Cytosolic S100A8/A9 promotes Ca2+ supply at LFA-1 adhesion clusters during neutrophil recruitment

Abstract

S100A8/A9 is an endogenous alarmin secreted by myeloid cells during many acute and chronic inflammatory disorders. Despite increasing evidence of the proinflammatory effects of extracellular S100A8/A9, little is known about its intracellular function. Here, we show that cytosolic S100A8/A9 is indispensable for neutrophil post-arrest modifications during outside-in signaling under flow conditions in vitro and neutrophil recruitment in vivo, independent of its extracellular functions. Mechanistically, genetic deletion of S100A9 in mice caused dysregulated Ca²⁺ signatures in activated neutrophils resulting in reduced Ca²⁺ availability at the formed LFA-1/F-actin clusters with defective β₂ integrin outside-in signaling during post-arrest modifications. Consequently, we observed impaired cytoskeletal rearrangement, cell polarization, and spreading, as well as cell protrusion formation in S100a9^-/- compared to wildtype (WT) neutrophils, making S100a9^-/- cells more susceptible to detach under flow, thereby preventing efficient neutrophil recruitment and extravasation into inflamed tissue.

Keywords: LFA-1 integrin clustering; acute inflammation; calcium signaling; immunology; inflammation; intracellular S100A8/A9; intracellular signaling; mouse; neutrophil recruitment.

Figure 1.. Cytosolic S100A8/A9 regulates leukocyte recruitment in vivo regardless of extracellular S100A8/A9.

(A) Enzyme-linked immunosorbent assay (ELISA) measurements of S100A8/A9 levels in supernatants of wildtype (WT) bone marrow neutrophils stimulated for 10 min with PBS, E-selectin, or lysed with Triton X-100 (mean + SEM, n=6 mice per group, RM one-way analysis of variance [ANOVA], Holm-Sidak’s multiple comparison). (B) Schematic model of the mouse cremaster muscle preparation for intravital microscopy and representative picture of a vessel showing rolling and adherent cells. WT and S100a9^-/- mice were stimulated intrascrotally (i.s.) with TNF-α 2 hr prior to cremaster muscle post-capillary venules imaging by intravital microscopy. Quantification of (C) number or rolling (rolling flux fraction) and (D) number of adherent neutrophils per vessel surface of WT and S100a9^-/- mice (mean + SEM, n=5 mice per group, 25 [WT] and 30 [S100a9^-/-] vessels, unpaired Student’s t-test). (E) Correlation between physiological vessel shear rates and number of adherent neutrophils in WT and S100a9^-/- mice (n=25 [WT] and 30 [S100a9^-/-] vessels of 5 mice per group, Pearson correlation). (F) Schematic model of sterile inflammation induced by exteriorizing WT and S100a9^-/- cremaster muscles. (G) Analysis of number of adherent leukocytes by intravital microscopy before and after S100A8/A9^mut intra-arterial injection (mean + SEM, n=3 mice per group, 3 [WT] and 3 [S100a9^-/-] vessels, two-way ANOVA, Sidak’s multiple comparison). (H) Representative Giemsa staining micrographs of TNF-α stimulated WT and S100a9^-/- cremaster muscles (representative micrographs, scale bar = 30 µm, arrows: transmigrated neutrophils) and (I) quantification of number of perivascular neutrophils (mean + SEM, n=5 mice per group, 56 [WT] and 55 [S100a9^-/-] vessels, unpaired Student’s t-test). ns, not significant; *p≤0.05, **p≤0.01, ***p≤0.001.

Figure 1—figure supplement 1.. Cytosolic S100A8/A9 is indispensable for neutrophil recruitment in vivo.

(A) Representative confocal images of S100A9 intensity in intravascular and extravascular neutrophils after TNF-α stimulation of wildtype (WT) cremaster muscle tissues (scale bars = 50 μm) and (B) quantification (mean + SEM, n=5 mice per group, 602 [intravascular] and 326 [extravasated] neutrophils, unpaired Student’s t-test). (C) WT neutrophils’ purity assessment and gating strategy (equivalent for S100a9^-/- neutrophils) and flow cytometry analysis of (D) CD18, (E) CD11a, (F) CD11b (G) CXCR2, (H) CD62L, (I) PSGL1, and (J) CD44 surface levels of WT and S100a9^-/- neutrophils (mean + SEM, n=3 mice per group, unpaired Student’s t-test). (K) Schematic model of the adhesion rescue experiments in TNF-α stimulated WT and S100a9^-/- cremaster muscles by intra-arterial application of mutS100A8/A9 (aa exchange N70A+E79A). (L) Quantification of number of adherent WT and S100a9^-/- leukocytes mm^–2 in the same vessel before and after mutS100A8/A9 i.v. injection (mean + SEM, n=4 mice per group, 4 [WT] and 4 [S100a9^-/-] vessels, two-way analysis of variance (ANOVA), Sidak’s multiple comparison). (M) Number of transmigrated WT and S100a9^-/- neutrophils in a classical transwell assay upon stimulation with PBS or CXCL1 (mean + SEM, n=3 mice per group, two-way ANOVA, Sidak’s multiple comparison). ns, not significant; *p≤0.05, **p≤0.01, ***p≤0.001.

Figure 2.. Loss of cytosolic S100A8/A9 impairs neutrophil adhesion under flow conditions without affecting β₂ integrin activation.

(A) Schematic representation of blood harvesting from wildtype (WT) and S100a9^-/- mice via a carotid artery catheter and perfusion into self-made flow cambers coated with E-selectin, ICAM-1, and CXCL1. Analysis of (B) number of rolling and (C) number of adherent leukocytes FOV^–1 (mean + SEM, n=4 mice per group, 10 [WT] and 12 [S100a9^-/-] flow chambers, paired Student’s t-test). (D) Number of adherent leukocytes FOV^–1 in self-made flow chambers coated with E-selectin, ICAM-1, CXCL1, and additionally with extracellular S100A8/A9 (mean + SEM, n=5 mice per group, ≥12 [WT] and 14 [S100a9^-/-] flow chambers, two-way analysis of variance (ANOVA), Sidak’s multiple comparison). (E) Schematic representation of the soluble ICAM-1 binding assay using bone marrow neutrophils stimulated with PBS control or CXCL1 (10 nM) assessed by (F) flow cytometry (MFI = median fluorescence intensity, mean + SEM, n=5 mice per group, two-way ANOVA, Sidak’s multiple comparison). (G) Spectroscopy fluorescence intensity analysis of percentage of adherent WT and S100a9^-/- neutrophils, seeded for 5 min on ICAM-1 coated plates and stimulated with PBS or CXCL1 (10 nM) for 10 min (mean + SEM, n=4 mice per group, two-way ANOVA, Sidak’s multiple comparison). ns, not significant; *p≤0.05, **p≤0.01, ***p≤0.001.

Figure 3.. Cytosolic S100A8/A9 is crucial for neutrophil spreading, crawling, and post-arrest modifications under flow.

(A) Representative bright-field pictures of wildtype (WT) and S100a9^-/- neutrophils spreading over E-selectin, ICAM-1, and CXCL1 coated glass capillaries (scale bars = 10 μm). Analysis of cell shape parameters (B) area, perimeter, (C) circularity (4π * [area/perimeter]) and solidity (area/convex area) over time (mean + SEM, n=103 [WT] and 96 [S100a9^-/-] neutrophils of 4 mice per group, unpaired Student’s t-test). (D) Rose plot diagrams representative of migratory crawling trajectories of WT and S100a9^-/- neutrophils in flow chambers coated with E-selectin, ICAM-1, and CXCL1 under flow (2 dyne cm^–2). Analysis of (E) crawling distance, (F) directionality of migration, and (G) crawling velocity of WT and S100a9^-/- neutrophils (mean + SEM, n=5 mice per group, 113 [WT] and 109 [S100a9^-/-] cells, paired Student’s t-test). Western blot analysis of ICAM-1-induced (H) Pyk2 and (I) paxillin phosphorylation of WT and S100a9^-/- neutrophils upon CXCL1 stimulation (10 nM) (mean + SEM, representative western blot of n≥4 mice per group, two-way analysis of variance (ANOVA), Sidak’s multiple comparison). ns, not significant; *p≤0.05, **p≤0.01, ***p≤0.001.

Figure 3—figure supplement 1.. S100A8/A9 deficient cells are more susceptible to increasing shear stress compared to WT cells.

(A) Analysis of number of adherent WT and S100a9^-/- neutrophils under flow as percentage related to the initial number of adherent neutrophils in E-selectin, ICAM-1, and CXCL1 coated flow chambers at indicated shear stress levels. Shear stress was increased every 30sec. [mean + SEM, n=3 mice per group, 3 (WT) and 3 (S100a9^-/-) flow chambers, unpaired Student’s t-test]. ns, not significant; *p≤0.05, **p≤0.01, ***p≤0.001.

Figure 4.. Cytosolic S100A8/A9 drives neutrophil cytoskeletal rearrangement by regulating LFA-1 nanocluster formation and Ca²⁺ availability within the clusters.

(A) Representative confocal images of LFA-1 staining in Lyz2xGCaMP5 and Lyz2xGCaMP5xS100a9^-/- crawling neutrophils on E-selectin, ICAM-1, and CXCL1 coated flow chambers (scale bar = 10 μm). (B) Segmentation of LFA-1 signals through automatic thresholding (scale bars = 10 μm). (C) Size-excluded LFA-1 nanoclusters of 0.15 μm² minimum size from previously thresholded images (scale bars = 10 μm). (D) Single-cell analysis of average number of LFA-1 nanoclusters in min 0–1, 5–6, and 9–10 of analysis of Lyz2xGCaMP5 and Lyz2xGCaMP5xS100a9^-/- neutrophils (mean + SEM, n=5 mice per group, 56 [WT] and 54 [S100a9^-/-] neutrophils, two-way analysis of variance (ANOVA), Sidak’s multiple comparison). (E) Representative confocal images of Ca²⁺ signals in Lyz2xGCaMP5 and Lyz2xGCaMP5xS100a9^-/- neutrophils (scale bars = 10 μm) and (F) Ca²⁺ signals in the previously segmented LFA-1 nanoclusters (scale bars = 10 μm). (G) Quantification of subcellular Ca²⁺ levels in the LFA-1 nanocluster area in min 0–1, 5–6, and 9–10 in Lyz2xGCaMP5 and Lyz2xGCaMP5xS100a9^-/- neutrophils (mean + SEM, n=5 mice per group, 56 [WT] and 54 [S100a9^-/-] cells, two-way ANOVA, Sidak’s multiple comparison). (H) Segmented LFA-1 cluster negative areas (scale bars = 10 μm) and (I) representative confocal images of Ca²⁺ signals in the LFA-1 cluster negative areas (scale bars = 10 μm) (E–I, scale bar color code: 0=black, 255=white). (J) Analysis of cytosolic Ca²⁺ levels in the LFA-1 cluster negative areas in min 0–1, 5–6, and 9–10 of Lyz2xGCaMP5 and Lyz2xGCaMP5xS100a9^-/- neutrophils (mean + SEM, n=5 mice per group, 56 [WT] and 54 [S100a9^-/-] neutrophils, two-way ANOVA, Sidak’s multiple comparison). (K) Representative confocal images showing S100A9 localization at LFA-1 nanocluster areas in stimulated WT neutrophils (scale bars = 10 μm). (L) Quantitative analysis of S100A9 levels in positive LFA-1 nanocluster areas compared to non-LFA-1 nanocluster areas in stimulated WT neutrophils. (mean + SEM, n=3 mice, 26 [WT] neutrophils, paired Student’s t-test). (M) Representative confocal micrographs of LFA-1 nanocluster spatial aggregation in Lyz2xGCaMP5 and Lyz2xGCaMP5xS100a9^-/- neutrophils, within 10 μm² area and minimum 10 LFA-1 nanoclusters considered (≥10 LFA-1 nanoclusters within 10 µm², yellow circles = spatial aggregation area, scale bars = 10 μm). (N) Analysis of spatially aggregated LFA-1 nanoclusters of Lyz2xGCaMP5 and Lyz2xGCaMP5xS100a9^-/- neutrophils (mean + SEM, n=5 mice per group, 56 [WT] and 54 [S100a9^-/-] cells, unpaired Student’s t-test). (O) Segmentation of Lyz2xGCaMP5 and Lyz2xGCaMP5xS100a9^-/- neutrophil area through Lyz2 channel automatic thresholding (scale bars = 10 μm) and (P) representative confocal images of respective F-actin signals (scale bars = 10 μm). (Q) Analysis of F-actin intensity normalized to the cell area in min 0–1, 5–6, and 9–10 of Lyz2xGCaMP5 and Lyz2xGCaMP5xS100a9^-/- neutrophils (mean + SEM, n=5 mice per group, 74 [WT] and 66 [S100a9^-/-] cells, two-way ANOVA, Sidak’s multiple comparison). ns, not significant; *p≤0.05, **p≤0.01, ***p≤0.001.

Figure 4—figure supplement 1.. S100A8/A9 deficient cells display higher frequencies but shorter duration of Ca²⁺ waves compared to WT cells.

(A) Analysis of basal Ca²⁺ levels normalized to the cell area of bone marrow derived Lyz2xGCaMP5 and Lyz2xGCaMP5xS100a9^-/- neutrophils (scale bars = 10 μm) seeded on poly-L-lysine coated slides [mean + SEM, n=234 (WT) and 192 (S100a9^-/-) neutrophils of 5 mice per group, paired Student’s t-test]. (B) Representative western blot images and quantification of total calmodulin levels normalized to GAPDH signal of Lyz2xGCaMP5 and Lyz2xGCaMP5xS100a9^-/- neutrophils [mean + SEM, representative western blot of n=4 mice per group, unpaired Student’s t-test]. (C) Schematic representation of the Ripley’s K used to determine spatial LFA-1 nanocluster correlation. (D) Representative western blot images and quantification of total actin levels (β-Actin) of Lyz2xGCaMP5 and Lyz2xGCaMP5xS100a9^-/- neutrophils [mean + SEM, representative western blot of n≥4 mice per group, unpaired Student’s t-test]. (E) Histogram of Ca²⁺ event frequency distribution and (F) quantification of Ca²⁺ event mean frequency in Lyz2xGCaMP5 and Lyz2xGCaMP5xS100a9^-/- neutrophils [mean + SEM, n=5 mice per group, paired Student’s t-test]. (G) Histogram of Ca²⁺ event duration and (H) quantification of average Ca²⁺ event duration in Lyz2xGCaMP5 and Lyz2xGCaMP5xS100a9^-/- neutrophils [mean + SEM, n=5 mice per group, paired Student’s t-test]. ns, not significant; *p≤0.05, **p≤0.01, ***p≤0.001.

Figure 5.. Cytosolic S100A8/A9 is dispensable for chemokine-induced endoplasmic reticulum (ER) store Ca²⁺ release and for the initial phase of store-operated Ca²⁺ entry (SOCE).

(A) Average flow cytometry kinetic graphs of Ca²⁺ store release in the absence of extracellular Ca²⁺ (Ca²⁺ free medium) in wildtype (WT) and S100a9^-/- neutrophils upon CXCL1 stimulation (traces are shown as mean + SEM, n=5 mice per group). (B) Rapid ER store Ca²⁺ release (MFI _peak/ MFI _0-30s) of WT and S100a9^-/- neutrophils (mean + SEM, n=5 mice per group, paired Student’s t-test). (C) Quantification of Ca²⁺ levels under baseline conditions (MFI _0-30s) (mean + SEM, n=5 mice per group, paired Student’s t-test). (D) Average flow cytometry kinetic graphs of Ca²⁺ influx in the presence of extracellular Ca²⁺ (HBSS medium, 1.5 mM Ca²⁺) of WT and S100a9^-/- neutrophils upon CXCL1 stimulation (traces are shown as mean + SEM, n=5 mice per group, double-headed arrow represents the time points of quantification). (E) Ca²⁺ levels before CXCL1 stimulation (MFI _0-30s) (mean + SEM, n=5 mice per group, paired Student’s t-test). (F) Quantification of ER store Ca²⁺ release and calcium release-activated channel (CRAC) store-operated Ca²⁺ entry (MFI _peak/ MFI _0-30s) (mean + SEM, n=5 mice per group, paired Student’s t-test). (G) Ca²⁺ influx after CXCL1 stimulation, from peak to peak half-life (AUC_{peak – ½ peak}) of WT and S100a9^-/- neutrophils (mean + SEM, n=5 mice per group, paired Student’s t-test). ns, not significant; *p≤0.05, **p≤0.01, ***p≤0.001.

Author response image 1.. Extracellular S100A8/A9 does not rescue the adhesion defect in Mrp14/- neutrophils.

Analysis of number of adherent leukocytes FOV-1 normalized to the WBC of WT and Mrp14-/- mice. Whole blood was harvested through a carotid artery catheter and perfused with a high precision pump at constant shear rate using flow cambers coated with either E-selectin, ICAM-1 and CXCL1 or E-selectin, ICMA-1, CXCL1 and S100A8/A9. [mean + SEM, n=5 mice per group, 12 (WT) and 14 (Mrp14-/-) flow chambers, 2way ANOVA, Sidak’s multiple comparison]. ns, not significant; *p≤0.05, **p≤0.01, ***p≤0.001.

Author response image 2.. Overall Ca2+ levels in WT and Mrp14-/- neutrophils (A) Representative confocal images of neutrophils from WT Lyz2xGCaMP5 and Mrp14-/- Lyz2xGCaMP5 mice, labeled with Lyz2 td Tomato marker.

The images illustrate overall cytosolic Ca2+ levels during neutrophil crawling flow chambers coated with E-selectin, ICAM-1, and CXCL1 (scale bars=10μm). (B) Quantitative analysis of total cytosolic Ca2+ intensity in single cells from WT Lyz2xGCaMP5 and Mrp14-/- Lyz2xGCaMP5 neutrophils measured over three time intervals: min 0-1, 5-6 and 9-10 [mean + SEM, n=5 mice per group, 56 (WT) and 54 (Mrp14-/-) neutrophils, 2way ANOVA, Sidak’s multiple comparison]. (C) Representative traces and (D) single cell analysis of total Ca2+ lifetime over the first 5 minutes in WT Lyz2xGCaMP5 and Mrp14-/- Lyz2xGCaMP5 neutrophils crawling on Eselectin, ICAM-1, and CXCL1 coated flow chambers recorded with FLIM microscopy [mean + SEM, n=3 mice per group, 111 (WT) and 95 (Mrp14-/-) neutrophils, 2way ANOVA, Sidak’s multiple comparison]. ns, not significant; *p≤0.05, **p≤0.01, ***p≤0.001.

Author response image 3.. Giemsa staining of extravasated leukocyte subsets.

(A) Representative image of Giemsa-stained cremaster muscle tissue post-TNF stimulation. The image clearly differentiates leukocyte subsets (white arrow = neutrophils, yellow arrow = eosinophils, red arrow = monocytes). Scale bar = 50µm.

Author response image 4.. Static Transmigration assay.

(a) Transmigration of WT and Mrp14-/- neutrophils in static transwell assays (3um pore size, 45min migration time) showing spontaneously migration (PBS) or migration towards CXCL1. [mean + SEM, n=3 mice per group, 2way ANOVA, Sidak’s multiple comparison]. ns, not significant; *p≤0.05, **p≤0.01, ***p≤0.001.