Carboxy-terminus of CXCR7 regulates receptor localization and function

Abstract

Chemokine receptor CXCR7 is essential for normal development, and this receptor promotes initiation and progression of diseases including cancer and autoimmunity. To understand normal and pathologic functions of CXCR7 and advance development of therapeutic agents, there is a need to define structural domains that regulate this receptor. We generated mutants of CXCR7 with deletion of different lengths of the predicted intracellular tail and analyzed effects on CXCR7 signaling and function in cell-based assays. While wild-type CXCR7 predominantly localized to intracellular vesicles, progressive deletion of the carboxy terminus redistributed the receptor to the plasma membrane. Truncating the intracellular tail of CXCR7 did not alter binding to CXCL12, but mutant receptors had reduced scavenging of this chemokine. Using a firefly luciferase complementation system, we established that deletions of the carboxy terminus decreased basal interactions and eliminated ligand-dependent recruitment of the scaffolding protein β-arrestin-2 to receptors. Deleting the carboxy terminus of CXCR7 impaired constitutive internalization of the receptor and reduced activation of ERK1/2 by CXCL12-CXCR7. Inhibiting dynamin, a molecule required for internalization of CXCR7, increased ligand-dependent association of the receptor with β-arrestin-2 and enhanced activation of ERK1/2. These studies establish mechanisms of action for CXCR7 and establish the intracellular tail of CXCR7 as a critical determinant of receptor trafficking, chemokine scavenging, and signaling.

Figure 1. Carboxy-terminal truncations redistribute CXCR7 to the cell membrane

A) Confocal fluorescence micrographs of cells transfected with CXCR7-WT-GFP, CXCR7-346-GFP, CXCR7-322-GFP, or CXCR4-GFP, respectively. Scale bar denotes 20 μm. B) Representative line profile analyses of fluorescence intensity and distribution profiles across cells shown in panel A. Absorbance is presented as arbitrary units for fluorescence intensity.

Figure 2. Carboxy-terminal truncations of CXCR7 reduce chemokine scavenging

A) Accumulation of CXCL12-GL after 30 minutes of incubation with 293T cells transiently transfected with CXCR7-WT-GFP, CXCR7-346-GFP, CXCR7-322-GFP, or vector control. Graph displays mean values + SEM for photon flux from CXCL12-GL normalized to photon flux from the firefly luciferase transfection control. *, p < 0.05; **, p < 0.01. B) Confocal microscopy of cells expressing CXCR7-WT-GFP, CXCR7-346-GFP, or CXCR7-322-GFP following 30 minutes of incubation with CXCL12-mCherry. Panels show GFP expression for receptor fusion proteins, fluorescence from CXCL12-mCherry, and a merged image for each condition. Scale bar denotes 20 μm.

Figure 3. Carboxy-terminal truncation mutants of CXCR7 co-localize with wild-type CXCR7 without affecting chemokine scavenging

A) Confocal microscopy of cells expressing CXCR7-WT-GFP, CXCR7-346-GFP, or CXCR7-322-GFP co-transfected with CXCR7-WT-mCherry. Panels show separate green and red fluorescent images for each condition. Arrows in merged images show sites of co-localization for both receptors depicted as yellow foci. Scale bar denotes 20 μm. B) Cells co-transfected with CXCR7-WT-mCherry and CXCR7-WT-GFP, CXCR7-346-GFP, CXCR7-322-GFP, or unfused GFP were incubated for 30 minutes with CXCL12-GL as in figure 2. Graph displays mean values + SEM. *, p < 0.05; **, p < 0.01.

Figure 4. Carboxy-terminus of CXCR7 regulates basal and inducible association with β-arrestin 2

Graph shows mean values + SEM for photon flux from firefly luciferase complementation between combinations of β-arrestin-2-CFLuc or c-fos-CLuc with CXCR7-WT, CXCR7-346, or CXCR7-322 fusions with NFLuc after 30 minutes of incubation with 0, 3, or 10 nM CCX733. Photon flux from firefly luciferase was normalized to photon flux from co-transfected Gaussia luciferase as a control for transfection efficiency. *, p <0.05.

Figure 5. CXCL12 binding to CXCR7 is not impaired by truncating the carboxy-terminus of the receptor

A) MDA-MB-231 cells expressing NGLuc-CXCR7-WT or NGLuc-CXCR7-322 were co-cultured with 231 cells secreting CXCL12-CGLuc for increasing periods of time. Graph shows mean values + SEM for ratios of photon flux for CXCL12-CGLuc complementation with NGLuc-CXCR7-322 relative to NGLuc-CXCR7-WT before (open bars) or after (filled bars) acid washing. **, p < 0.01. B) Co-cultures of 231-NGLuc-CXCR7-WT or 231-CXCR7-NGLuc-322 cells with 231-CXCL12-CGLuc were incubated with increasing concentrations of CCX771 for 4 hours before quantifying photon flux from Gaussia luciferase complementation. Graph shows mean values ± SEM. C) Co-cultured cells were incubated with 100 nM CCX771 for increasing periods of time through 6 hours. Graph shows mean values for photon flux from Gaussia luciferase complementation. Error bars are smaller than the symbols.

Figure 6. CXCR7-322 mutation impairs receptor internalization

A) Graph shows mean values + SEM for relative internalization of CXCR7-WT or CXCR7-322 receptors from the cell membrane after 5 or 15 minutes. Receptors were analyzed by flow cytometry. *, p < 0.05.

Figure 7. CXCR7-322 blocks activation of ERK1/2 in response to CXCL12

A) Western blot of 293T cells transfected with CXCR7-GFP or CXCR7-322-GFP following treatment with 100 ng/ml CXCL12 for various periods of time through 30 minutes. Blots show phosphorylated and total ERK1/2, respectively. The gel format minimized separation between bands for ERK1 and ERK2. B) Graph displays relative intensities of phosphorylated to total ERK1/2 for each samples expressed as ratios of arbitrary units.

Figure 8. Inhibiting dynamin eliminates sustained activation of ERK1/2 by CXCR7

A) Confocal microscope images of 293T cells transfected transiently with CXCR7-WT-GFP along with dominant negative K44A dynamin (K44A) or control plasmid. Scale bar denotes 20 μm. B) Representative line profile analyses of fluorescence intensity and distribution across cells depicted in panel A. C) Graph depicts bioluminescence from interaction of CXCR7-NLuc with β-arrestin-2-CLuc in transiently transfected 293T cells co-transfected with dominant negative K44A dynamin or control plasmid. Cells were incubated with 100 ng/ml CXCL12 for 30 minutes before quantifying bioluminescence. Data were graphed as mean values + SEM for firefly luciferase photon flux normalized to co-transfected Gaussia luciferase. *, p < 0.05; **, p < 0.01. D) Western blot of phosphorylated and total ERK1/2 in 293T cells transfected with CXCR7-WT-GFP with K44A dynamin or control plasmid as described in panel A. Cells were treated with 100 ng/ml CXCL12 for 0, 5, 15, or 30 minutes as shown. E) Graph displays relative band intensities for phosphorylated to total ERK1/2 as quantified by Image J.