To delineate the difference of acute responses upon influenza infection between young and aged mice, we firstly explored the establishment of an appropriate infection model through a pilot experiment (Supplementary Figure S1). We found that the manifestation of disease in adolescent (6–8 weeks) mice was much more pronounced than that of young adult (3–5 months) or aged (22–24 months) mice. A previous study suggests that aged mice (16–19 months) are more resistant to influenza infection (Lu et al., 2018). Our observation confirmed this finding and suggested that adult mice are more appropriate to be used as controls while investigating the impact of aging on influenza infection, because they displayed similar clinical susceptibility to influenza infection with aged mice. Based on the above observation, we sought to observe whether and how aging might deviate the acute responses against influenza infection via dissecting viral replication and innate lymphocyte composition in the lungs of adult (8–12 months) and aged mice (22–24 months) at 8 days post infection (Figure 1A).

Our results showed that the average body weights of both groups of mice decreased gradually after infection and were significantly lower on day 8 than day 0 (Figure 1B). While no statistical difference was observed between the dynamic body weight changes of the adult and aged mice, persistent viral replication was detected in a higher proportion of the aged mice (Supplementary Table S1) and the viral titers were significantly higher than those of the adult mice (Figure 1C). These data indicate that, despite similar clinical manifestations, aged mice exhibited impaired viral clearance relative to adult mice.

To investigate the impact of aging on pulmonary innate immune responses, we first established a multiparameter flow cytometry assay to define major innate immune subsets in mouse lung. As shown in Supplementary Figure S2A, major innate immune cell populations were gated based on established markers, including Ly6C⁺monocytes (Mono, CD45⁺CD11b⁺CD64⁻Ly6C⁺), neutrophils (Neu, CD45⁺CD11b⁺CD64⁻Ly6G⁺), eosinophils (Eos, CD45⁺CD11b⁺CD64⁻SiglecF⁺), alveolar macrophages (AM, CD45⁺CD11b⁻CD64⁺SiglecF⁺), interstitial macrophages (IM, CD45⁺CD11b⁺CD64⁺SiglecF⁻), and transitional monocytes (Trans, CD45⁺CD11b⁺CD64⁺SiglecF⁺).

Using this approach, we first examined the composition of innate immune cells in the lungs of uninfected adult and aged mice. No significant differences were observed in the frequencies of these subsets between the two groups (Supplementary Figure S2B).

We then applied the same strategy to delineate pulmonary innate immune cell composition at day 8 post infection and analyzed the correlations of these innate immune cell subsets with disease severity. The loss of body weight and the viral titers in the lungs were used as indicators of disease severity. In adult mice, the loss of body weight tended to positively correlate with viral titers, but this association did not reach statistical significance (Figure 2A) partially because the virus had been completely cleared in 4 out of 7 mice. Due to the same reason, we did not observe significant correlations between lung viral titers and the compositions of innate immune cell subsets or the relative transcription levels of inflammatory cytokines (Figure 2B). By contrast, significant negative correlations were found between body weight loss and both the number and frequency of AMs (Figure 2C), which was consistent with previous findings (Schneider et al., 2014) and indicated that AMs play a protective role during influenza infection.

In aged mice, viral titers also tended to positively correlate with body weight loss (p = 0.0530) (Figure 3A), while the patterns of correlations between disease severity and the compositions of innate immune cell subsets or the relative transcription levels of inflammatory cytokines were obviously different from those in adult mice. The viral titers in lungs of aged mice significantly correlated with the IFN-γ transcription, and also tended to positively correlate with the percentage of transitional monocytes (p = 0.0762) (Figure 3B). Meanwhile, despite no significant correlation was observed between the loss of body weight and the measured parameters, a strong tendency toward statistical significance was observed between body weight loss and the frequency of monocytes or IFN-γ transcription level (Figure 3C).

The mice were further stratified into sub-groups according to their viral status at day 8 post infection: adult virus-cleared (Adult-V⁻), adult virus-persistent (Adult-V⁺), aged virus-cleared (Aged-V⁻), and aged virus-persistent (Aged-V⁺). Principal component analyses (PCA) were, respectively, performed using the datasets of innate immune cell composition and transcription levels of cytokines.

PCA of innate immune cell composition (Figure 4A) revealed that Adult-V⁺ mice were clearly separated from the other three sub-groups, whereas the Aged-V⁺ mice clustered more closely with both the Adult-V⁻ and Aged-V⁻ sub-groups. Quite interestingly, when looking at the PCA of cytokine transcription, we found that all the sub-groups clustered together (Figure 4B). These findings collectively suggested that the innate immune cell might not be efficiently activated in Aged-V⁺ mice although the inflammatory milieu was similar to that of the Adult-V⁺.

Comparisons of innate immune cell composition between Adult-V⁺ and Aged-V⁺ mice further highlighted the following differences: Adult-V⁺ mice displayed significantly higher absolute numbers of neutrophils and eosinophils (Figure 4C left). In terms of proportions, Adult-V⁺ mice exhibited elevated Neu%, IM%, and Eos%, indicating a granulocyte-dominant response (Figure 4C right). In contrast, Aged-V⁺ mice showed relatively enriched macrophage lineages, particularly transitional monocytes. Notably, as shown in Supplementary Figure S3, the total CD45⁺ cell numbers did not differ between adult and aged mice, indicating that these age-associated differences did not arise from impaired lymphocyte infiltration. These findings indicate that age-related differences under viral persistence are primarily reflected in immune cell recruitment and composition rather than cytokine expression, with adult mice mounting a granulocyte-skewed response while aged mice exhibit relative macrophage enrichment.

H1N1 A/Puerto Rico/8/34 (H1N1) (PR8) was propagated in MDCK cells. Briefly, confluent MDCK cells plated in 75 cm² cell culture flasks were inoculated with PR8 at a multiplicity of infection (MOI) of 0.01. After 1-h adsorption at 37 °C with 5% CO₂, 10 mL of fresh maintenance medium, Dulbecco’s Modified Eagle Medium (DMEM) containing 0.5% bovine serum albumin (BSA), 1% penicillin–streptomycin (PS), 25 mM 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid (HEPES), and 2 μg/mL TPCK-treated trypsin, was added to each flask. The cell cultures were maintained at 37 °C with 5% CO₂ for another 3 days. After that, propagated PR8 was harvested from the culture supernatant by centrifuging at 20,000 g for 2 h at 4 °C. Next, the harvested PR8 virus stock was titrated using a plaque assay. Briefly, confluent monolayers of MDCK cells in 12-well plates were infected with 10-fold serial dilutions of the virus stock, ranging from 10⁻⁴ to 10⁻⁸. Then, the plates were incubated at 37 °C with 5% CO₂ for 1 h. After absorption, the inoculum was removed and the cells were covered with 1.5 mL of overlay medium (DMEM containing 0.5% low-melting Agar, 0.5% BSA, 1% PS, 25 mM HEPES and 2 μg/mL TPCK-Trypsin) and incubated at 37 °C with 5% CO2 for 4 days. Post incubation, the overlay medium was removed and the plates were fixed with 2% neutral formalin and stained with 1% crystal violet. Plaques were visually counted, and the viral titers were calculated and expressed as plaque-forming units (PFU) per milliliter.

Experiments using mice were approved by the Research Ethics Review Committee of the Shanghai Public Health Clinical Center Affiliated to Fudan University. Female C57BL/6 J mice were purchased from Shanghai Heyu Biotechnology Co., Ltd. After a 3-day acclimation in the animal biosafety level 2 (ABSL-2) lab, mice were intranasally infected with PR8. Briefly, mice were transiently anesthetized using an animal anesthesia machine (ABS small animal anesthesia machine; Yuyan Instruments, Shanghai, China) with isoflurane (Cat# R510-22, RWD Life Science, China) at 3–4% in oxygen (1.0 L/min) for approximately 1–2 min, until loss of the righting reflex. Subsequently, mice were instilled intranasally with 50 μL phosphate-buffered saline (PBS) containing 1 × 10⁴ PFU of PR8 virus. Following infection, mice were monitored daily for changes of body weight and mortality.

Eight days post infection, mice were euthanized by cervical dislocation. Lung tissues were then freshly isolated and homogenized using a high-throughput tissue grinding machine (Cat# Scientz-192, NingBo Scientz Biotechnology Co., China). After that, the homogenates were centrifuged at 2,000 g for 10 min and supernatants were collected for viral titration following procedures similar to those described above.

The real-time PCR assay for mouse cytokines was conducted according to the method described in our previous works (Wang et al., 2022; Zhang et al., 2025) with minor modification. Briefly, total RNA was extracted from freshly isolated mouse lung tissues using the Trizol reagent (Cat# 15596018, Invitrogen). 1 μg of total RNA from each sample was reversely transcribed into cDNA using the PrimeScript™ RT Reagent Kit (Cat# RR037A, Takara). The panel of cytokine genes examined by qRT-PCR includes IFN-γ, TNF-α, IL-2, IL-1β, IL-17a, IL-4, IL-12, and IL-6, with GAPDH serving as the reference gene. The primers of mouse cytokines and GAPDH genes used for qRT-PCR are listed in Supplementary Table S2. Real-time PCR mix was prepared according to the manufacturer’s instructions (TB Green Premix Ex Taq II kit, Cat# RR820A, Takara). PCR reactions were performed using the ABI 7500 Real-Time PCR system (Thermo Fisher Technology Co., USA) under the following conditions: 95 °C pre-denaturation for 30 s, 1 cycle; 95 °C denaturation for 5 s, 60 °C annealing, extension for 34 s, and 40 cycles. Transcription levels of cytokine genes relative to the GAPDH gene were calculated using the 2^−ΔCt method, where ΔCt = Ct of the target gene - Ct of GAPDH.

Freshly isolated mouse lung tissue was rinsed with sterile PBS and finely minced before being digested with 0.4 mg/mL Collagenase IV (Cat# V900893-1G, Merck) and 0.2 mg/mL DNase I (Cat# 10104159001, Roche) dissolved in Roswell Park Memorial Institute Medium 1,640 (RPMI 1640) at 37 °C in a shaking incubator. 30 min later, digestion was terminated by adding RPMI 1640 containing 5% Fetal Bovine Serum (R5) into the mixture. Next, the digested lung tissue mixture was transferred into 70-μm cell strainers and ground the remaining large aggregates of tissue through the strainer with the rubber end of a syringe plunger. The sieved cell suspension was centrifuged at 400 g for 5 min, and the cell pellet was resuspended in 2 mL red blood cell lysis buffer and incubated for 3 min at room temperature. After incubation, the lysis was stopped by adding 5 mL R5 to each sample and nucleated cells were collected via centrifugation. Finally, purified immune cells were obtained through Percoll density gradient centrifugation. Briefly, 5 mL of 40% isotonic Percoll was added to a 15 mL conical centrifuge tube, and 2 mL of cell suspension was carefully layered on the top. The sample was then centrifuged horizontally at 800 g for 30 min. Cell pellet at the bottom of the tube was collected for downstream flow cytometry assay.

Firstly, the isolated cells were suspended in 100 μL of PBS buffer and stained with 0.1 μL cell viability dye (Zombie Aqua™ Fixable Viability Kit, Cat# 423101, BioLegend) at room temperature. 15 min later, the staining was stopped by adding RPMI 1640 containing 10% Fetal Bovine Serum (R10) to each sample and the cells were centrifuged at 600 g for 5 min. Cell pellet was resuspended in 50 μL of R10 and stained with fluorochrome-conjugated antibodies for 30 min at 4 °C (CD45 (Brilliant Violet 785™, Cat# 103149, BioLegend); CD11b (Brilliant Violet 605™, Cat# 101257, BioLegend); CD64 (PerCP/Cyanine5.5, Cat# 139307, BioLegend); Ly6G (Alexa Fluor® 488, Cat# 127625, BioLegend); Siglec-F (PE/Cyanine7, Cat# 155527, BioLegend); Ly-6C (Alexa Fluor® 700, Cat# 128024, BioLegend)). After a 30-min incubation, the stained cells were washed with 200 μL R10, and resuspended in R10. Flow cytometry acquisition was performed using a BD LSRFortessa™ Cell Analyzer (BD Biosciences). Data were analyzed with FlowJo™ v10.8.1 software.

All statistical analyses were performed using GraphPad Prism 9 (GraphPad Software, USA). Data distribution was examined for normality prior to subsequent testing. For two-group comparisons, unpaired t-tests were applied, while differences across multiple groups were assessed using one-way ANOVA. Survival analyses were carried out using the Mantel-Cox log-rank test. A threshold of p ≤ 0.05 was considered statistically significant. Principal component analysis (PCA) and correlation plots were generated using https://www.bioinformatics.com.cn (last accessed on 10 December 2024), an online platform for data analysis and visualization.