Schematic overview of the C-terminal tail region of 23 chemokine receptors.

Potential serine/threonine phosphorylation sites are colored red, and serine/threonine clusters are marked in grey.

CCR7-stimulated βarr recruitment, receptor internalization, and Gi/o signaling.

(A) Schematic representation of the experimental design used to monitor EbBRET between RlucII-βarr1/2 and rGFP-CAAX or RlucII-βarr1/2 and rGFP-Rab5 upon chemokine stimulation of CCR7. (B-C) EbBRET signal between (B) RlucII-βarr1 or (C) RlucII-βarr2 recruitment to plasma membrane-anchored rGFP-CAAX or early endosome-anchored rGFP-Rab5 in response to 100 nM CCL19, 100 nM CCL21, or vehicle control stimulation. Data represent the mean ± SE of N=5 experiments. (D) Schematic representation of the experimental design used to monitor CCR7 internalization by detecting loss of luminescence generated by CCR7-SmBiT and LgBiT-CAAX or endosomal CCR7 translocation by measuring gain of luminescence by CCR7-SmBiT and LgBiT-FYVE. (E) Change in luminescence signal generated between CCR7-SmBiT and LgBiT-CAAX in response to 100 nM CCL19 or 100 nM CCL21. The response to chemokine stimulation was normalized to vehicle control. (F) Area under the curve (AUC) was used to calculate the total internalization response for each chemokine ligand. Data represent the mean ± SE of N=3 experiments. (G) Change in luminescence signal generated between CCR7-SmBiT and LgBiT-FYVE in response to 100 nM CCL19 and 100 nM CCL21. The response to chemokine stimulation was normalized to vehicle control. (H) Area under the curve (AUC) was used to calculate the total internalization response for each chemokine ligand. Data represent the mean ± SE of N=4 experiments. (I) HEK293-CCR7 cells transiently expressing the real-time cAMP sensor CAMYEL were challenged with 10 μM forskolin (or vehicle buffer) to increase cAMP production. 5 min later, the cells were stimulated with 100 nM CCL19, 100 nM CCL21, or vehicle buffer, and inhibition of cAMP production was followed as an indirect measurement of Gi/o activation. (J) AUC was used to calculate the total cAMP for each chemokine ligand. Data represent the mean ± SE of N=3-4 experiments. (F, H, and J) One-way ANOVA or (B and C) two-way ANOVA with (B, C, F, H) Turkey’s or (J) Sidak’s multiple comparison post hoc tests were performed to determine statistical differences between the distinct conditions (*p < 0.05; **p < 0.01; ***p <0.001; ****p < 0.0001).

Chemokine-induced CCR7–Gi/o–βarr complex formation.

(A) Schematic illustration of the working hypothesis. CCR7-mediated G protein signaling does not appear to be affected by βarr recruitment or receptor internalization. Therefore, we hypothesized that CCR7 associates and internalizes with βarr in the ‘tail’ conformation where βarr does not block the G protein-binding site within CCR7. As this site is available, CCR7 can interact simultaneously with G protein and βarr to form a CCR7–Gi/o–βarr megaplex, which enables the receptor to stimulate G proteins while being internalized into endosomes. (B) EbBRET signal between RlucII-βarr1ΔFL recruitment to plasma membrane-anchored rGFP-CAAX or endosomally-anchored rGFP-Rab5 in response to 100 nM CCL19, 100 nM CCL21, or vehicle control stimulation. Data represent the mean ± SE of N=5 experiments. (C) Schematic representation of the experimental design used to monitor luminescence upon proximity between SmBiT-βarr1 and LgBiT-miniG protein in response to CCR7 activation. (D) Change luminescence measured upon stimulation of HEK293-CCR7 cells expressing SmBiT-βarr1 and LgBiT-miniGi in response to 100 nM CCL19 or 100 nM CCL21 stimulation. The response to chemokine stimulation was normalized to vehicle control. (E) Area under the curve (AUC) was used to calculate the total response for each chemokine ligand. Data represent the mean ± SE of N=4 experiments. (F) Confocal microscopy imaging displaying HEK293 cells co-expressing, CCR7, βarr2-Strawberry and Halo-miniGi protein. In the experiment the cells were treated with either 100 nM CCL19, 100 nM CCL21, or vehicle control for 30 min. (G) Co-localization quantification analysis of βarr2-Strawberry and Halo-miniGi from the confocal microscopy images. (G) Student’s t test, (E) one-way ANOVA, or (B) two-way ANOVA with Turkey’s multiple comparison post hoc tests were performed to determine statistical differences between the distinct conditions (*p < 0.05; **p < 0.01; ***p < 0.001; ****p < 0.0001).

Endosomal Gi/o activation by internalized CCR7.

(A) Schematic representation of the experimental design used to measure luminescence upon proximity between Rap1GAP-SmBiT and LgBiT-FYVE in response to CCR7 activation. (B) Change in luminescence measured upon stimulation of HEK293-CCR7 cells expressing Rap1GAP-SmBiT and LgBiT-FYVE in response to 100 nM CCL19 or 100 nM CCL21 stimulation. The response to chemokine stimulation was normalized to vehicle control. (C) Area under the curve (AUC) was used to calculate the total response for each chemokine ligand. Data represent the mean ± SE of N=4 experiments. (D) Schematic representation of the EbBRET-based assay to monitor proximity between RlucII-miniGi and the endosomal marker rGFP-Rab5 upon CCR7 activation from endosomes. (E) EbBRET measurements from CCR7-expressing HEK293 cells co-transfected with RlucII-miniGi and rGFP-Rab5 upon stimulation with 100 nM CCL19, 100 nM CCL21, or vehicle control. Data represents the mean ± SE from N=5 independent experiments. (C) Confocal microscopy imaging displaying CCR7-expressing HEK293 cells co-transfected with the plasma membrane marker RFP-Lck and Halo-miniGi. The cells were stimulated with 100 nM CCL19, 100 nM CCL21, or vehicle control for 10 min. (D) Co-localization quantification analysis of RFP-Lck and Halo-miniGi from the confocal microscopy images. (E) Confocal microscopy imaging displaying CCR7-expressing HEK293 cells co-transfected with the endosomal marker RFP-EEA1 and Halo-miniGi. The cells were stimulated with 100 nM CCL19, 100 nM CCL21, or vehicle control for 30 min. (F) Co-localization quantification analysis of RFP-EEA1 and Halo-miniGi from the confocal microscopy images. (C, E, G, and I) One-way ANOVA with Turkey’s multiple comparison post hoc tests were applied to determine statistical differences between the distinct treatments (*p < 0.05; **p < 0.01; ***p < 0.001; ****p < 0.0001).

APEX2-mediated biotinylation of proteins in proximity of CCR7.

(A) Schematic illustration of the workflow behind CCR7-APEX2-mediated biotinylation created with BioRender.com. First, the cells are loaded with biotin-tyramide (biotin-T) followed by stimulation of CCR7 by 100 nM CCL19 or CCL21 for 0 min, 2 min, 10 min, or 25 min. For the last 1 min of chemokine-stimulation hydrogen peroxide is added, which initiates the APEX2-mediated oxidation of biotin-tyramide into highly reactive and short-lived radicals. These radicals bind to proteins within close proximity to the APEX2 enzyme (∼20nm), and thus, label proteins that are in complex with or in close proximity chemokine-stimulated CCR7. Next, the cells are lysed and the resulting biotinylated proteins are captured on neutravidin (Neu) beads followed by extensive washing. Finally, all biotin-labeled proteins are eluted, identified, and analyzed by LC-MS. (B) Western blot analysis of HEK293-CCR7-APEX2 cell lysates, which shows that the biotinylation only takes place in the presence of both biotin-tyramide and hydrogen peroxide. Biotinylated proteins were detected using streptavidin-alexa488. (C) Silver staining of the pull-down experiments demonstrating that the enrichment of biotinylated proteins using neutravidin-coated beads is highly specific. In the control sample, the bound fraction only displays neutravidin band on the SDS-PAGE, whereas the labeled sample shows multiples bands of biotinylated proteins.

© 2024, BioRender Inc. Any parts of this image created with BioRender are not made available under the same license as the Reviewed Preprint, and are © 2024, BioRender Inc.

Identification of protein enrichment in the proximal proteome of CCR7 following agonist stimulation.

(A) Heatmap visualizing select proteins with significant increase (p < 0.05 and Log2 fold-increase > 1) in the proximal proteome of CCR7 for at least one timepoint following chemokine stimulation. Data represent the mean of N=4 experiments. The proteins are clustered according to their function with corresponding significance score, which was calculated by combining the absolute value of Log2 fold-change and-Log10 p value as previously suggested61. (B) Schematic representation showing organelle markers fused to APEX2 used for spatial controls. (C) Confocal microscope images of HEK293-CCR7 cells expressing each of the organelle markers (PM-APEX2, ENDO-APEX2, and CYTO-APEX2) at their respective subcellular locations. (D) Volcano plots of enrichment differences of PM vs CYTO, ENDO vs PM, and ENDO vs CYTO, respectively. Proteins represented as green dots are significant in PM-CYTO/ENDO-PM pair for plasma membrane protein, and red dots in ENDO-CYTO/ENDO-PM pair for endosomal proteins. (E) Schematic illustration showing the roles of RhoGAP, RhoGEF, and RhoGDI in regulation of the family of Rho-GTPase function. (F) Differential enrichment of RhoGAP, RhoGEF, RhoGDI and other proteins that regulate Rho-GTPase signaling with significant change in the CCR7 proximal proteome following chemokine stimulation. Student’s t tests were applied to determine statistical differences between unstimulated cells and chemokine stimulation at different time points (*p < 0.05; **p < 0.01; ***p < 0.001) (G) Interaction network of RhoGAP, RhoGEF, and RhoGDI proteins with significant change in the CCR7 proximal proteome following chemokine stimulation.

Compartmentalized CCR7 signaling and regulation of RhoA, Rac1, and Cdc42 signaling as well as chemotaxis.

(A) Schematic description of the RhoA/Rac1/Cdc42 NanoBiT assay. (B, D, and F) Change in luminescence measured upon stimulation of HEK293-CCR7 cells expressing (B) SmBiT-PKN1/LgBiT-RhoA, (D) SmBiT-Rac1/LgBiT-PAK1, or (F) SmBiT-Cdc42/LgBiT-WAS1 in response to 100 nM CCL19 or 100 nM CCL21 stimulation. The response to chemokine stimulation was normalized to vehicle control. (C, E, and G) Area under the curve (AUC) was used to calculate the total (C) RhoA, (E) Rac1, and (G) Cdc42 activity response for each chemokine ligand. Data represent the mean ± SE of N=4-6 experiments. (H, J, L, and N) HEK293-CCR7 cells were either pre-treated with (H) 100 ng/ml PTX or control buffer for 16 hours, (J) 30 μM of the endocytosis inhibitor Dyngo-4a or the inactive Dyngo control compound for 30 minutes, (L) co-transfected with the dominant negative HA-Dyn-K44A mutant or mock pcDNA3.1 control plasmid, or (N) pre-treated with 10 μM of the endocytosis inhibitor PitStop2 or the inactive PitNot control compound for 30 minutes. (P) HEK293 cells were co-transfected with either wild-type CCR7 or CCR7-ΔST. (H, J, L, N, and P) Changes in luminescence were measured upon stimulation of HEK293-CCR7 cells expressing SmBiT-Rac1/LgBiT-PAK1 in response to 100 nM CCL19 or 100 nM CCL21 stimulation. The response to chemokine stimulation was normalized to vehicle control. (I, K, M, O, and Q) AUC was used to calculate the total Rac1 activity response for each condition. Data represent the mean ± SE of N=4-6 experiments. (C, E, and G) One-way ANOVA or (I, K, M, O, and Q) two-way ANOVA with (C, E, and G) Turkey’s or (I, K, M, O, and Q) Sidak’s multiple comparison post hoc tests were performed to determine statistical differences between the distinct treatments (*p < 0.05; **p < 0.01; ***p < 0.001; ****p < 0.0001).

CCR7-mediated chemotaxis of Jurkat T-cells.

(A) CCR7 stimulation of Gi/o signaling in Jurkat T cells. Jurkat T cells were challenged with 10 μM forskolin (or vehicle buffer) to increase cAMP production either with or without 100 nM CCL19 and 100 nM CCL21. The accumulation of cAMP was determined using the Cisbio cAMP dynamic assay. Data represent the mean ± SE of N=4 experiments. (B) Chemotaxis of Jurkat T cells towards a gradient of 100 nM CCL19, 100 nM CCL21, or vehicle control. Total cells migrated through the transwell chamber were normalized to the CCL19 response. Data represent the mean ± SE of N=8 experiments. (C-F) Chemotaxis of Jurkat T cells that were either pre-treated with (C) 100 ng/ml PTX or control buffer for 16 hours, (D) 30 μM of the endocytosis inhibitor Dyngo-4a or the inactive Dyngo control compound for 30 minutes, € 10 μM of the endocytosis inhibitor PitStop2 or the inactive PitNot control compound for 30 minutes, (F) or 10 μM of the Rac1 inhibitor EHT1864 or DMSO-containing control buffer for 30 minutes. Data represent the mean ± SE of N=6-10 experiments. (A-B) One-way ANOVA or (C-F) two-way ANOVA with (A and C-F) Sidak’s or (B) Turkey’s multiple comparison post hoc tests were performed to determine statistical differences between the distinct treatments (*p < 0.05; **p < 0.01; ***p < 0.001; ****p < 0.0001). (G) Schematic illustration of how endosomal CCR7 signaling promotes chemotaxis.

(A-B) EbBRET signal between (B) RlucII-βarr1 or (C) RlucII-βarr2 recruitment to plasma membrane-anchored rGFP-CAAX or early endosome-anchored rGFP-Rab5 in response to 100 nM CCL19, 100 nM CCL21, or vehicle control stimulation in mock transfected HEK293 cells. Data represent the mean ± SE of N=3 experiments. (C) Schematic illustration of the DiscoverX enzyme fragment complementation assay used to monitor the recruitment of βarr2-EA (EA; N-terminal deletion mutant of β-galactosidase enzymatic acceptor) to CCR7-PK (PK; small enzyme donor ProLink™) upon chemokine stimulation. (D) βarr2 recruitment to CCR7 in HEK293 cells upon 100 nM CCL19, 100 nM CCL21 or vehicle control using the DiscoverX assay. Data represent the mean ± SE of N=3 experiments. (E-F) Change in luminescence signal generated between V2R-SmBiT and (E) LgBiT-CAAX or (F) LgBiT-FYVE in response to 100 nM CCL19 or 100 nM CCL21. The response to chemokine stimulation was normalized to vehicle control. Representative data of N=3 experiments. (G) Mock transfected HEK293 cells expressing the real-time cAMP sensor CAMYEL were challenged with 10 μM forskolin (or vehicle buffer) to increase cAMP production. 5 min later, the cells were stimulated with 100 nM CCL19, 100 nM CCL21, or vehicle buffer, and inhibition of cAMP production was followed as an indirect measurement of Gi/o activation. (H) AUC was used to calculate the total cAMP for each chemokine ligand. Data represent the mean ± SE of N=4 experiments. (D and H) One-way ANOVA with (D) Turkey’s or (H) Sidak’s multiple comparison post hoc test was performed to determine statistical differences between the distinct conditions (*p < 0.05; ****p < 0.0001).

(A) EbBRET signal between RlucII-βarr1ΔFL recruitment to plasma membrane-anchored rGFP-CAAX or endosomally-anchored rGFP-Rab5 in response to 100 nM CCL19, 100 nM CCL21, or vehicle control stimulation in mock transfected HEK293 cells. Data represent the mean ± SE of N=3 experiments. (B) Change in luminescence measured upon stimulation of HEK293-CCR7 cells co-expressing SmBiT-βarr1 and either LgBiT-miniGi or LgBiT-miniGs in response to 100 nM CCL19 stimulation. The response to CCL19 was normalized to vehicle control. (C) Area under the curve (AUC) was used to calculate the total response for each chemokine ligand. Data represent the mean ± SE of N=4 experiments, and one-way ANOVA with Dunnett’s multiple comparison post hoc test was performed to determine statistical differences between the distinct treatments (***p < 0.01; ****p < 0.0001). Change in luminescence signal generated between SmBiT-βarr1 and LgBiT-miniGi in response to 100 nM CCL19 or 100 nM CCL21 stimulation in mock transfected HEK293 cells. The response to chemokine stimulation was normalized to vehicle control. Representative data of N=3 experiments.

(A) Change in luminescence measured upon stimulation of HEK293-CCR7 cells expressing Rap1GAP-SmBiT and LgBiT-FYVE in response to 100 nM CCL19 or 100 nM CCL21 stimulation. The response to chemokine stimulation was normalized to vehicle control. The cells had been pre-incubated with 100 ng/ml PTX or vehicle control for 16 hours. (B) Area under the curve (AUC) was used to calculate the total response for each chemokine ligand. Data represent the mean ± SE of N=4 experiments, and two-way ANOVA with Sidak’s multiple comparison post hoc test was performed to determine statistical differences between the distinct conditions (***p < 0.001; ****p < 0.0001). (C) Change in luminescence measured upon stimulation of mock transfected HEK293 cells expressing Rap1GAP-SmBiT and LgBiT-FYVE in response to 100 nM CCL19 or 100 nM CCL21 stimulation. The response to chemokine stimulation was normalized to vehicle control. Representative data of N=3 experiments.

CCR7-APEX2 stimulation of Gi/o signaling. HEK293-CCR7-APEX2 cells were challenged with 10 μM forskolin (or vehicle buffer) to increase cAMP production either with or without 100 nM CCL19 followed by cAMP determination using the Cisbio cAMP dynamic assay. Data represent the mean ± SE of N=3 experiments, and student’s t test was performed to determine statistical difference between the forskolin and forskolin/CCL19 treatments (***p < 0.001).

Identification of enrichment in proximal proteome of CCR7 following agonist stimulation. (A) Heatmap visualizing proteins with significant change (p < 0.05 and Log2 fold-change > 1) in the proximal proteome of CCR7 for at least one timepoint following chemokine stimulation. (B) Volcano plots showing changes in CCR7 proximity proteome following agonist stimulation. Total of 5582 proteins were analyzed by Student’s t test against the MS data from the unstimulated samples.

Western blot analysis of HEK293 cells transfected with PM-APEX2, ENDO-APEX2, or CYTO-APEX2 shows that the biotinylation only takes place in the presence of both biotin-tyramide and hydrogen peroxide, and that they have different proximity proteome. Biotinylated proteins were detected using streptavidin-alexa488.

(A-C) Change in luminescence measured upon stimulation of HEK293 cells expressing (A) SmBiT-PKN1/LgBiT-RhoA, (B) SmBiT-Rac1/LgBiT-PAK1, or (C) SmBiT-Cdc42/LgBiT-WAS1 in response to 100 nM CCL19 or 100 nM CCL21 stimulation. The response to CCL19 and CCL21 was normalized to vehicle control. Representative data of N=3 experiments. (D, F, and I) Change in luminescence signal in HEK293 cells between CCR7-SmBiT and LgBiT-CAAX in response to 100 nM CCL19. The response to CCL19 stimulation was normalized to vehicle control. The experiments were conducted in the presence of (D) 30 μM Dyngo-4a or the inactive Dyngo control compound, (F) overexpression of the dominant negative HA-Dyn-K44A mutant or mock transfection, (I) or pre-treatment of 10 μM PitStop2 or the inactive PitNot control compound. (K and M) Luminescence signal between LgBiT-CAAX and either wild-type CCR7-SmBiT or CCR7-ΔST-SmBiT mutant in response to (K) 100 nM CCL19 or (M) at resting state. (E, G, J, and L) Area under the curve (AUC) was used to calculate the total internalization response. (E, G, J, L, and M) Data represent the mean ± SE of N=3-4 experiments, and student’s t tests were performed to determine statistical differences between the distinct conditions (**p<0.01; ***p < 0.001; ****p<0.0001). (H) Western blot analysis of HEK293 cells transfected with the CCR7-SmBiT/LgBiT-CAAX NanoBiT biosensors, and pcDNA3.1 (left lane) or HA-Dyn-K44A (right lane). Expression of HA-tagged proteins was detected using a primary anti-HA antibody (upper panel) and the expression of β-tubulin was visualized using a primary anti-β-tubulin antibody.