macrophages (Hwaiz et al., 2015), was increased ∼6 fold from young mice after infection (p < 0.001) (Figure 3B). mRNA encoding (C-C motif) ligand 2 (CCL2 or monocyte chemoattractant protein-1 (MCP-1)), a chemokine responsible for recruiting monocytes to peripheral tissues (Deshmane et al., 2009) was transcriptionally upregulated 7.5-fold in aged mice, compared to young mice (p < 0.001) (Figure 3C). Lung neutrophils from aged-infected mice had 4.5-fold higher expression of CXCR2, the receptor for CXCL1, relative to lung neutrophils from young mice (p < 0.001) (Figure 3D) and may, therefore, be primed for increased recruitment into tissues in the aged animals by CXCL1 or other chemokines.

It has been demonstrated that the gut microbiome can alter the neutrophilic response to S. pneumoniae intranasal infection by promoting neutrophil clearance from the lung (Felix et al., 2018). Therefore, to evaluate the effects of age and S. pneumoniae infection on the gut microbiome in our model, we performed 16s rRNA sequencing of fecal pellets from young and aged mice at 24 h prior to and at 24 h after infection. Fecal bacterial profiling revealed broad alterations of bacterial phyla in both young and aged mice after infection with S. pneumoniae (Figure 4A). Comparison of beta diversity between treatment groups indicated a small but significant shift in the species diversity in the infection groups compared to the uninfected groups (p < 0.05) (Figure 4B). Analysis of three common measures of alpha biodiversity, Chao, Shannon and Simpson, showed no effect of either age or infection on bacterial richness (Chao), but S. pneumoniae induced a significant change in bacterial community evenness (Shannon, p < 0.01) and diversity (p < 0.01) compared to the age-matched uninfected group, indicating an infection-induced expansion of bacterial species within the microbiome of young and aged mice (Figure 4C). Comparison of individual taxa pre- and post-infection showed a significant infection-dependent increase in two distinct genera belonging to the phylum Bacteroidetes, Barniesiella (p < 0.05) and Parabacteroides (p < 0.01) in both the young and the aged mice compared to uninfected microbiome (Figure 4D). Infection in young, but not aged, mice also led to increased expansion of the beneficial bacteria Akkermansia compared to the young uninfected group (p < 0.05). Taken together, these data demonstrate that intranasal infection with S. pneumoniae can markedly alter the gut microbiome in both young and aged mice.

We next wanted to determine if the S. pneumoniae-induced gut microbiome changes differed between young and aged mice. Analysis of individual bacterial phyla showed an increase in Proteobacteria in the feces following infection in the aged, but not young mice (p < 0.05) (Figure 5A). Proteobacteria is a major phylum of Gram-negative bacteria in the gut microbiome that has been associated with dysbiosis in a number of disease models (Zeng et al., 2017). Further analysis of the Proteobacteria population revealed an 11-fold increase (p < 0.05) in the relative abundance of the Enterobacteriaceae family in the aged-infected feces, compared to young-infected mice (Figure 5B). The E. coli genus accounted for >95% of the Enterobacteriaceae in the gut of these mice (Figure 5C).

To determine if expansion of these potentially pathogenic symbiotes in the gut microbiome parallels changes in the lung we performed a correlation analysis of fecal Enterobacteriaceae and lung expression of Ly6g mRNA and found a significant positive correlation between the two variables (Figure 5D). Furthermore, aged mice had significantly higher levels of bacteria in the blood, compared to young mice, following infection (Figure 5E). Therefore, we performed PCR for Enterobacteriaceae and E. coli DNA on the lungs of young and aged mice with and without S. pneumoniae infection. We were unable to detect any DNA for either Enterobacteriaceae or E. coli in the lungs of uninfected young or aged mice, confirming that Enterobacteriaceae is not a prominent bacterial family in the healthy murine lung, regardless of age (Figure 5F). Enterobacteriaceae was, however, detected in the lungs of both aged and young mice 24 h after S. pneumoniae infection. Interestingly, we found Enterobacteriaceae in significantly more of the aged-infected than young-infected animals (25% v. 80%, p < 0.05) (Table 1). Furthermore, E. coli DNA was only found in the lungs of aged-infected mice, with 40% of the mice having detectable amounts in the lung at 24 h after infection (Figure 5F; Table 1). These data confirm the occurrence of age-specific bacterial dysbiosis in both the gut and the lung following infection with S. pneumoniae, suggesting changes in the gut-lung axis could be contributing to the altered response to this pathogen in aged mice.