Age-related changes in the local milieu of inflamed tissues cause aberrant neutrophil trafficking and subsequent remote organ damage

Abstract

Aging is associated with dysregulated immune functions. Here, we investigated the impact of age on neutrophil diapedesis. Using confocal intravital microscopy, we found that in aged mice, neutrophils adhered to vascular endothelium in inflamed tissues but exhibited a high frequency of reverse transendothelial migration (rTEM). This retrograde breaching of the endothelium by neutrophils was governed by enhanced production of the chemokine CXCL1 from mast cells that localized at endothelial cell (EC) junctions. Increased EC expression of the atypical chemokine receptor 1 (ACKR1) supported this pro-inflammatory milieu in aged venules. Accumulation of CXCL1 caused desensitization of the chemokine receptor CXCR2 on neutrophils and loss of neutrophil directional motility within EC junctions. Fluorescent tracking revealed that in aged mice, neutrophils undergoing rTEM re-entered the circulation and disseminated to the lungs where they caused vascular leakage. Thus, neutrophils stemming from a local inflammatory site contribute to remote organ damage, with implication to the dysregulated systemic inflammation associated with aging.

Keywords: ACKR1; CXCR2; Neutrophils; aging; chemokines; diapedesis; endothelium; extravasation; inflammation; mast cells.

Graphical abstract

Figure 1

Inflamed aged stroma promotes aberrant neutrophil transendothelial cell migration (A–G) Young (2–4 months) and aged (≥18 months) mice were treated intrascrotally (i.s.) with PBS or IL-1β and neutrophil responses in cremasteric post-capillary venules analyzed. Leukocyte (A) rolling flux and (B) firm adhesion in WT mice as quantified by brightfield IVM (n = 3–16 mice/group). Neutrophil (C) normal TEM events (n = 5-7 mice/group; Video S1), (D) total extravasation (n = 5–7 mice/group), and (E) related representative images of Lyz2-EGFP-ki venules, as assessed by confocal IVM (scale bar: 15 μm). (F) Time-lapse confocal images (Video S2) showing a neutrophil rTEM event in an IL-1β-stimulated aged Lyz2-EGFP-ki venule with the neutrophil in the sub-endothelial space (t = 17 min) re-entering the vascular lumen (t = 26 min to t = 46 min). Top panel: en face luminal view; bottom panel: cross-sections; arrows: direction of neutrophil motility (scale bar: 10 μm). (G) Frequency of neutrophil rTEM in Lyz2-EGFP-ki stimulated tissues (n = 5–6 mice/group). (H) The generation of Y→Y, A→Y, or Y→A chimeras (young ‘Y’; or aged ‘A’) and (I–K) their analysis post treatment with i.s. PBS or IL-1β. Cremaster muscle (I) leukocyte firm adhesion as assessed by brightfield IVM (n = 3-10 mice/group), (J) neutrophil normal TEM events (n = 4-5 mice/group) and (K) frequency of neutrophil rTEM as assessed by confocal IVM (n = 3-5 mice/group). Means ± SEM, #p < 0.05, ##p < 0.01, ###p < 0.001, ####p < 0.0001 relative to aged-matched controls and ^∗p < 0.05, ^∗∗p < 0.01, ^∗∗∗p < 0.001, ^∗∗∗∗p < 0.0001, n.s. not significant, as indicated. See also Figure S1.

Figure 2

CXCL1 drives aging-associated neutrophil reverse TEM (A–D) Young and aged mice were treated i.s. with PBS or IL-1β. (A) Inflammatory mediator analysis in homogenized cremaster muscles as assayed by protein array (n = 3 mice/condition). (B) CXCL1 levels in cremaster muscles (n = 4-7 mice/group) or (C) plasma (n = 4–8 mice/group) as quantified by ELISA. (D) Frequency of neutrophil rTEM in Y→Y or Y→A chimeras (generated as detailed in Figure 1H) treated i.v. with isotype control, anti-CXCL1 or anti-CXCL2 blocking mAbs (n = 3–5 mice/group). (E) Representative confocal images of mast cells (MCs; Avidin) associated with post-capillary venules (CD31) in young and aged unstimulated WT cremaster muscles (scale bar: 20 μm) and quantification in (F) cremaster muscles, and (G) ear skin (n = 5-7 mice/group). (H–I) Analysis of CXCL1 expression in MCs of young and aged IL-1β-stimulated cremasteric tissues by confocal microscopy with (H) showing representative images and (I) quantification by MFI (scale bar: 5 μm; n = 3–7 mice/group). (J) Representative confocal images of MCs (CD117) in young and aged unstimulated WT ear skin (scale bar: 10 μm) and quantification of MC volume (n = 4 mice/group). (K) Peritoneal MCs acquired from unstimulated young and aged mice assayed for SA-β-galactosidase activity by flow cytometry (n = 6-13 mice/group). (L) Frequency of neutrophil rTEM in control and MC depleted IL-1β-stimulated cremaster muscles of aged chimeras (see Figure 1H; n = 4-5 mice/group). (M) Frequency of neutrophil rTEM in IL-1β-stimulated ear skin of aged MC deficient (Mcpt5-Cre-R-DTA) mice and littermate controls (n = 5 mice/group). Means ± SEM, #p < 0.05, ##p < 0.01, ###p < 0.001, ####p < 0.0001 relative to controls, ^∗p < 0.05, ^∗∗p < 0.01, ^∗∗∗p < 0.001, ^∗∗∗∗p < 0.0001 as indicated. See also Figure S2.

Figure 3

ACKR1 is elevated in aged tissues and retains mast cell-derived CXCL1 at EC junctions (A–G) Young and aged WT mice were treated i.s. with PBS or IL-1β and cremaster muscles analyzed by confocal microscopy. (A) Representative confocal images of post-capillary venules (PCVs) immunostained for CD31 and CXCL1 (scale bar: 4 μm; dashed boxes delineate magnified areas) and (B) quantification of CXCL1 expression (MFI) at EC junctional (junc.) and non-junctional (non-junc.) regions (n = 6-7 mice/group). (C) EC CXCL1 expression (MFI) in control and mast cell-depleted aged cremaster tissues (n = 3-5 mice/group). (D) Representative confocal images illustrating ACKR1 expression in PCVs (CD31; scale bar: 10 μm) and ACKR1 quantification (MFI) within (E) whole ECs, and EC (F) non-junctional or (G) junctional regions (n = 3 mice/group). (H) Generation of EC Ackr1^+/+ and EC Ackr1^−/− chimeras. (I-K) Young and aged chimeras as generated in (H) were treated i.s. with IL-1β. (I) Representative confocal images of cremasteric PCVs immunostained for CD31 and CXCL1 (scale bar: 4 μm), (J) quantification of CXCL1 expression (MFI) within EC junctional and non-junctional regions (n = 3-8 mice/group) and (K) plasma CXCL1 as quantified by ELISA (n = 3-8 mice/group). Means ± SEM, #p < 0.05, ##p < 0.01, ###p < 0.001 relative to controls, ^∗p < 0.05, ^∗∗p < 0.01, n.s. not significant, as indicated. See also Figure S3.

Figure 4

GRK2-dependent CXCR2 downregulation promotes neutrophil rTEM in aged tissues. (A–F) Young and aged mice were treated i.s. with IL-1β. (A) Representative confocal images of cremasteric post-capillary venules (PCVs) of WT mice immunostained for CXCR2, MRP14 (neutrophils) and CD31. Arrows indicate CXCR2^lo neutrophils (scale bar: 10 μm; dashed boxes delineate magnified areas). (B) Percentage of luminal CXCR2^lo neutrophils in cremasteric PCVs of EC Ackr1^+/+ and EC Ackr1^−/− chimeras (n = 3-5 mice/group). (C) Generation of neutrophil Grk2^+/+ and Grk2^−/− chimeras. (D–F) Young and aged chimeras as generated in (C) were treated i.s. with IL-1β. (D) Percentage of luminal CXCR2^lo neutrophils (n = 3-4 mice/group). (E) Total neutrophil TEM events and (F) frequency of neutrophil rTEM as assessed by confocal IVM (n = 3-4 mice/group). Means ± SEM #p < 0.05, ####p < 0.0001 as compared to young, ^∗p < 0.05, ^∗∗∗p < 0.001 as indicated. See also Figure S4.

Figure 5

rTEM neutrophils stemming from locally injured aged tissues accumulate in the lungs. Young and aged mice were subjected to sham or cremasteric IR injury. (A) Representative confocal images of post-capillary venules (PCVs) immunostained for CD31 and MRP14 (neutrophils) in WT mice (scale bar: 20 μm). (B) Representative confocal images and quantification of lung vascular leakage in WT mice 4 h post reperfusion (scale bar: 20 μm; n = 4-5 mice/group). (C) Neutrophil normal TEM events and (D) frequency of neutrophil rTEM in Y→Y or Y→A chimeras (see Figure 1H) as assessed by confocal IVM (n = 6 mice/group). (E-I) Mice were injected i.v. with a biotinylated anti-Ly6G mAb and AF647-Strept locally applied to the cremaster muscle. (E) Time-lapse confocal IVM images (Video S4) of a neutrophil rTEM event in an aged Lyz2-EGFP-ki cremaster muscle during IR injury illustrating that the neutrophil exhibiting rTEM is AF647-Strept^hi (Top panel: en face luminal view; bottom panel: isolated neutrophil; scale bar: 4 μm). (F-I) Representative flow cytometry profiles and frequency of AF647-Strept^hi neutrophils in (F-G) blood and (H-I) pulmonary vascular washouts in WT mice (n = 4-11 mice/group). Numbers indicate the percentage of AF647-Strept^hi neutrophils. Means ± SEM, #p < 0.05, ###p < 0.001, ####p < 0.0001 as compared to age-matched controls, ^∗p < 0.05, ^∗∗∗p < 0.001, ^∗∗∗∗p < 0.0001 as indicated. See also Figure S5.

Figure 6

rTEM neutrophils are programmed toward an activated state in aged lungs and are directly noxious to the lung tissue (A–C) Young and aged WT mice were injected i.v. with biotinylated anti-Ly6G mAb, subjected to sham or cremasteric IR injury and AF647-Strept applied locally to the cremaster muscle. Expression levels of indicated markers on AF647-Strept^hi neutrophils relative to levels on AF647-Strept^lo neutrophils within the pulmonary vasculature (A) 1 h or (B) 4 h post-reperfusion (n = 5–9 mice/group) and (C) representative histograms of indicated markers on pulmonary vascular neutrophils of aged mice 4 h post-reperfusion. (D) Flow cytometry sorting strategy of AF647-Strept^lo and AF647-Strept^hi neutrophils from whole blood of young or aged mice 1 h post-reperfusion and subsequent i.v. injection into naive young or aged mice. (E–G) Extravasation of i.v. Evans blue in lung tissue in (E) aged recipients 4 h post i.v. injection of PBS or neutrophils sorted from young donors, (F) aged recipients 4 h or 24 h post i.v. injection of neutrophils sorted from aged donors, and (G) young recipients 4 h post i.v. injection of PBS or neutrophils sorted from aged donors (n = 4–7 mice/group). Means ± SEM. ^∗p < 0.05, ^∗∗p < 0.01, ^∗∗∗∗p < 0.0001, n.s. not significant as indicated or as compared to AF647-Strept^lo neutrophils of the same group. See also Figures S6 and S7.

Figure 7

CXCL1 blockade protects aged mice from excessive lung injury (A) Representative confocal images of whole mount lung (scale bar: 20 μm) and (B) lung vascular leakage quantification in young and aged WT mice subjected to sham or cremasteric IR injury and treated with an isotype control or anti-CXCL1 blocking mAb (n = 4 mice/group). Means ± SEM. ^∗p < 0.05, ^∗∗p < 0.01 as indicated.