Host Aging Induces a Senescent‐Like Phenotype in Neutrophils and Altered Transcriptional Responses to Streptococcus pneumoniae

TL;DR

Aging causes neutrophils to adopt a senescent-like state, reducing their ability to fight pneumococcal infections in older individuals.

Contribution

The study identifies a senescent-like phenotype in aged neutrophils and shows that blocking TNFα can restore antibacterial function.

Findings

Aged neutrophils fail to upregulate glycolysis and ROS production needed for bacterial killing.

Aged neutrophils show increased SASP factors like IL-10 and TNFα, and reduced apoptosis.

Blocking TNFα improves antibacterial activity and host resistance to pneumococcal infection in aged mice.

Abstract

Aging drives increased susceptibility to respiratory infections by
Streptococcus pneumoniae
(pneumococci). Polymorphonuclear leukocytes (PMNs) are among the first responders in the lung following pneumococcal infection and are required for bacterial clearance. However, PMN antimicrobial function declines with age. To identify mechanisms underlying this decline, we performed RNA sequencing on PMNs in the lungs of young and old mice following pulmonary infection with
S. pneumoniae
. We observed significant transcriptomic differences across host age. Transcriptional analysis followed by functional validation revealed that in infected mice, PMNs from aged hosts failed to upregulate several effector activities including glycolysis and subsequent mitochondrial reactive oxygen species (ROS) production, which are necessary for bacterial killing by PMNs. Conversely, PMNs in aged mice displayed a higher senescence‐associated secretory phenotype (SASP) score and upregulated pathways involved in cellular senescence. Follow‐up functional characterization found that in uninfected hosts, PMNs in aged mice expressed higher levels of SASP factors IL‐10, TNFα, and ROS, had a lower incidence of apoptosis, and had a higher proportion of cells positive for senescence‐associated β‐galactosidase, features of a senescent‐like phenotype. Importantly, blocking TNFα, one of the SASP factors, altered the senescent‐like phenotype and boosted the antibacterial activity of PMNs from aged hosts and increased host resistance to
S. pneumoniae
pulmonary infection. In conclusion, host aging is associated with altered PMN phenotype, including a shift toward senescent‐like energy‐deficient cells, which contribute to impaired host defense and represent potential targets for improved interventions against infection in older adults.

Aging is associated with a senescent‐like phenotype in neutrophils that emerges in bone marrow cells; however, many facets of this phenotype are only fully acquired upon entry into peripheral tissues. This neutrophil senescent‐like phenotype is in part driven by TNFα and impairs resistance of aged hosts to
S. pneumoniae. Created in BioRender. Boughanem (2026). https://BioRender.com/sbef1sl.

Figures

Experimental design. Diagram of experimental layout (A). Principal component analysis (PCA) plot showing variance in expression of the top 500 variable genes in S. pneumoniae challenged mice (B). Venn diagram indicating overlap in DEGs identified as upregulated or downregulated in young and old mice from 12 to 24HPI (C).

Transcriptome of lung PMNs in aged hosts is associated with reduced activation and maturation. Volcano plot of DEGs identified between old and young mice at 24HPI (A). Activation (B) scores for lung PMNs from young and old mice at 12 and 24HPI. Top 10 results of DAVID analysis for terms enriched in DEGs identified as downregulated (C) and upregulated (D) in old vs. young mice at 24HPI.

Glycolysis upregulation is required for PMN killing of S. pneumoniae . Percent bacterial killing using BM PMNs from young mice treated with VC, lonidamine, or 2‐DoG at the listed concentrations (A). Data are pooled from six experiments with significance determined by one‐way ANOVA followed by Tukey's multiple comparisons test (A). Data showing technical replicates of bacterial killing using human peripheral PMNs from five separate donors in the presence of a VC or 2‐DoG at the listed concentrations (B). Significance was determined by one‐way ANOVA (B). Representative results of Glycolysis Stress Test conducted on BM PMNs from young and old mice (C). Basal glycolysis (D) and max glycolysis (E) measurements as derived from Glycolysis Stress Test. Data are pooled from five experiments each with significance determined by unpaired t‐test (C–E). (F) Flow cytometry data showing % glycolytic PMNs in the lung of young and old mice relative at baseline and 24HPI. Data are pooled from 7 to 9 mice per group across three separate experiments (F). Statistical significance was determined by one‐way ANOVA followed by Tukey's multiple comparisons test.

Old mice display a senescence‐like phenotype in PMNs in peripheral organs. Relative SASP Score of lung PMNs in young and old mice 12HPI with significance determined by unpaired t‐test (A). Bubble plot showing curated list of GO terms enriched in PMNs in old vs. young mice at 12HPI (B). Flow cytometry data showing % TNFα+ and % IL10+ PMNs (C, D) and % SA‐β‐gal+ PMNs (E) in the lung and spleen of young and old uninfected mice. Flow cytometry data showing % apoptotic PMNs in the lungs of young and old mice highlighting the percentage of mice displaying less than 10% apoptotic PMNs (F). Flow cytometry data showing relative gMFI for CellROX in the lungs and spleen of young and old uninfected mice normalized to young controls (G). Flow cytometry data showing relative gMFI for MitoSOX in the lungs and spleen of young and old uninfected mice normalized to young controls (H). Flow cytometry data showing gMFI for CellROX (I) and MitoSOX (J) of lung PMNs in young and old Sp‐challenged mice relative to own uninfected controls at 24 h post infection. Data are pooled from 9 mice per group across three separate experiments (C–J). Statistical significance was determined by either Mann–Whitney test (C, D (spleen), E) or unpaired t‐test (C, D (lung), G, H) between indicated groups and unpaired t‐test with respect to uninfected controls (I, J).

Phenotype of circulating and bone marrow PMNs in young versus old mice. RT‐qPCR data showing relative expression of p16 (A), FasL (B), and Ki‐67 (C) in BM PMNs of young and old uninfected mice. Data are pooled from 3 mice per group across three separate experiments. Flow cytometry data showing % TNFα+ (D) and % IL10+ (E) PMNs in the BM and circulation of young and old uninfected mice. Data are pooled from three mice per group across two separate experiments. Flow cytometry data showing relative gMFI for CellROX in the BM and circulation of young and old uninfected mice normalized to young controls (F). Flow cytometry data showing % SA‐β‐gal+ PMNs in the BM and circulation of young and old uninfected mice (G). Flow cytometry data showing % apoptotic PMNs in the BM and circulation of young and old mice highlighting the percentage of mice displaying less than 10% apoptotic PMNs (H). Data are pooled from 6 to 9 mice per group across three separate experiments (F–H). Statistical significance was determined by unpaired t‐test between indicated groups.

Blocking TNFα improves PMN antibacterial activity and host resistance against S. pneumoniae . Flow cytometry data showing geometric mean fluorescent intensity (gMFI) of TNFα (A) and IL10 (B) in PMNs as well as %SA‐β‐gal+ (C) and % apoptotic (D) PMNs in the lung of old uninfected mice following treatment with TNFα blocking antibody or isotype control. Data are pooled from 3 to 5 uninfected mice per treatment group across two separate experiments (A–D). Percent bacterial killing using BM PMNs from young mice and old mice treated with TNFα blocking antibody or isotype control (E). Data are pooled from six mice per treatment group across two separate experiments (E). Data showing bacterial burden in the lungs of old mice treated with TNFα blocking antibody or isotype control prior to Sp‐challenge at 24 h post infection. Data are pooled from six infected mice per treatment group across two separate experiments (F). Statistical significance was determined by unpaired t‐test (C), Kruskal–Wallis followed by Dunn's multiple comparison test (E) and Mann–Whitney test (F) between indicated groups.

Full text

Introduction

1

Streptococcus pneumoniae (Sp)* is a Gram‐positive bacterial colonizer of the nasopharynx that can disseminate to cause severe disease (Narciso et al. 2024). These infections are most prevalent in age extremes, with people above 65 having higher incidence and mortality (Centers for Disease Control and Prevention 2022). As the number of older adults is projected to grow in the next 35 years from ~56 to 95 million (Vespa et al. 2020), understanding the mechanisms that underlie this susceptibility is vital to improving therapies and outcomes in this rapidly growing demographic.

Among the first immune responders to pulmonary infections by Sp are neutrophils (known as polymorphonuclear leukocytes [PMNs]). Circulating PMNs are rapidly recruited to the lungs and during infection new PMNs are generated in the bone marrow (BM) in a process called emergency granulopoiesis resulting in different PMN populations responding to the infection over time (Manz and Boettcher 2014). PMN are vital for host defense against Sp as their depletion results in impaired bacterial clearance and host survival (Bou Ghanem et al. 2015; Garvy and Harmsen 1996; McNamee and Harmsen 2006). However, persistent PMN activity in the context of pneumonia can lead to increased tissue damage, paradoxically leading to increased bacterial numbers and bacterial dissemination (Bou Ghanem et al. 2015; Taenaka et al. 2024).

PMN responses are altered with age (Bhalla et al. 2020; Fortin et al. 2007; Simell et al. 2011; Simmons et al. 2024). In the context of pneumococcal pneumonia, aged hosts have delayed initial beneficial recruitment of pulmonary PMNs with over‐exuberant detrimental influx later in infection (Simmons et al. 2024), and their PMNs are defective in bacterial killing (Bhalla et al. 2020). Furthermore, adoptive transfer of PMNs from young mice reversed the susceptibility of aged mice to pneumococcal pneumonia (Bhalla et al. 2020). This emphasizes the importance of PMNs; however, the underlying cause of their altered responses with age is not fully understood.

Immune cell dysfunction can arise from many sources including the aging process (Lee et al. 2022). Aging is associated with molecular, cellular, and systemic hallmarks (Lopez‐Otin et al. 2013) that affect genomic stability, metabolism, and cellular communication. These impair the ability of the host to fight infection (Ruiz et al. 2017; Stupka et al. 2009). This impairment is in part driven by inflammaging, the low‐grade chronic inflammation (Krone et al. 2014) that accompanies aging and blunts the ability of immune cells to acutely respond to infection. An underpinning cause of inflammaging is cellular senescence, which is defined as a state of stable cell cycle arrest in response to stressors whereby cells remain viable and metabolically active but lose the ability to divide, become resistant to apoptosis, and develop altered responses including a senescence‐associated secretory phenotype (SASP) characterized by the production of inflammatory mediators (Hodes et al. 2016). Cellular senescence plays both beneficial and detrimental roles (Gonzalez‐Gualda et al. 2021; Herranz and Gil 2018; Hodes et al. 2016; Kuehnemann and Wiley 2024); however, senescent cells accumulate with age and are implicated in several aging‐associated diseases (Chaib et al. 2022; Gonzalez‐Gualda et al. 2021; Herranz and Gil 2018; Hodes et al. 2016).

Cellular senescence is well characterized in mitotic cells (Liu et al. 2023), but postmitotic cells can also enter a state of senescence and acquire SASP phenotype (Sapieha and Mallette 2018; Zhao et al. 2024). In immune cells, this process is well described in T cells (Liu et al. 2023). Due to being short‐lived and terminally differentiated, an age associated senescence‐phenotype has not been explored in PMNs. PMNs are abundant with around 10^11^ cells produced in the BM daily (Simmons et al. 2021) that exit to the circulation as post‐mitotic and terminally differentiated cells. Under homeostasis PMNs are short lived with a half‐life of hours, but upon inflammation, they are recruited into tissues where they receive antiapoptotic signals and can live up to days (Ovadia et al. 2023). Given their sheer number, continuous production, and potential for extended lifespan in tissues, it is possible PMNs play a role in senescence. One study found that PMNs drive DNA damage, telomere dysfunction and senescence in hepatocytes via ROS production (Lagnado et al. 2021). PMNs showing a senescent‐like phenotype were recently described in tumors (Rys and Calcinotto 2024). In prostate cancer, apolipoprotein E (APOE) induces a senescent‐like phenotype in PMNs characterized by elevated DNA damage markers, elevated ROS, reduced apoptosis, and immunosuppressive activity (Bancaro et al. 2023). In breast cancer, tumor infiltrating PMNs exhibited a senescence gene signature and elevated senescence‐inducing exosomes (Ou et al. 2022). These data suggest a microenvironment‐dependent induction of senescence in PMNs. However, there are no studies exploring the effect of host aging on senescence of PMNs. Further, although it is well established that the antibacterial function of PMNs declines with aging (Simmons et al. 2021), whether this is driven by cellular senescence of PMNs themselves has not been tested. In this study, we assessed age‐driven changes in PMN responses in a pulmonary infection model and asked if senescent‐like PMNs arise because of host aging.

Results

2

PMNs Display Significantly Different Transcriptomes Across Host Age

2.1

While PMNs were thought of as transcriptionally quiescent, findings in the past decade showed significant changes in transcription following stimulation (Montaldo et al. 2022) and tissue infiltration (Giacalone et al. 2020; Sumagin 2021). To test if there are age‐associated transcriptomic changes in PMNs, we performed RNA‐seq on PMNs isolated from young and old mice that were pulmonary challenged with Sp. A highly enriched population of lung PMNs was obtained 12 and 24 h post infection (HPI) (Figure S1A,B). Differential gene expression analysis was conducted following bulk RNA sequencing (Figure 1A). Three mice were analyzed at each timepoint. One young mouse at 24HPI was removed due to failing a quality control step. Differentially expressed genes (DEGs), defined by an absolute log2 fold‐change expression of ≥ 0.75 and a p‐adj value < 0.05, were identified. PCA analysis showed a similar transcriptomic profile in PMNs from young and old mice 12HPI, that diverges by 24HPI (Figure 1B). This was not due to differences in bacterial burdens in the lungs, as young and old mice had comparable numbers in response to the Sp challenge dose at 12 and 24HPI (Figure S2). The shift in transcriptomic profile by 24HPI coincides with a shift in PMN phenotype from protective to detrimental in the context of Sp infection (Bou Ghanem et al. 2015). Assessment of overlapping gene signatures indicated a conserved shift of 542 upregulated and 1108 downregulated DEGs that occurs from 12 to 24HPI (Figure 1C). However, responses were highly divergent across age with nearly 70% of DEGs identified being unique to young or old mice.

Experimental design. Diagram of experimental layout (A). Principal component analysis (PCA) plot showing variance in expression of the top 500 variable genes in S. pneumoniae challenged mice (B). Venn diagram indicating overlap in DEGs identified as upregulated or downregulated in young and old mice from 12 to 24HPI (C).

We first assessed gene expression changes over the course of infection in young mice. This revealed significant up and downregulation of many genes (Figure S3A). DAVID analysis on the DEGs identified in young mice recapitulated previous findings of upregulation of immune response pathways (Figure S3B,C). This included enrichment of inflammatory response terms involving NF‐κB, IFNγ, and TNFα, conserved proinflammatory pathways important for clearance of Sp by PMNs (Hackert et al. 2023; Jones et al. 2005; Khoyratty et al. 2021).

Analysis of the DEGs in PMNs from old mice revealed broadly similar overall changes in the transcriptome across time as compared to young controls (Figure S4A–C). Of note were the enrichment of the terms “cell cycle” and “cell division” (Figure S4C). Given that mature PMNs are terminally differentiated, this suggests that infiltrating PMNs from aged mice may be more immature and might be retaining a progenitor gene signature as influx of immature PMNs occurs during emergency granulopoiesis (Paudel et al. 2022), which has been observed to increase in old hosts (Gullotta et al. 2023). These findings indicate an age and time‐associated effect on PMN gene expression in response to Sp infection.

Aging Is Associated With Decreased PMN Activation

2.2

We next wanted to study the impact of age on the gene expression at each timepoint following infection. As indicated by the PCA plot, few DEGs were identified at the 12HPI timepoint (Figure S5A) that differed between the two age groups. A significant portion of DEGs identified were part of the Igκ and heavy chain variable gene family previously reported in Sp infected BM PMNs from old mice (Bhalla et al. 2021). Pathway analysis of the upregulated and downregulated genes in old vs. young mice at 12HPI also showed enrichment of pathways associated with the immune response in both up and downregulated genes (Figure S5B,C). Analysis of the transcriptome in young and old mice 24HPI indicated significant differential expression of many more genes (Figure 2A). To determine the activation status of PMNs, we devised an Activation Score derived from normalized expression of genes annotated to the GO terms “Neutrophil Activation”, “Positive Regulation of Neutrophil Activation”, and “Negative Regulation of Neutrophil Activation”. In young mice there was a significant increase in Activation Score from 12 to 24HPI that was not observed in old mice (Figure 2B, Figure S6), where PMNs were less responsive to Sp infection progression. These data suggest that PMNs that influx into the lung of old mice later in infection are less capable of activation. In line with these data, pathway analysis of DEGs in old versus young mice 24HPI indicated downregulation of pathways involved in immune responses (Figure 2C).

Transcriptome of lung PMNs in aged hosts is associated with reduced activation and maturation. Volcano plot of DEGs identified between old and young mice at 24HPI (A). Activation (B) scores for lung PMNs from young and old mice at 12 and 24HPI. Top 10 results of DAVID analysis for terms enriched in DEGs identified as downregulated (C) and upregulated (D) in old vs. young mice at 24HPI.

To assess if aging alters expression of PMN effector function related genes, we focused on genes involved in granule production, phagocytosis, and reactive oxygen species (ROS) production. PMN granules are categorized as primary, secondary, and tertiary defined by their contents (Othman et al. 2022). Primary granules develop early in granulopoiesis (Lehman and Segal 2020), and carry out antibacterial function while contributing to lung pathology through damage of epithelial tissues (Dickerhof et al. 2020; Haegens et al. 2009; Voynow and Shinbashi 2021). PMNs from old hosts exhibited elevated expression of many primary granule components including Elane and Mpo at 24HPI compared to young hosts (Figure S7). This corroborates previous results indicating increased neutrophil elastase activity in older donors (Bou Ghanem et al. 2017) and suggests that PMNs in old mice may be more damaging to tissues (Xu et al. 2024). Conversely, PMNs from young hosts exhibited an infection‐dependent elevated expression of secondary (Figure S8) and tertiary (Figure S9) granules. These granules, which develop later in PMN maturation, have roles in pathogen clearance and tissue remodeling (Gigon et al. 2021). As mature PMNs have many of their granular components already made and prepackaged (Sheshachalam et al. 2014), these changes in transcript abundance may not fully account for the reported age‐driven changes in degranulation (Simmons et al. 2021).

Phagocytosis is required for pneumococcal killing by PMNs (Siwapornchai et al. 2020), while Nox‐mediated ROS production has a limited role in PMN‐mediated killing of Sp (Herring et al. 2022) but is important for killing of other pathogens (Rada et al. 2004). PMNs from young hosts exhibited a near uniform increase in response to infection over time in expression of genes involved in phagocytosis, phagosome maturation, and Nox (Figures S10 and S11). Conversely, PMNs from the old host exhibited little to no change in expression of genes related to phagocytosis and Nox upon progression of the infection from 12 to 24HPI (Figures S10 and S11). These findings are in line with previous reports of aberrant effector functions in PMNs from old hosts, validating the dataset (Simmons et al. 2021).

Altered Transcription Factor Networks in PMNs From Old Mice

2.3

While pathway analysis provides clues as to the altered function and maturity of PMNs, it does not provide mechanisms that underlie these changes. To assess if there are altered transcription factor (TF) networks in PMNs that impact gene expression, we utilized the tool ChEA3, which predicts TF enrichment based on integration of Encode ChIP and co‐expression data (Keenan et al. 2019). Through analysis of the DEGs up and downregulated in both age groups, a total of 98 TFs (24 unique to young, 34 unique to old, 40 overlapping) were identified as significant (TopRanked Score < 0.01) (Figure S12A), with many previously identified as controlling PMN development and function (Xie et al. 2020). Pathway analysis of the identified TFs enriched in young mice supports previous data suggesting a robust immune response by pulmonary infiltrating PMNs (Figure S13), with significant enrichment of terms such as “response to cytokine” and “innate immune response”.

To identify the networks regulated by the TFs identified by ChEA3, we used STRING software (Szklarczyk et al. 2023), a tool that shows physical and regulatory interaction networks. STRING analysis identified 13 functional clusters of TFs (Figure S12B, Table S1) including enrichment of Ets family TFs (Figure S12A,B), which are important for lymphocyte development and function (Sharrocks 2001). Due to similar DNA‐binding motifs, Ets1 family members often share similar targets (Wei et al. 2010). In order to isolate potentially important Ets family members, we utilized TRANSFAC (Matys et al. 2006), another TF prediction tool, whose results were cross referenced with that of ChEA3 resulting in identification of Ets1 (Figure S12C). Ets1 is a TF expressed in lymphoid cells that controls their development and function (Garrett‐Sinha 2023; Lee et al. 2019; Sunshine et al. 2019), but has not been studied in PMNs.

We then wanted to study the role of Ets1 in PMN differentiation and effector function. As full knockouts of Ets1 (Garrett‐Sinha 2023) have broad systemic effects and primary PMNs are difficult to genetically manipulate, we used the Hoxb8 system, which allows for CRISPR/Cas9 mediated gene knockouts and generation of PMNs in vitro (Shannon and Hinnebusch 2023; Wang et al. 2006). Hoxb8 derived PMNs were generated as previously described (Nguyen et al. 2020) resulting in CD11b^+^Ly6G^+^ cells (Figure S14A). Loss of Ets1 had no impact on PMN maturation where Ets1KO cells displayed similar prevalence of mature CD11b^+^Ly6G^+^ cells as compared to the parental line (Figure S14B). As Ets1 may play a role in activation of PMNs, similar to its role in B and T cells (Garrett‐Sinha 2013), we assessed the killing capacity of BM PMNs isolated from Ets1+/+, Ets1+/−, and Ets1−/− mice. Sera from the wild‐type Ets1+/+ and Ets1−/− mice were used for opsonization to control for autoimmune mediated hypocomplementemia (Garrett‐Sinha 2023). PMNs isolated from all three genotypes exhibited similar killing regardless of the opsonin (Figure S14C). In addition, Ets1 transcript levels did not change upon in vitro infection of PMNs isolated from human donors (Figure S14D), suggesting Ets1 does not play a role in PMN function.

Aging Is Associated With Altered Metabolic Signatures in Pulmonary PMNs

2.4

Metabolic shifts are often seen in age‐associated senescence of immune cells that affect function and differentiation (Wiley and Campisi 2021), we therefore examined metabolic responses in PMNs. Activated PMNs rely heavily on aerobic glycolysis, which has been shown to be important for ROS production, phagocytosis, and NETosis (Awasthi et al. 2019; Ettel and Weichhart 2024; Toller‐Kawahisa et al. 2023). Pathway analysis indicated metabolic shifts in pulmonary infiltrating PMNs with significant upregulation of genes involved in the Citric Acid Cycle (TCA) (Figure 2D) in old versus young hosts. In contrast, in young mice there was upregulation of nearly all genes involved in glycolysis accompanied by downregulation of nearly all TCA components during the course of infection (Figure S15A–C). Conversely, old mice displayed the opposite trend with little to no change in expression of glycolytic components and an upregulation of TCA component genes over time (Figure S15A–C). PMN precursors in the BM and circulating immature PMNs were reported to increase utilization of oxidative phosphorylation (Riffelmacher et al. 2017), which is immediately downstream of the TCA cycle.

To assess whether glycolysis is required for bacterial killing, PMNs from young mice were treated with 2‐DoG or lonidamine, known inhibitors of glycolysis. In a dose‐dependent manner, we found that glycolysis inhibition reduced opsonophagocytic killing of Sp (Figure 3A). Similarly, we found that inhibition of glycolysis by 2‐DoG resulted in near‐complete abrogation of S. pneumoniae killing by human PMNs from five donors (Figure 3B). These data corroborate the previous work (Fan et al. 2021), showing the requirement of glycolysis for Sp killing.

Glycolysis upregulation is required for PMN killing of S. pneumoniae . Percent bacterial killing using BM PMNs from young mice treated with VC, lonidamine, or 2‐DoG at the listed concentrations (A). Data are pooled from six experiments with significance determined by one‐way ANOVA followed by Tukey's multiple comparisons test (A). Data showing technical replicates of bacterial killing using human peripheral PMNs from five separate donors in the presence of a VC or 2‐DoG at the listed concentrations (B). Significance was determined by one‐way ANOVA (B). Representative results of Glycolysis Stress Test conducted on BM PMNs from young and old mice (C). Basal glycolysis (D) and max glycolysis (E) measurements as derived from Glycolysis Stress Test. Data are pooled from five experiments each with significance determined by unpaired t‐test (C–E). (F) Flow cytometry data showing % glycolytic PMNs in the lung of young and old mice relative at baseline and 24HPI. Data are pooled from 7 to 9 mice per group across three separate experiments (F). Statistical significance was determined by one‐way ANOVA followed by Tukey's multiple comparisons test.

To determine whether the change in metabolic gene expression constituted an intrinsic shift in the metabolic capacities of PMNs, we assessed glycolytic capacity using Agilent Seahorse. For feasibility with the numbers needed for the assay, we focused on BM PMNs. As PMNs are primarily dependent on glycolysis for their energetic needs, inhibition of respiration is not sufficient to reach maximal glycolytic capacity (Grudzinska et al. 2023). We therefore used opsonized heat‐killed (HK) Sp as a stimulus to mimic pathogen specific responses (Figure S16A), which increased glycolysis beyond that induced by respiration inhibition. We found that PMNs from young and old mice displayed similar basal glycolysis and max glycolytic capacity (Figure 3C–E). As pathway analysis indicated upregulation of the TCA cycle (Figure 2D), we also tested if there were differences in the respiratory capacity of PMNs. To do so, we utilized a Cell Mito Stress Test to characterize respiration (Figure S16B). Results indicated no significant differences in basal respiration (Figure S16C), max respiration (Figure S16D), spare capacity (Figure S16E), and ATP production (Figure S16F). Nevertheless, all bio‐energetic parameters were trending to be lower in old mice, indicating that mitochondrial ability to produce energy slowly declines, with great variability in individual PMNs, as their aging is not homogenous/synchronized. In agreement with less efficient ATP production, the proton leak was most reduced in PMNs from old mice.

To assess if the energetic state of PMNs is altered in the lungs, we used flow cytometry to assess glycolytic (MitoTracker Gr^+^ MitoTracker DR^+^) and respiratory (MitoTracker Gr^−^ MitoTracker DR^+^) PMNs in the lungs as previously described (Díaz‐Basilio et al. 2024). In the absence of infection, young mice displayed a significantly higher portion of glycolytic PMNs as compared to old mice (Figure 3F). Upon infection, the proportion of glycolytic PMNs decreased in young mice corresponding with an increase in respiratory PMNs (Figure S16H). In old mice, there was no observed change in PMN energetics following infection (Figure 3F, Figure S16H). Together, these data indicate significant changes in the metabolic pathways in PMNs in a tissue environment‐dependent manner.

PMNs in Old Hosts Display a Senescence‐Like Phenotype

2.5

Senescent cells are characterized by increased lysosomal content, cell cycle arrest, a secretory phenotype, apoptotic resistance, and changes in metabolism (Hernandez‐Segura et al. 2018). Given the changes in metabolism we observed above in pulmonary PMNs, to assess if aging is associated with a senescence‐like phenotype in PMNs, we examined the RNAseq dataset. Pathway analysis of DEGs upregulated in old relative to young mice at 24HPI found enrichment of pathways related to DNA damage and respiration including “DNA repair” and “Citrate cycle (TCA)” (Figure 2D). Similarly, terms associated with TFs enriched in DEGs from PMNs in old mice suggest an immune response with a senescent phenotype (Figure S13) with enrichment of terms such as “immune system process”, “regulation of cell cycle”, and “cellular senescence”. One hallmark of cellular senescence is the senescence‐associated secretory phenotype (SASP) (Li et al. 2023). To determine if this was evident in the data, we devised a score based on the normalized expression of 22 SASP factors (Bleve et al. 2023; Chaib et al. 2022; Lagnado et al. 2021). Analysis at 12HPI indicated an increased cytokine phenotype in old mice that is consistent with SASP and driven by significantly higher expression of factors such as IL‐10 and TNFα (Figure 4A, Figure S17A). Examining pathways identified by DAVID revealed that at 12HPI there was significant enrichment for several pathways associated with a senescent‐like phenotype (Figure 4A) including positive regulation of cytokine production (secretory phenotype), response to oxidative stress (changes in metabolism), negative regulation of cell proliferation (cell cycle arrest) accompanied by negative regulation of apoptotic processes (resistance to apoptosis) and increase in signaling pathways in response to DNA damage (Figure 4B, Figure S5).

Old mice display a senescence‐like phenotype in PMNs in peripheral organs. Relative SASP Score of lung PMNs in young and old mice 12HPI with significance determined by unpaired t‐test (A). Bubble plot showing curated list of GO terms enriched in PMNs in old vs. young mice at 12HPI (B). Flow cytometry data showing % TNFα+ and % IL10+ PMNs (C, D) and % SA‐β‐gal+ PMNs (E) in the lung and spleen of young and old uninfected mice. Flow cytometry data showing % apoptotic PMNs in the lungs of young and old mice highlighting the percentage of mice displaying less than 10% apoptotic PMNs (F). Flow cytometry data showing relative gMFI for CellROX in the lungs and spleen of young and old uninfected mice normalized to young controls (G). Flow cytometry data showing relative gMFI for MitoSOX in the lungs and spleen of young and old uninfected mice normalized to young controls (H). Flow cytometry data showing gMFI for CellROX (I) and MitoSOX (J) of lung PMNs in young and old Sp‐challenged mice relative to own uninfected controls at 24 h post infection. Data are pooled from 9 mice per group across three separate experiments (C–J). Statistical significance was determined by either Mann–Whitney test (C, D (spleen), E) or unpaired t‐test (C, D (lung), G, H) between indicated groups and unpaired t‐test with respect to uninfected controls (I, J).

To validate the age‐associated senescence‐like PMN phenotype seen in the transcriptional signature of pulmonary PMNs, we measured several markers in vivo in the absence of infection. We first assessed production of SASP‐related cytokines TNFα and IL‐10 in lung PMNs. Analysis of pulmonary PMNs in old mice indicated a significant increase in the percent of cells expressing TNFα (Figure 4C) and IL‐10 (Figure 4D), which was mimicked in relative expression (Figure S18A,B). Together these data suggest that at baseline, PMNs in the lungs of old hosts display a phenotype consistent with SASP.

Another feature of cellular senescence is increased activity of senescence‐associated β‐galactosidase (SA‐β‐gal) (Lee et al. 2006). Analysis of PMNs in the lungs in the absence of infection revealed increases in SA‐β‐gal+ PMNs in aged hosts compared to young hosts (Figure 4E). Interestingly, most SA‐β‐gal+ PMNs in the lung displayed a Ly6G^lo^ phenotype, associated with immature PMNs, which was more prevalent in old mice (Figure S18C). This suggests that there is an age‐related SA‐β‐gal expression primarily in immature PMNs in the lungs.

We next assessed if there was a senescent‐like resistance to apoptosis in pulmonary PMNs (Hu et al. 2022). As expected, in young hosts, we observed upregulation of most genes involved in apoptosis that were not observed in old mice (Figure S17B,C). To assess if there was alteration in actual apoptosis, we measured apoptotic PMNs in the lungs. We observed a greater proportion of old mice with low levels of apoptosis (< 10%) as compared to young mice (Figure 4F). Together, these data suggest that PMNs in old hosts experience greater apoptosis resistance in the lungs.

As our data suggested increased oxidative stress in old mice, we then measured ROS at baseline and after infection. We found that at baseline, PMNs in the lungs of old mice had increased production of ROS (Figure 4G). Following infection, PMNs in the lungs of young but not old mice displayed a significant increase in total ROS (Figure 4I). This implies that lung PMNs in old hosts are experiencing greater oxidative stress even in the absence of infection and are unable to respond to acute stimuli. These findings are in line with the lack of transcriptional changes in nox genes over the course of infection seen in old mice (Figure S11). We next measured production of mitochondrial ROS (mitoROS), which are vital for PMN mediated clearance of Sp (Herring et al. 2022). We found that PMNs in old mice produced significantly lower mitoROS in the lungs as compared to young controls (Figure 4H). Further, PMNs in the lungs of old mice failed to upregulate mitoROS production in response to Sp infection, opposite to what was observed in young mice (Figure 4J). This indicates that PMNs in old hosts are unable to properly upregulate protective mitoROS following migration to the lungs. This suggests that the lung microenvironment in old mice is abrogating PMN acute responses to Sp infection.

To determine whether this senescent‐like phenotype was intrinsic to pulmonary PMNs or was a feature of PMNs in peripheral tissues, we measured responses in the spleen. Similar to what we observed in pulmonary PMNs, splenic PMNs from old mice displayed significantly increased cytokine production (Figure 4C,D), increased SA‐β‐gal (Figure 4E), decreased apoptosis (Figure 4F), and increased ROS (Figure 4G) at baseline. Further, splenic PMNs from old mice also failed to upregulate ROS and mitoROS production (Figure 4I,J) in response to infection. Overall, these findings suggest that in old hosts, PMNs in peripheral organs have features associated with a senescent‐like phenotype and fail to efficiently respond to infection.

To test whether this senescent‐like phenotype is acquired during development in the bone marrow or is tissue‐specific and acquired upon organ entry, we assessed all the above parameters in bone marrow and circulating PMNs. As senescence is also associated with elevated DNA damage response, reduced proliferation, and an accompanied resistance to normal apoptotic clearance (Hernandez‐Segura et al. 2018; Hu et al. 2022), we first looked at these parameters in the BM. In the absence of infection, PMNs within the BM of old mice exhibited elevated expression of p16 (Figure 5A), a cell cycle arrest protein that is upregulated in the event of DNA damage (Rodier et al. 2009), FasL (Figure 5B), a programmed cell death receptor ligand reported to be upregulated in many senescent cells (Lagunas‐Rangel 2023), and lower expression of Ki67 indicative of lower proliferation (Figure 5C). However, when we examined other markers of senescence, we found that none of them appeared in BM PMNs. At baseline, BM PMNs from young and old mice expressed comparable levels of the cytokines we measured (Figure 5D,E) and similar levels of ROS (Figure 5F), displayed no difference in SA‐β‐gal (Figure 5G) and had comparable levels of apoptosis (Figure 5H). Similarly, circulating PMNs also had comparable levels of the above parameters (Figure 5D–H). These findings suggest that there are certain aspects of senescent‐like phenotype, namely DNA damage and reduced proliferation, emerging in PMNs in the bone‐marrow niche where they are generated. However, many facets of senescent‐like phenotype are not evident until the cells migrate to the periphery.

Phenotype of circulating and bone marrow PMNs in young versus old mice. RT‐qPCR data showing relative expression of p16 (A), FasL (B), and Ki‐67 (C) in BM PMNs of young and old uninfected mice. Data are pooled from 3 mice per group across three separate experiments. Flow cytometry data showing % TNFα+ (D) and % IL10+ (E) PMNs in the BM and circulation of young and old uninfected mice. Data are pooled from three mice per group across two separate experiments. Flow cytometry data showing relative gMFI for CellROX in the BM and circulation of young and old uninfected mice normalized to young controls (F). Flow cytometry data showing % SA‐β‐gal+ PMNs in the BM and circulation of young and old uninfected mice (G). Flow cytometry data showing % apoptotic PMNs in the BM and circulation of young and old mice highlighting the percentage of mice displaying less than 10% apoptotic PMNs (H). Data are pooled from 6 to 9 mice per group across three separate experiments (F–H). Statistical significance was determined by unpaired t‐test between indicated groups.

PMNs display an aging phenotype characterized by loss of expression of CD62L and increase expression of CXCR4 and CD49d, in a circadian rhythm dependent manner (Casanova‐Acebes et al. 2013) that has altered function (Ovadia et al. 2023). To test if the observed increase in senescent‐like PMNs is indicative of increased aged PMNs, we tested for their presence by flow cytometry. Analysis indicated that at baseline, old mice displayed similar presence of aged (CD49d+CXCR4+CD62L−) PMNs in the BM (Figure S18D) but decreased prevalence in the lung (Figure S18E). All together, these data suggest that in the periphery PMNs from old mice displayed altered secretory phenotype, apoptosis, and ROS production at baseline and in response to Sp infection, indicative of adoption of a senescence‐like phenotype that is distinct from aged PMNs described previously.

Blocking TNFα Restores PMN Anti‐Microbial Function and Boosts Host Resistance to

S. pneumoniae Infection

2.6

The aged lung tissue microenvironment is known to display inflammaging and express elevated levels of SASP factors including TNFα, which was shown to contribute to the senescence of pulmonary epithelial cells and macrophages in old mice (Hinojosa et al. 2009; Kruckow et al. 2024; Shivshankar et al. 2011). To test whether TNFα contributes to the senescence‐like phenotype of PMNs, we treated mice with anti‐TNFα or isotype control antibodies and compared PMN phenotype in the lungs. We found that in uninfected controls, blocking TNFα slightly reduced production of cytokines (Figure 6A,B), significantly reduced levels of SA‐β‐gal (Figure 6C), and increased the proportion of apoptotic PMNs (Figure 6D) in the lungs. When we examined the effect of TNFα on PMN antibacterial function, we found that blocking TNFα rescued the impairment of bacterial killing by PMNs from old mice to levels comparable to those of young controls (Figure 6E). Given these findings, we then tested the effect of blocking TNFα in old mice on S. pneumoniae pulmonary infection. We found that pre‐treatment of mice with a TNFα blocking antibody starting 2 days prior to infection resulted in a significant 10‐fold reduction in bacterial burden in the lungs compared to isotype treated controls (Figure 6F). Taken together, these findings suggest that the inflammatory environment in old mice, driven in part by TNFα, contributes to the senescence‐like phenotype of PMNs and their age‐driven dysfunction and impairs host resistance to pneumococcal pneumonia.

Blocking TNFα improves PMN antibacterial activity and host resistance against S. pneumoniae . Flow cytometry data showing geometric mean fluorescent intensity (gMFI) of TNFα (A) and IL10 (B) in PMNs as well as %SA‐β‐gal+ (C) and % apoptotic (D) PMNs in the lung of old uninfected mice following treatment with TNFα blocking antibody or isotype control. Data are pooled from 3 to 5 uninfected mice per treatment group across two separate experiments (A–D). Percent bacterial killing using BM PMNs from young mice and old mice treated with TNFα blocking antibody or isotype control (E). Data are pooled from six mice per treatment group across two separate experiments (E). Data showing bacterial burden in the lungs of old mice treated with TNFα blocking antibody or isotype control prior to Sp‐challenge at 24 h post infection. Data are pooled from six infected mice per treatment group across two separate experiments (F). Statistical significance was determined by unpaired t‐test (C), Kruskal–Wallis followed by Dunn's multiple comparison test (E) and Mann–Whitney test (F) between indicated groups.

Discussion

3

Here we examined the effect of host aging on transcriptional changes in PMNs during Sp pulmonary infection. In young mice our results are consistent with previous studies finding significant upregulation of effector function related genes following activation of PMNs in vitro and in vivo (Gomez et al. 2016; Gour et al. 2024; Khoyratty et al. 2021; Lu et al. 2021; Matarazzo et al. 2024; Minhas et al. 2020). However, analysis of the transcriptome in old hosts revealed a significantly different response characterized by reduced effector functions, altered metabolism and maturation, and a senescent‐like phenotype. In vivo functional confirmation of these changes revealed tissue specific phenotypes, where several of the age‐driven alterations were observed in the lungs but not in the BM. This suggests that it is in the periphery that many facets of PMN dysfunction are acquired during host aging. It further suggests that heterogeneity in PMN phenotype arises in part in response to tissue instruction (Casanova‐Acebes et al. 2018; Ganesh and Joshi 2023). These changes may be due to altered lung microenvironment during aging mediated by elevated cytokine or damage related signaling molecules (Leblanc et al. 2024). Whether these changes are restricted to a bona fide subset that arises in the BM (Xie et al. 2020) and acquire phenotypes in the periphery is an open question.

In examining PMN metabolism in infected and uninfected states we found host aging results in elevated respiration in PMNs. As PMNs have been primarily described as glycolytic (Ettel and Weichhart 2024; Fan et al. 2021; Leblanc et al. 2024) this may be indicative of age‐related energy deficit, where PMNs from old hosts are unable to maintain their energetic requirements through glycolysis, despite similar concentrations of glucose being available. We hypothesize that this prompts compensatory elevated respiration to generate energy, which is reflected in changes in gene expression profile during infection progression. Previous studies have found that in many cases aging correlates with decreased mitochondrial function (Gómez and Hagen 2012). As our data suggest that mitochondria in lung PMNs in young but not old mice become more respiratory in response to infection, it is possible that PMNs from old mice have dysfunctional mitochondria that are unable to acutely respond and boost energy production when challenged. This deficit in energy production may in part account for PMN reduced effector capacity. Many of these metabolic shifts have been characterized in B and T cells during aging (Frasca et al. 2020; Liu et al. 2023). Significant changes were observed in T cell metabolism following development of age‐related senescence including reduced cellular respiration (Ron‐Harel et al. 2018) and ATP production (Fang et al. 2016). While there is a classical reliance on glycolysis in senescent cells (Liu et al. 2023), it was observed that glycolysis can be reduced in T cells in aging hosts (Ron‐Harel et al. 2018). Our data suggest that this is also true for PMNs in aged hosts, that have lower basal glycolysis, and are unable to acutely upregulate cellular respiration in response to Sp infection in the lung microenvironment. Together these studies suggest that in aging there may be a shared dysregulation of carbon metabolism in multiple immune cell types in response to exogenous activation.

Lastly, we observed a senescent‐like phenotype in PMNs that emerges primarily in the periphery characterized by increased IL‐10, TNFα, and SA‐β‐gal, reduced incidence of apoptosis and altered metabolism. It is possible that senolytic or senomorphic therapeutics, which clear senescent cells or inhibit production of SASP factors (Zhang, Pitcher, et al. 2023), may rescue PMN anti‐pneumococcal responses in old hosts. These therapies have been shown to improve immune responses following treatment of senescent T cells, B cells, macrophages, and hematopoietic cells in malignancies (Cai et al. 2020; Chang et al. 2016; Cobanoglu et al. 2023; Yousefzadeh et al. 2021). Recent studies have also found removal of senescent‐like PMNs to be beneficial in tumor models (Bancaro et al. 2023; Guo et al. 2023). In infection, studies in hamsters and mice have shown that senolytics significantly decreased serious illness and improved survival upon SARS‐CoV 2 infection, although the effect on PMNs was not directly tested (Camell et al. 2021; Zhang, Suda, and Zhu 2023). Indeed, here we found that blocking TNFα reversed some features of senescence in PMNs, boosted the antimicrobial function of these cells, and improved overall host resistance to pneumococcal infection. These findings are consistent with prior work demonstrating that elevated TNFα during aging contributes to senescence of pulmonary epithelial cells (Hinojosa et al. 2009; Shivshankar et al. 2011) and drives the age‐driven impairment in macrophage function against S. pneumoniae (Kruckow et al. 2024). In summary, this study provides evidence for an age‐associated senescence phenotype in PMNs that is apparent in peripheral tissues and suggests that strategies targeting senescent‐like PMNs could improve disease outcomes in the future.

Materials and Methods

4

Animals

4.1

Young (8–12 weeks) and old (20–22 months) C57BL/6 male mice were obtained from the National Institute on Aging colony or from Jackson Laboratories (Bar Harbor, ME). Experiments were performed in males as they are more susceptible to pneumococcal infection (Gutierrez et al. 2006). All mice were placed at the University at Buffalo in specific pathogen free housing for at least 2 weeks prior to use. Ets1KO mice were generated as previously described (Barton et al. 1998) and maintained on a mixed 129/SVxC57BL/6 background (Wang et al. 2005). All experiments were conducted in accordance with the Institutional Animal Care and Use Committee guidelines.

Human Donors

4.2

Male and female donors, 24–41 years old, were recruited. Individuals taking medications, pregnant, or with infections within the last 2 weeks were excluded. All enrolled donors signed approved informed consent forms in conjunction with IRB approval. Blood was collected into tubes containing sodium citrate at 9 AM to control for circadian rhythm.

Bacteria

4.3

Streptococcus pneumoniae TIGR4 strain was a kind gift from Andrew Camilli. Briefly, bacteria were grown to mid‐exponential phase at 37°C at 5% CO_2_ in Todd Hewitt broth supplemented with 0.5% yeast extract and Oxyrase as previously described (Siwapornchai et al. 2020). Heat‐killed (HK) Sp was generated as described (Herring et al. 2022) by incubation at 55°C for 2 h.

Mouse Infections

4.4

Anesthetized mice were infected by assisted aspiration using the tongue pull method that we call intratracheal (i.t.) with 2 × 10^6^ CFU of Sp where the bacterial inoculum is pipetted into the trachea with the tongue gently pulled out to ensure direct delivery into the lungs. Animals were euthanized for organ harvest. Organs were homogenized and plated on blood agar plates for enumeration of colony forming units (CFU).

PMN Isolation

4.5

For RNAseq

4.5.1

PMN negative selection kit (customized; StemCell) was used per the manufacturer's protocol to obtain a highly enriched population. Briefly, lungs were harvested at 12‐ and 24HPI, digested as described (Herring et al. 2022), and used for PMN isolation. PMN purity was determined by flow cytometry (Figure S1B).

For Opsonophagocytic Killing Assays

4.5.2

Bone Marrow (BM) cells were collected from the femurs and tibias of uninfected mice via density‐gradient centrifugation using histopaques as previously described (Bhalla et al. 2020).

For Seahorse Metabolic Assays

4.5.3

BM cells were collected from the femurs and tibias of uninfected mice. Cells were flushed and red blood cells were then lysed using ACK lysis buffer, and the remaining pellet was resuspended in HBSS. PMNs were then isolated using the StemCell EasySep Mouse Neutrophil Enrichment Kit.

For Human Peripheral Blood PMNs

4.5.4

PMNs were isolated utilizing the StemCell EasyStep Human PMN Isolation Kit as per the manufacturer's protocol.

RNA Purification, Illumina Library Preparation, and RNA Sequencing Analysis

4.6

RNA was isolated from PMNs using a RNeasy Mini Kit according to the manufacturer's specifications. RNA quality, cDNA library prep, sequencing, and data processing were conducted as previously described (Bhalla et al. 2021). We defined significant up‐ or downregulation of gene expression as an absolute fold change of ≥ 0.75 with a p adj value of < 0.05. Heatmaps were generated using normalized transcript per million (TPM) values on the webtool heatmapper.ca (Babicki et al. 2016).

Pathway Analysis

4.7

Pathway and functional enrichment analysis were conducted using the Database for Annotation, Visualization, and Integrated Discovery (DAVID). Terms and pathways with a p value < 0.05 were considered significant. For pathways of the Kyoto Encyclopedia of Genes and Genomes (KEGG), the pathway was retrieved from https://www.genome.jp/kegg/ and edited using GNU Image Manipulation Program to colorize differential expression.

Activation Score

4.8

Summed expression normalized to the average of young mice at 12HPI was used to generate the activation score. The gene list was determined by GO terms for neutrophil activation, positive regulation of neutrophil activation, and negative regulation of neutrophil regulation. Activation Score was calculated using the following formula: sum (neutrophil activation gene expression) + sum (positive regulation of neutrophil activation gene expression)−sum (negative regulation of neutrophil activation gene expression).

Transcription Factor Signature Enrichment and Protein Interaction Network Analysis

4.9

DEGs identified were input into the ChEA3 online webtool and analyzed using TopRanked scoring (Keenan et al. 2019). TFs with a score < 0.01 were considered significantly enriched. TFs identified as significantly enriched were input into the STRING interaction network analysis webtool (Szklarczyk et al. 2023). The interaction score was set to high confidence (0.7). Clusters were identified using the MCL clustering (inflation parameter = 2.0). The resulting clustered interaction network was edited using GIMP to colorize to indicate enrichment in young (yellow), old (pink), or both (black). TRANSFAC analysis was conducted using gProfiler functional enrichment analysis webtool utilizing default settings (Kolberg et al. 2023).

Hoxb8 PMN Differentiation

4.10

Hoxb8 cell lines were maintained and differentiated as previously described (Nguyen et al. 2020). In brief, Hoxb8 progenitors were maintained in RPMI 1640 supplemented with 10% FBS, β‐estradiol (E2), and stem cell factor (SCF). Cells were washed three times in PBS and cultured in E2‐free RPMI 1640 supplemented with FBS, G‐CSF, IL‐3, and SCF for 2 days. Cells were then resuspended in E2‐free RPMI supplemented with FBS and G‐CSF for 1 day.

Extracellular Flux Assays

4.11

Extracellular flux assay protocols were modified from previous protocols (Grudzinska et al. 2023; Mathuram et al. 2022). In brief, PMNs resuspended in RPMI1640 supplemented with glucose, pyruvate, glutamine, and FBS were seeded on Seahorse Cell Culture plate precoated with murine sera. Extracellular flux was measured using a Seahorse Xfe96 analyzer using the setups below.

For Respiration

4.11.1

Oxygen consumption rate was obtained following sequential treatment with 2.5 μM oligomycin, FCCP, and 1 μM mixture of rotenone/antimycin A. Basal respiration, max respiration, respiratory capacity, proton leak, and ATP production were calculated as previously described (Mathuram et al. 2022).

For Glycolysis

4.11.2

Extracellular acidification rate (ECAR) was converted to proton efflux rate (PER) utilizing Agilent Wave software following treatment with Vehicle Control (VC) (unbuffered RPMI1640), murine sera opsonized Heat‐killed Sp (MOI = 10), and 60 mM 2‐deoxyglucose (2‐DoG). Basal glycolysis and max glycolysis were calculated as described previously (Grudzinska et al. 2023).

Opsonophagocytic Killing Assay

4.12

The ability of PMNs to kill S. pneumoniae ex vivo was measured using an opsonophagocytic killing assay as previously described (Bou Ghanem et al. 2015). Briefly, PMNs treated with either VC (DMSO), lonidamine at a range of concentrations (100‐300 μM) used in prior studies to inhibit glycolysis in immune cells (Chen et al. 2022; Cheng et al. 2019) or 2‐DoG at concentrations equimolar to that of glucose in HBSS (6 mM) to 10‐fold higher (600 mM). Cells were then infected with Sp (MOI = 0.01) opsonized with mouse serum. Reactions were rotated at 37°C for 45 min. Percent killing was determined plating on blood agar plates in comparison to the no‐PMN controls.

RT‐qPCR

4.13

RNA was isolated from the indicated samples utilizing a Qiagen RNeasy Mini Kit as per the manufacturer's instructions. cDNA conversion was done using the SuperScript VILO cDNA Synthesis kit. RT‐qPCR was done using iQ SYBR Green Supermix to detect relative levels of transcript cDNA using the following primers in Table S2.

Flow Cytometry

4.14

The lungs were digested as previously described (Siwapornchai et al. 2020). Splenocytes were harvested following physical disruption of the spleen. BM cells were collected from the femurs and tibias. Red blood cells were lysed using ACK lysis buffer, and the remaining pellet was resuspended in HBSS. Cells were stained with antibodies, kits, and stains purchased from BD Biosciences, Invitrogen, Biolegend, and Dojindo Molecular Technologies used as listed in Table S3. 20,000–50,000 events were collected utilizing a BD Celesta or BD Fortessa Cytometer. Data were analyzed using FlowJo with representative gating depicted in Figure S19.

Bocking TNFα

4.15

Anti‐mouse TNFα (Bioxcell, clone XT3.11) or InVivoMAb rat IgG1 isotype control antibodies were used in old mice. The antibodies were diluted in phosphate buffered saline (PBS) and administered at a dose of 100 μg/mouse (Gao et al. 2022) in 100 μL volume intraperitoneally. The injections were given once daily for 3 days, on days −2, −1, and 0 with respect to infection or PMN harvest.

Statistics

4.16

Data was analyzed using GraphPad Prism 10. Data are presented as mean and standard deviation. All data were tested for normality using Shapiro–Wilk Test. Significance was tested using Student's t‐test, Mann–Whitney test, one‐sample t‐test (different than 1), One‐way ANOVA followed by Tukey's multiple comparisons test or Kruskal–Wallis test as appropriate (ns not significant, *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001).

Author Contributions

M.C.B. and M.B. conducted research, analyzed data, and wrote paper. B.M., A.B., A.P.L., R.H., M.C.K., L.H., and S.L. conducted research. P.A.L., L.A.G.‐S., J.M., and A.B.‐P. provided materials. E.N.B.G. designed research, edited the paper, and had responsibility for final content. All authors read and approved the manuscript.

Funding

This work was supported by National Institute of Health grants R21 AI167956‐01A1 and R01 AG068568‐01A1 to E.N.B.G.

Conflicts of Interest

The authors declare no conflicts of interest.

Supporting information

Appendix S1: acel70435‐sup‐0001‐AppendixS1.pdf.