T cell extravasation: demonstration of synergy between activation of CXCR3 and the T cell receptor

Abstract

Endothelial cells present chemokines to T cells and can also stimulate the T cell antigen receptor by presentation of peptide-MHC antigen complexes. This study was designed to investigate the potential synergy between stimulation of the chemokine receptor CXCR3 and the human T cell receptor complex. Transendothelial T cell migration towards CXCL10 was modified by crosslinking CD3 immediately before addition to the endothelium. When resting endothelium was used, T cells which had been activated by crosslinking CD3 for only 1 min showed a significant reduction (p<0.0001) in migration when compared with untreated T cells. By contrast, endothelial cells which had been activated by stimulation with interferon-gamma and tumour necrosis factor-alpha supported a specific increase in the migration of activated T cells; this was most apparent after CD3 had been activated for 90 min (p<0.0001). The molecular basis for synergy between CXCR3 and the T cell receptor complex was investigated by measurement of fluorescence resonance energy transfer. This showed that CXCL10 induced a close (<10 nm) spatial association between CXCR3 and the CD3epsilon subunit on the cell-surface. These data demonstrate that stimulation of both CXCR3 and the T cell receptor has the potential to enhance specifically both the proliferation and extravasation of specific T cells during episodes of local inflammation.

Fig. 1

Analysis of the function of CXCR3. Chemotactic migration of activated T cells through an EA.hy926 endothelial cell monolayer towards various concentrations of CXCL10. Prior activation of the endothelium with IFNγ and TNFα increased the number of migrant cells. Representative results are shown from one of three similar assays; the bars show mean values ± SEM.

Fig. 2

CXCR3 and CD3 activation have synergistic effects on T cell proliferation. A range of OKT3 concentrations were bound to a 96-well plate overnight. 10 nM CXCL10 was also bound to the plate overnight (-▾-), added in solution to the lymphocyte suspension (-▴-), or controlled with PBS (-■-). The assay was incubated for 72 h and T cell proliferation was determined by measuring ³H-thymidine incorporation. Representative results are shown from one of three similar assays; the bars show mean values ± SEM.

Fig. 3

Validation of the transendothelial T cell migration model. (A) Representative flow cytometric dotplots showing analysis of transendothelial chemotaxis assays. Region P1 contains the total T cell population retrieved from the bottom chamber of a well following transendothelial migration. P2 contains the fluorosphere counting beads added to the sample. P3 contains the Alexa 633 labelled T cells within P1. P4 contains the CFSE labelled T cells within P1. Enumeration of events in these regions allowed the number of CFSE and Alexa 633 labelled cells to be calculated for analysis as described in Section 2. (B) Confocal microscopy showing the migration of T cells through cytokine-stimulated endothelium. By separate use of lasers at 488 nm and 633 nm wavelengths it was possible to visualise CFSE labelled (untreated) T cells (green) and Alexa 633 labelled (CD3 crosslinked) T cells (red) in separate images. Overlaying these images allowed the spatial relationship of both cell populations to be compared at different time points. The representative image was captured after 120 min and shows an Alexa 633 labelled cell in the process of migration through the membrane, whilst CFSE labelled cells remain on the surface of the endothelial cell monolayer. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of the article.)

Fig. 4

The effect of CD3 crosslinking on chemotaxis of activated T cells through an endothelial monolayer. In both graphs, the bars indicate the proportion of Alexa 633 (crosslinked; filled bar) and CFSE labelled (untreated; open bar) T cells. (A) Using resting endothelial cells. A significant reduction in the proportion of Alexa 633 labelled cells was seen in the migrant population, when CD3 crosslinking was performed 1 min before the assay. When CD3 crosslinking was carried out 90 min before the assay there was no significant change in the proportion of these cells in the migrant population. (B) Using endothelial cells which had been previously stimulated with IFNγ and TNFα. Significant increases in the proportion of Alexa 633 labelled cells are seen in the migrant cell populations following CD3 crosslinking for either 1 min or 90 min. In both cases representative results are shown from one of three similar experiments; the bars show mean values ± SEM.

Fig. 5

CXCR3 can be spatially associated with CD3. (A) Mean FRET efficiencies were calculated for titrations of anti-CXCR3-PE and anti-CD3-APC demonstrating that the FRET seen is non-random. Fluorescence intensity is expressed as a percentage of the maximal median fluorescence intensity (MFI). The increasing percentage values corresponded to increasing anti-CD3ɛ antibody concentrations. (B) Control experiments demonstrated that the FRET signal is not due to non-specific spectral overlap (CD45/CXCR3 control), is specific for CXCR3 (CCR2/CD3 control), is ligand-dependent and is not entirely a consequence of receptor internalisation (reduction following acid wash). The bars show mean values ± SEM.