Temporal dynamics of immune cells in the lung of young and aged hosts following influenza virus infection.

A. Experimental design for temporal analysis of lung cellular responses in young or aged hosts post influenza virus infection. Young (2-3 months old) and aged (∼24 months old) mice were infected with the same dose of PR8. Samples were collected on 0, 2, 9, 14, 30, and ∼60 days post infection (d.p.i). To label circulating/vasculature-associated immune cells, anti-CD45 antibodies were injected intravenously (i.v.) 5 minutes before euthanasia. The right lung homogenate was used to prepare single-cell suspensions for scRNAseq and high-dimensional flow cytometry (CYTEK), while the left lung was processed for histology. This panel was created using BioRender.com. B. Representative H&E staining of the left lung. (C-E). Results were analyzed using all events that passed quality control and pooled from all collected samples (n=3-5 for each sample). C. UMAP visualization of combined scRNA-seq samples. Cells were colored by knowledge based cell type annotation. D. Dot plot displaying module scores of selected GSEA pathways. across all time points comparing lungs from young and aged mice. Dot size and color represents the percentage of cells and average expression level, respectively, in given groups (rows) and pathways (columns). E. Bar graphs showing the proportional cell type composition of time points. The cell types were annotated in (C). Results from (C-E) were analyzed with respect to the number of high-quality cells in each scRNAseq samples. (F-M). Kinetics of immune cells in the lung, quantified by CYTEK, including: Ly6Chi monocytes (F), neutrophils (G), i.v.- NK cells (H), i.v.- γδ T cells (I), eosinophils (J), macrophages (K), i.v.-CD4 T cells (L), and i.v.- CD8 T cells (M). Each dot represents the mean value for one sample, with error bars reflecting variability. The statistical analysis was performed using two-way ANOVA to assess whether age is a significant source of variation: ns, p ≥ 0.05; *, p < 0.05; **, p < 0.01; ***, p < 0.001; ****, p < 0.0001. See also Supplementary Fig. 1.

Enhanced TRM and TRH populations are observed in the aged lung during the memory phase.

(A-B). UMAP of αβ T cells and invariant T cells identified in Fig. 1C. Cells were colored subsequent re-clustering (A, left), cell sub type annotation (A, right), and cell cycle stage (B). C. Dot plot illustrating the expression level of selected genes in CD8+ T cell-related clusters. The color and size of dots represent the average expression and percentage of expressed cells, respectively, for the genes (rows) in clusters (columns). D. Kinetics of age-associated T cells (top) and Cd69+Itgae+ T cells (bottom) quantified from scRNAseq. (E-G). Kinetics of CD8+ T cell subsets in the lung, quantified by CYTEK, including CD69+PD-1hi CD8+ T cells (representing age-associated T cells) (E), CD69+CD103+ CD8+ T cells (F), and antigen-specific T cells (G). H. Dot plot showing the expression level of selected genes in CD4+ T cell-related clusters. The color and size of dots represent the average expression and percentage of expressed cells, respectively, for the genes (rows) in clusters (columns). I. Kinetics of regulatory T cells (Tregs) (top) and tissue-resident helper T cells (TRH) (bottom) quantified from scRNAseq. (J-K). Kinetics of CD4+ T cells in the lung, quantified by CYTEK, including TRH (J) and i.v.-Tregs (K). L. Violin plot depicting the module score for T cell receptor signaling pathway (MsigDB, MM8546) in TRH cells. The violin shape indicates the distribution density of the scores. The embedded boxplot displays the median (center line), the 25th and 75th percentiles (lower and upper hinges of the box), and whiskers extend to 1.5 times the interquartile range from the hinges. Data in (C, H) are shown as dots. Data were pooled from at least three animals per data point in (E-G; J, K). Each dot represents the mean value for that sample, with color fill indicating error bars. Statistical analysis was performed using two-way ANOVA. The impact of age as a source of variation is indicated as: ns, p ≥ 0.05; *, p < 0.05; **, p < 0.01; ***, p < 0.001; ****, p < 0.0001. See also Supplementary Fig. 2.

Increased accumulation of class-switched B cells displaying features of age-associated B cell phenotype is observed in aged hosts.

(A-C). UMAP of B cells identified in Fig. 1C. Cells were colored subsequent re-clustering (A, left), cell sub type annotation (A, right), cell trajectory (B) and cell cycle stage (C). D. Kinetics of bona fide memory B cells (left) and age-associated B cells (right). E. Bar graph showing normalized enrichment score of selected GSEA pathways generated from ranked differential expressed genes comparing two class-switched B cell populations at 61 days post-infection (d.p.i.). F. Bar graph illustrating normalized enrichment score of selected GSEA pathways generated from ranked differential expressed genes comparing bona fide memory B cells between young and aged hosts across all time points. Data in (E-F) were analyzed with GSEA. Statistical significance is indicated as: *, p < 0.05; **, p < 0.01; ***, p < 0.001; ****, p < 0.0001. See also Supplementary Fig. 3.

IMs emerge as a promising candidate regulating the accumulation of adaptive immune cells in aged hosts.

(A-B). UMAP of monocytes and macrophages identified in Fig. S4A. Cells were colored subsequent re-clustering (A, left), cell sub type annotation (A, right), and cell trajectory (B). C. Lineage tracing using Ccr2iCre ZsGreen+ mice. Mice were infected with PR8, followed by daily tamoxifen treatment starting at 13 d.p.i. for 5 days before tissue harvest at 30 d.p.i. The pie chart indicates the composition of ZsGreen+ macrophages. D. Kinetics of monocytes/macrophages quantified by scRNAseq (from Fig. 4A) expressed as a proportion of total quality-controlled events presented in Fig. 1C. E. Kinetics of monocytes/macrophages in the lung quantified by high dimensional flow cytometry (CYTEK). Cells examined include monocyte-derived macrophages (MoM), interstitial macrophages (IMs), Siglec-Fhi alveolar macrophages (AMs), and patrolling monocytes (Patrol. Mo). F. Bar graph showing IM quantification at ∼60 d.p.i. G. Representative plots (upper left) and bar graphs quantifying IM subsets. H. A group of young mice infected with a higher dose of PR8 (orange) was included along with young (yellow) and aged (purple) mice infected with the same lower dose of PR8. Lungs were harvested at 0, 10, 14, 30, and ∼60 d.p.i. Macrophage composition at each time point is presented in radar plots, where each axis represents a distinct macrophage population. Radar plot axes are arranged in an anticlockwise direction to roughly reflect the inferred differentiation trajectory of macrophages shown in Fig 4B. Data were pooled from at least three animals per data point (E-H). In (E), each dot represents the mean value of a sample, with color-filled error bars. Statistical analysis was performed using two-way ANOVA. In (F), results were combined from two experiments; each dot represents one animal and statistical analysis was performed with an unpaired Student’s t test with Welch’s correction. Data are shown with respect to whether the age of the mice served as a source of variation: ns, p ≥ 0.05; *, p < 0.05; **, p < 0.01; ***, p < 0.001; ****, p < 0.0001. See also Supplementary Fig. 4 and Supplementary Fig. 5.

Exuberant type I and type II interferon signaling synergistically drives chronic sequelae in the aged lung.

A. Module scores for “Hallmark, IFNα response (MSigDB, MM3877)” and “Hallmark, IFNγ response (MSigDB, MM3878)” across all samples (top) and across all cells (bottom). Dot size and color intensity represents the percentage of cells and average expression level, respectively, in given pathways (rows) and groups/cell types (columns). B. Experimental design. Starting on 14 d.p.i., infected aged mice received weekly treatments of anti-IFNγ ± anti-IFNαR1 monoclonal antibodies (i.p.). Mice were euthanized on 42 d.p.i., with the left lung collected for histology and the right lung for flow cytometry. C. Representative H&E or Masson’s trichrome staining of lung sections from indicated groups. D. Evaluation of resistance (Rrs) and compliance (Cst) of the respiratory system with flexiVent in infected aged mice post indicated treatment. (E-H). Quantification of macrophage populations, including alveolar macrophages (AMs, E-F) and interstitial macrophages (IMs, G-H). I. Representative flow cytometry plots of TRH in the lung and TFH in the mLN for each treatment group. J. Bar graph showing quantification of TRH. K. Bar graph showing quantification of TFH. Statistical analysis in (D-H, J-K) was performed using repeated measures (RM) one-way ANOVA with Geisser-Greenhouse correction and multiple comparisons. Significance is indicated as *, p < 0.05; **, p < 0.01; ***, p < 0.001; ****, p < 0.0001. See also Supplementary Fig. 6.

Cell-cell interaction analyses reveal unique pathways in young and aged hosts, as well as potential interactions between IMs and CD4+ T cells.

A. Unique interactions in young or aged hosts at 14, 30, and 61 d.p.i. (B-D). Chord plot illustrating UGRP1 signaling (Scgb3a2-Marco) in young hosts at 61 d.p.i. (B), IL-17 signaling (Il17a-(Il17ra+Il17rc)) in aged hosts at 30 d.p.i. (C), NRXN signaling (Nrxn3-Nlgn2) in aged hosts at 61 d.p.i. (D). E. Chord plots displaying CXCL signaling in young and aged hosts during the memory phase. F. Dot plot showing ligands involved in CXCL signaling within mononuclear phagocytes (MNPs) (Fig. S4A, C). Dot size and color intensity represents the percentage of cells and average expression level, respectively, in given cell types (rows) and genes (columns). G. Chord plots displaying co-stimulatory signaling at 61 d.p.i. In the chord plots (B-E, G), each arrow indicates an inferred ligand-receptor interaction. Each arrow originates from the signal-sending cell type and points toward the signal-receiving cell type. Arrow color represents the identity of the signal-sending cell, while arrow thickness reflects the inferred interaction strength. See also Supplementary Fig. 7.

Characterization and immune profiling of the lung in young and aged hosts following influenza virus infection.

Young (2-3 months old) and aged (∼24 months old) mice were infected with the same dose of mouse-adapted influenza virus A/PR/8/34 (PR8). Samples were collected at 0, 2, 9, 14, 30, and ∼60 d.p.i. A. Body weight loss curve. B. Features used to define the general immune populations shown in Fig. 1C. Dot size and color intensity represents the percentage of cells and average expression level, respectively, in given genes (rows) and cell types (columns). C. Module scores for selected pathways, displayed as a dot plot for the cell types defined in Fig. 1C. Dot size and color represents the percentage of cells and average expression level, respectively, in given cell types (rows) and pathways (columns). D-F. Gating strategies for myeloid cells (D), αβ T cells (E), and invariant T cells (F) used in spectral flow cytometry. G. Kinetics of plasmacytoid dendritic cells (pDCs) in the lung, quantified by spectral flow cytometry. Data in (G) are pooled from at least three animals per data point. Each dot represents the mean value for that sample, with error bars filled in color. Statistical analysis was performed using two-way ANOVA. Results are shown to indicate whether the age of the mice served as a source of variation: ns, p ≥ 0.05; *, p < 0.05; **, p < 0.01; ***, p < 0.001; ****, p < 0.0001.

Quantification of αβ T cells using scRNAseq and high-dimensional flow cytometry.

A. Kinetics of proliferating αβ T cells quantified from scRNAseq data (Fig. 2A), expressed as a proportion of all events passing quality control in Fig. 1C. B. Kinetics of CD8+ αβ T cells quantified from scRNAseq (Fig. 2A), expressed as a proportion of all quality-controlled events in Fig. 1C. C. Gating strategy for identifying CD69+PD-1hi CD8+ T cells (representing age-associated T cells) and CD69+CD103+CD8+ T cells by spectral flow cytometry. D. Bar graph showing normalized enrichment score of selected GSEA pathways generated from ranked differential expressed genes comparing two TRM clusters at 61 d.p.i. E. Kinetics of CD4+ αβ T cells quantified from scRNAseq (Fig. 2A), expressed as a proportion of all quality-controlled events in Fig. 1C. F. Gating strategy for identifying TRH cells and i.v.- Treg among CD4+ T cells by spectral flow cytometry. G. Bar graph showing normalized enrichment score of selected GSEA pathways generated from ranked differential expressed genes comparing TRH populations between young and aged hosts at 30 d.p.i. Data in (D) and (G) were analyzed by GSEA, with significance indicated as *, p < 0.05; **, p < 0.01; ***, p < 0.001; ****, p < 0.0001.

Characterization of B cell clusters in young and aged hosts.

A. Dot plot illustrating the defining features of each B cell cluster. B. Kinetics of B cells quantified from the scRNAseq data presented in Fig. 3A, expressed as a proportion of all events passing quality control in Fig. 1C. In (B), each dot’s size represents the percentage of events in each sample expressing the indicated gene, and the color intensity reflects the average expression level.

Characterization of mononuclear phagocytes in young and aged hosts.

(A-B). UMAP of Mononuclear phagocyte (MNP) identified in Fig. 1C. Cells were colored subsequent re-clustering (A), and cell cycle stage (B). C. Dot plot showing selected features used to define dendritic cell (DC) clusters. Dot size and color intensity represents the percentage of cells and average expression level, respectively, in given genes (rows) and cluster (columns). D. Kinetics of DCs quantified from the scRNAseq data presented in (A), expressed as a proportion of the total quality-controlled events in Fig. 1C. E. Dot plot illustrating selected features used to characterize MNPs by flow cytometry. Dot size and color intensity represents the percentage of cells and average expression level, respectively, in given cell types (rows) and genes (columns). F. Kinetics of DCs (gating strategy shown in Fig. S1D) in the lung, quantified by flow cytometry. G. Gating strategy for DC populations. (H-J). Bar graphs displaying the quantification of DC subsets at ∼60 d.p.i., including cDC1 (H), CD11b+PD-L1- DCs (I), and CD11b+PD-L1+ DCs (J). Data were pooled from at least three animals per time point (F-J). In (F), each dot represents the mean value for a sample, with color-filled error bars. Statistical analysis was performed using two-way ANOVA. In (H-J), results were pooled from two experiments; each dot represents one animal, and statistical analysis was performed using an unpaired Student’s t test with Welch’s correction. Significance levels regarding the impact of age as a source of variation are shown as: ns, p ≥ 0.05; *, p < 0.05; **, p < 0.01; ***, p < 0.001; ****, p < 0.0001.

Characterization of monocytes and macrophages in young and aged hosts.

A. Dot plot displaying selected features used to define subsets of monocytes and macrophages. Dot size and color intensity represents the percentage of cells and average expression level, respectively, in given clusters (rows) and genes (columns). B. UMAP plots showing the distribution of monocyte/macrophage clusters across with respect to individual samples. C. Gating strategy for identifying macrophage populations by flow cytometry. (D-E). Kinetics of alveolar macrophages (AMs) in the lung, quantified by CYTEK. AM subsets include CD11b+ AMs (D, left) and CD11b- AMs (D, right). Within the CD11b- AM population, Siglec-Fhi (as shown in Fig. 4E) and Siglec-Flo AMs were quantified (E). (F-G). Bar graphs showing normalized enrichment score of selected GSEA pathways generated from ranked differential expressed genes comparing comparing IM1 (F) or IM2 (G) between young and aged hosts. Data were pooled from at least three animals per data point (D-E). Each dot represents the mean value per sample, with error bars filled in color. Statistical analysis was performed using two-way ANOVA. Data in (F-G) were analyzed by GSEA. Significance levels are indicated as: ns, p ≥ 0.05; *, p < 0.05; **, p < 0.01; ***, p < 0.001; ****, p < 0.0001.

Exuberant type I and type II interferon signaling synergistically drives chronic sequelae in aged lungs.

A. Dot plot displaying Ifng expression by defined cell types (Fig. 1C). Dot size and color represents the percentage of cells and average expression level, respectively, in given cell types (rows) and genes (columns). B. Dot plot illustrating expression of type I IFN genes in defined cell types (Fig. 1C). Dot size and color represents the percentage of cells and average expression level, respectively, in given cell types (rows) and genes (columns). C. Dot plot showing mRNA levels of type I and type II IFN receptors across defined cell types (Fig. 1C). Dot size and color represents the percentage of cells and average expression level, respectively, in given cell types (rows) and genes (columns). D. Gating strategy for Tregs and TRH/TFH cells. E. Representative plots of Tregs in the lung and mLN for each experimental group. F. Bar graph quantifying Tregs in the lung. G. Bar graph quantifying Tregs in the mLN. In (F, G), each dot represents one animal. Statistical analysis was performed using repeated measures (RM) one-way ANOVA with Geisser-Greenhouse correction and multiple comparisons. Significance is indicated as *, p < 0.05; **, p < 0.01; ***, p < 0.001; ****, p < 0.0001.

Cell-cell interaction analyses reveal unique pathways in young and aged hosts, and potential interactions between IMs and CD4+ T cells.

A. Unique interactions identified in young or aged hosts at 0, 2, and 9 d.p.i. B. Kinetics of Rorc (RORγt) + cells quantified from scRNAseq data (Fig. 2A), expressed as a proportion of all events passing quality control in Fig. 1C. C. Dot plot showing selected features used to define subsets of Rorc (RORγt)+ cells. Dot size and color represents the percentage of cells and average expression level, respectively, in given clusters (rows) and genes (columns). D. Chord plot illustrating CCL signaling in young and aged hosts during the memory phase. Each arrow indicates an inferred ligand-receptor interaction. Each arrow originates from the signal-sending cell type and points toward the signal-receiving cell type. Arrow color represents the identity of the signal-sending cell, while arrow thickness reflects the inferred interaction strength. E. Dot plot displaying ligands involved in CCL signaling within MNPs (Fig. S4D). Dot size and color represents the percentage of cells and average expression level, respectively, in given genes (rows) and cell types (columns).