Secondary influenza challenge triggers resident memory B cell migration and rapid relocation to boost antibody secretion at infected sites

Abstract

Resident memory B (BRM) cells develop and persist in the lungs of influenza-infected mice and humans; however, their contribution to recall responses has not been defined. Here, we used two-photon microscopy to visualize BRM cells within the lungs of influenza -virus immune and reinfected mice. Prior to re-exposure, BRM cells were sparsely scattered throughout the tissue, displaying limited motility. Within 24 h of rechallenge, these cells increased their migratory capacity, localized to infected sites, and subsequently differentiated into plasma cells. Alveolar macrophages mediated this process, in part by inducing expression of chemokines CXCL9 and CXCL10 from infiltrating inflammatory cells. This led to the recruitment of chemokine receptor CXCR3-expressing BRM cells to infected regions and increased local antibody concentrations. Our study uncovers spatiotemporal mechanisms that regulate lung BRM cell reactivation and demonstrates their capacity to rapidly deliver antibodies in a highly localized manner to sites of viral replication.

Keywords: antibody response; humoral immunity; influenza virus; live imaging; memory B cell; mucosal immunity; plasma cells; resident memory B cells; tissue-resident immunity; two-photon microscopy.

Graphical abstract

Figure 1

Tracking lung-resident memory B cells (A) Schematic of BAT reporter mice. (B) FACS plots of naive (B220⁺ tdTomato^neg mVenus^neg GL7^neg) and memory (tdTomato⁺ mVenus^neg GL7^neg) cells from spleens and lungs of BAT mice 6 weeks postinfection. In the lung, plots are also pregated on parenchymal cells (in vivo CD45^neg). (C) Frequencies of cells expressing the indicated markers gated as in (B). Data represent one of 3 independent experiments. Statistical analysis was made using an ordinary one-way ANOVA. Error bars represent SD. ^∗∗p < 0.01; ^∗∗∗p < 0.001; ^∗∗∗∗p < 0.0001. (D–F) Bulk RNA-seq of naive B cells and HA⁺ memory B cells from lung and spleen derived from BAT mice. Lung B cells were also pre-gated on parenchymal CD69⁺ cells. Gating shown in Figure S1D. (D) Heatmap of key migration and residency genes. (E) Full heatmaps of all differentially expressed genes with an adjusted p value (FDR) < 0.05. (F) PCA plots, applied to the top 500 most variable genes. Shown are data from 3 samples collected in 3 independent experiments (n = 6 per experiment). (G) HA specific memory B cell, as defined as in (E and F) and PC numbers (gating shown in Figure S1D) in BAT mice, quantified over time. Data in (G) are pooled from 6 independent experiments with 4–6 mice per group. See also Figure S1.

Figure 2

Lung-resident memory B cells increase motility following rechallenge (A) Snapshots from two-photon laser-scanning microscopy (TPLSM) of live explant lungs of BAT >6 weeks postinfection. Top, a typical iBALT-like structure (left), airway-associated PC clusters (middle), and BRM cells (right). Bottom, a BRM cell probing an alveolus. Time lapse is shown in minutes:seconds. White dotted line, alveolus boundaries. Data are representative of 9 videos. (B) Top, experimental design. Bottom left, snapshots from live imaging 24 h post rechallenge or resting memory. Yellow, BRM cell migration tracks. Bottom right, plots displaying tracks of BRM cells from common origin. (C) Time lapse of BRM cells migrating 24 h post rechallenge. Yellow line, migration path of a BRM cell. (D) Mean velocities (left) and displacement (right) of BRM cells in lungs treated as indicated. Each dot represents one tracked cell. (E) Frequency of BRM cell migration velocities. Data in (D and E) were pooled from 4 independent experiments with a total of 4–6 mice per group. Statistical analysis in (D) was made using Kruskal-Wallis tests. Error bars represent SD. ^∗∗p < 0.01; ^∗∗∗p < 0.001; ^∗∗∗∗p < 0.0001. See also Figure S2.

Figure 3

Memory B cells colocalize with infected cells early after activation (A) Point patterns of BRM cell distribution before and after rechallenging. Plots display clustering L values of observed data versus complete spatial random simulation (simulation n = 1,000). (B) L values at r = 200 of BRM cells. Data are pooled from 3–5 independent experiments per group. Each circle represents the mean L value calculated for one mouse (± SD), based on multiple images tiles collected from each mouse. (C) BRM cells 24 h post rechallenge with CFP-S-Flu. Dotted lines demarcate border between highly infected and uninfected areas. Cells are highlighted using Imaris-created spots. (D) BRM cell density in infected and uninfected sites 24 h after rechallenge. Each pair of points represents the average BRM cell density in one mouse, obtained by averaging data from multiple large tiles per animal. (E) A representative image from mice treated with PTX 2 h after CFP-S-Flu rechallenge. (F) L value plot of BRM cells in PTX-treated rechallenged mice. (G) BRM cell density in infected and uninfected areas of PTX-treated, rechallenged lungs. Data are pooled from 4 (C and D) or 3 (E and G) independent experiments. Statistical analyses were made using Mann-Whitney test (B) and paired t tests (D and G). Error bars represent SD. ^∗p < 0.5; ^∗∗p < 0.01. See also Figure S3.

Figure 4

Plasma cells appear within infected alveoli within 4 days of rechallenge (A) TPLSM images of infected and uninfected BAT mice 4 days post rechallenge with CFP-S-Flu. PCs are highlighted using Imaris-created spots (yellow). (B) Alveolar PC density in infected or uninfected areas 4 days after rechallenge with CFP-S-Flu. Data are pooled from 4 independent experiments. (C) Left, confocal microscopy of alveolar PCs near infected cells. Right, zoom of indicated boxed region. (D) L values of observed PC positioning compared with spatial random simulation. Data show the results of one experiment out of 5 performed (at least one mouse per group per experiment). (E) PC L values at r = 200. (F) Densities of alveolar and clustered PCs. Each circle represents the mean densities measured in one mouse, obtained by averaging data from multiple large tiles per animal. Data in (E) and (F) are pooled from 3–5 independent experiments. Statistical analyses were made using a paired t test (B), one-way ANOVA (E), and Mann-Whitney U test (F). Error bars represent SD. ^∗∗p < 0.01;^∗∗∗∗p < 0.0001.

Figure 5

Alveolar plasma cells are derived from memory B cells (A) Experimental outline for (A–C). (B) TPLSM images of CD19^−/− mice transferred with BAT B cells as described in (A) before and after rechallenge. PCs are highlighted using Imaris-created spots (yellow). (C) PCs L values at r = 200. (D) Alveolar PC density in infected or uninfected sites. Right, data represented as the fold-change difference. (E) Left, experimental outline. Right, PCs L values in untreated resting memory mice or rechallenged mice treated with anti-CD40L. Statistical analysis were made using an unpaired t test (C), a paired t test (D), and a Mann-Whitney U test (E). Error bars represent SD. ^∗p < 0.05; ^∗∗p < 0.01; ^∗∗∗p < 0.001. See also Figure S4.

Figure 6

Depletion of alveolar macrophages leads to loss of resident memory B cell mobilization and plasma cell differentiation in infected lungs (A) Experimental setup for (A–C). (B) Plots of BRM cell tacks migrating from a common origin. (C) Mean velocities and displacements of BRM cells. Data are pooled from 4 independent experiments with a total of 3–4 mice per group. (D) BRM density in uninfected and infected sites of mice treated as in (A) and infected with CFP-S-Flu. Right, data represented as the fold-change difference. Data are pooled from 4 independent experiments. (E) PCs 4 days post rechallenge in PBS- and CLL-treated mice. Left, representative images using TPLSM. Right, L values of PCs at r = 200. Plots are pooled from 4 independent experiments. (F) Lung PCs analyzed by flow cytometry 4 days post rechallenge of mice treated with CLL or PBS liposomes as in (A). Data represent one of 3 independent experiments. (G) Left, experimental design. Right, ELISA of anti-influenza (flu) ratios between concentrations of anti-influenza and hamster antibodies measured in the serum and BAL of PBS and CLL-treated rechallenged mice. Each circle represents one mouse. Data are pooled from 3 independent experiments performed. Lines indicate matched data from individual animals. Statistical analysis were made using Kruskal-Wallis tests (C), a paired t test (D, left plot), unpaired t tests (E and F), and Mann-Whitney U test (D, right plot, and G). Error bars represent SD. ^∗p < 0.05; ^∗∗p < 0.01; ^∗∗∗∗p < 0.0001. See also Figure S5.

Figure 7

Migration of resident memory B cells to sites of infection is regulated by CXCR3 (A–C) scRNA-seq of lung leukocytes prior to rechallenge and 1 day post rechallenge of PBS or CLL-treated mice (treated as in Figure 6A). (A) UMAP showing clusters of total cells detected under all 3 conditions combined. (B) Dot plots showing the expression of chemokines and cytokines significantly downregulated in at least one cell type in the CLL versus PBS treated rechallenge mice. Genes in which most significant changes are observed are at the top (p ranges from <10⁻²³ to 10⁻³). Significant differences are indicated by asterisks (Wilcoxon tests, >1.5× change, BH adjusted p < 0.05). (C) UMAPs showing the expression of selected chemokines and cytokines of interest under all conditions combined. (D) Heatmap of qPCR data from the indicated cell subsets sorted from mice treated as in (A), showing results of one of two independent experiments. The average expression from 3 biological replicates is shown. Statistical symbols indicate results of one-way ANOVA, comparison between “rechallenged CLL” and “rechallenged” groups. (E) Experimental setup for (E–G). Animals received one of the two blocking treatments options shown. (F and G) BRM cell density (left), and fold difference of BRM cell densities between infected and uninfected areas (right), in rechallenged mice treated with anti-CXCR3 (F), or anti-IFNγ (G). Data in (F) and (G) are pooled from 4–7 independent experiments. Statistical analysis (F and G) was done using a paired t test (left plots) and Mann-Whitney U tests (right plots). Error bars represent SD. ^∗p < 0.05; ^∗∗p < 0.01; ^∗∗∗p < 0.001; ^∗∗∗∗p < 0.0001. See also Figures S6 and S7.