Interference with glycosaminoglycan-chemokine interactions with a probe to alter leukocyte recruitment and inflammation in vivo

Abstract

In vivo leukocyte recruitment is not fully understood and may result from interactions of chemokines with glycosaminoglycans/GAGs. We previously showed that chlorite-oxidized oxyamylose/COAM binds the neutrophil chemokine GCP-2/CXCL6. Here, mouse chemokine binding by COAM was studied systematically and binding affinities of chemokines to COAM versus GAGs were compared. COAM and heparan sulphate bound the mouse CXC chemokines KC/CXCL1, MIP-2/CXCL2, IP-10/CXCL10 and I-TAC/CXCL11 and the CC chemokine RANTES/CCL5 with affinities in the nanomolar range, whereas no binding interactions were observed for mouse MCP-1/CCL2, MIP-1α/CCL3 and MIP-1β/CCL4. The affinities of COAM-interacting chemokines were similar to or higher than those observed for heparan sulphate. Although COAM did not display chemotactic activity by itself, its co-administration with mouse GCP-2/CXCL6 and MIP-2/CXCL2 or its binding of endogenous chemokines resulted in fast and cooperative peritoneal neutrophil recruitment and in extravasation into the cremaster muscle in vivo. These local GAG mimetic features by COAM within tissues superseded systemic effects and were sufficient and applicable to reduce LPS-induced liver-specific neutrophil recruitment and activation. COAM mimics glycosaminoglycans and is a nontoxic probe for the study of leukocyte recruitment and inflammation in vivo.

Figure 1. Binding of mouse chemokines to COAM and heparan sulphate.

Concentration ranges from 200of mouse (A) GCP-2_9–78, (B) KC/CXCL1, (C) MIP-2/CXCL2, (D) IP-10/CXCL10, (E) I-TAC/CXCL11 and (F) RANTES/CCL5 was run over SA sensor chips surfaces onto which biotinylated COAM and heparan sulphate (HepS) were immobilized. Binding was measured by SPR technology and the resulting experimental sensorgrams are shown in gray. For curve fitting, shown as black lines, the following concentrations were used in two-fold dilution series: 50–6.25 nM for GCP-2_9–78, 25–3.13 nM for KC/CXCL1, 200–25 nM for MIP-2/CXCL2, 6.25–0.78 nM for IP-10/CXCL10, 50–6.25 nM for I-TAC/CXCL11 and 400–50 nM for RANTES/CCL5.

Figure 2. In vivo recruitment of neutrophils into the peritoneal cavity.

Mice received an i.p. dose of 1-2_(7–98), or a mixture of both COAM (1 mg) and mouse GCP-2_(7–98) (100 ng). After 1 h (A) or 4 h post-treatment (B), peritoneal lavage fluids were collected and the percentages and absolute numbers of neutrophils, recognized as CD11b and Ly6G double positive cells, determined by FACS analysis, are shown. The net numbers of CD11b⁺ Ly6G⁺ cells were determined by multiplying the percentages of CD11b⁺ Ly6G⁺ cells with total peritoneal leukocyte counts. Histograms and dots represent group medians and spreading of individual data points from each mouse, respectively. *P<0.05, **P<0.01, ***P<0.001, as determined by Kruskal-Wallis test.

Figure 3. In vivo recruitment of neutrophils to the cremaster muscle.

Mice received an intrascrotal dose of 0.2(A and B) or 24 (C and D) h prior to induction of anesthesia, surgical preparation of the cremaster muscle and onset of MIP-2/CXCL2 superperfusion. The number of adherent neutrophils (A and C) were quantified within a 100 µm length of venule, and the number of emigrated neutrophils within the field of view (B and D) were quantified prior to or following 30, 60 and 90 min of MIP-2/CXCL2 superperfusion. *P<0.05, **P<0.01, ***P<0.001, as determined by students’ t-test.

Figure 4. Effects of COAM co-treatment on LPS-induced systemic inflammation in the liver.

(A) Co-application of COAM (2 mg/mouse) with intraperitoneally administrated LPS (1 mg/kg) decreases neutrophil infiltration to the liver at 4 h of inflammation; (B) representative images of neutrophils present in the liver sinusoids of LPS- and LPS plus COAM-treated mice (green cells – autofluorescent hepatocytes; 20x; scale bars represent 50 µm). Quantification of extracellular neutrophil elastase (C) and histone (D) within the livers of LPS and LPS+COAM-treated animals (mean area of staining per 20×FOV ± SD; scale bars represent 45 µm). Intravital visualization of NET deposition in the liver vasculature of LPS-treated and LPS plus COAM-treated mice (E). Staining for extracellular neutrophil elastase (NE) and histone illustrates clear deposition of these characteristic molecules of NETs in the liver after either treatment. In addition, overlay of histone and elastase staining is shown. Staining for extracellular DNA is presented with a higher magnification to clearly picture Sytox green deposition along the liver sinusoids; areas of the extDNA deposition are marked with red arrows. Neutrophil, elastase and histones were measured in five FOV/mouse, n = 5–7 animals per group; *P<0.05.