Reduced chemokine presentation by synovial endothelial cells in sdc-3−/−mice. In the same samples as in Figure 1 sections of CXCL1-injected joints of wild-type and sdc-3−/−mice were stained with a CXCL1 antibody using tyramide amplification, and DAPI, and viewed by immunofluorescence. (A and B) Synovial endothelial cells of wild-type mouse joint; CXCL1 occurs as clusters with white arrows showing examples of luminal chemokine and red arrows intracellular or abluminal chemokine. The endothelial cell layer is labelled (e) and the lumen of the blood vessel (L) containing red blood cells. (C) No CXCL1 clusters are present in the endothelial cells of PBS-injected controls (arrows); this image is from a wild-type mouse. ( D ) There are less CXCL1 clusters in endothelial cells of following treatment with heparanase I and III to degrade heparan sulphate prior to immunostaining. (E) is a negative control of the synovium of wild-type mouse in the absence of CXCL1 antibody. Bar = 30 μm in A to E. (F) Quantification of CXCL1 staining shows decrease in the numbers of endothelial CXCL1 clusters per blood vessel in sdc-3−/− (n = 6) compared to wild-type (n = 6) joints. Data are mean ± SEM. ***P = 0.0003. (G) shows the data in (F) expressed as number of CXCL1 clusters at the luminal surface or intracellularly/abluminally in synovial endothelial cells. There is a reduction of the numbers of luminal CXCL1 clusters in sdc-3−/−mice compared to wild type, ***P = 0.0002. Data are means ± SEM. CXCL1, chemokine C-X-C ligand 1; DAPI, 4',6-diamidino-2-phenylindole; PBS, phosphate-buffered saline.