A spatiotemporal map of co-receptor signaling networks underlying B cell activation

Abstract

The B cell receptor (BCR) signals together with a multi-component co-receptor complex to initiate B cell activation in response to antigen binding. Here, we take advantage of peroxidase-catalyzed proximity labeling combined with quantitative mass spectrometry to track co-receptor signaling dynamics in Raji cells from 10 s to 2 h after BCR stimulation. This approach enables tracking of 2,814 proximity-labeled proteins and 1,394 phosphosites and provides an unbiased and quantitative molecular map of proteins recruited to the vicinity of CD19, the signaling subunit of the co-receptor complex. We detail the recruitment kinetics of signaling effectors to CD19 and identify previously uncharacterized mediators of B cell activation. We show that the glutamate transporter SLC1A1 is responsible for mediating rapid metabolic reprogramming and for maintaining redox homeostasis during B cell activation. This study provides a comprehensive map of BCR signaling and a rich resource for uncovering the complex signaling networks that regulate activation.

Keywords: B cell signaling; CP: Immunology; co-receptor; phosphoproteomics; proteomics; proximity labeling.

Figure 1.. Overview of CD19-APEX system

(A) Schematic of CD19-APEX proximity labeling. CD19, the signaling subunit of the B cell co-receptor complex, is tagged with the ascorbate peroxide APEX2 on its C terminus. In the presence of hydrogen peroxide (H₂O₂), APEX2 converts biotin phenol into a biotin phenoxy radical, which then covalently labels proteins within close spatial proximity (about 20 nm) of the APEX2 fusion. (B) Timeline of the eight time points at which CD19-APEX cells were labeled after activation. (C) Principal-component analysis of biological replicates. See also Figures S1 and S2.

Figure 2.. Response patterns to BCR stimulation

(A) Enrichment pattern heatmap of all proteins and phosphosites with a fold change difference of at least 2.5 upon activation, highlighting five distinct patterns of response revealed by hierarchical clustering. Each protein’s enrichment level across different time points is normalized to its maximum enrichment signal. (B–F) Enrichment profile of three exemplary proteins showing (B) an early enrichment response pattern, (C) a late enrichment response pattern, (D) a sustained enrichment response pattern, (E) an early depletion response pattern, and (F) a late depletion response pattern. Error bars represent mean ± SEM of two biological replicates. In all panels, phospho-marks are annotated with the protein name followed by “_amino acid position” of the phospho-mark. See also Figures S3–S5.

Figure 3.. Kinetics of recruitment of core components of the BCR signaling pathway to CD19

(A) Enrichment pattern heatmap of known components of the BCR signaling pathway grouped by pathway. (B) Enrichment profiles of known components of the BCR signaling pathway highlighting the distinct patterns of response of each member. Error bars represent mean ± SEM of two biological replicates, and dashed gray lines represent phospho-marks. (C) Summary of the proteins recruited to CD19 during activation. Time is plotted on a log10 scale for better separation of the early time points. (D) Summary of phosphorylation within the CD19 microenvironment during activation. Time is plotted on a log10 scale for better separation of the early time points. (C) and (D) contain the same data as shown in the heatmap in (A) plotted in different ways to highlight distinct patterns of response. In all panels, phospho-marks are annotated with the protein name followed by “_amino acid position” of the phospho-mark.

Figure 4.. CD19-APEX2 as a resource to uncover previously uncharacterized B cell biology

(A) Enrichment pattern heatmap of proteins involved in cytoskeletal regulation and organization. (B) Enrichment pattern heatmap of the palmitoyltransferase ZDHHC5 and several phosphorylation sites. Error bars represent mean ± SEM of two biological replicates. (C) Enrichment pattern heatmap of syntaxins (STXs). Error bars represent mean ± SEM of two biological replicates. (D) Enrichment profile of SWAP70. Error bars represent mean ± SEM of two biological replicates. (E) Western blot of SWAP70 knockout Raji cells. “P” denotes parental Raji cells. (F) Fold change in CD69 surface staining after activation with anti-IgM F(ab′)2 in parental Raji cells and SWAP70 knockout clones. (G) Fold change in CD86 surface staining after activation with anti-IgM F(ab′)2 in parental Raji cells and SWAP70 knockout clones. For (F) and (G), error bars represent mean ± SEM of three independent experiments. Statistical analysis was performed in GraphPad Prism using an unpaired t test. *p ≤ 0.05; **p ≤ 0.01; ***p ≤ 0.001; ****p ≤ 0.0001; ns, not significant. In all panels, phospho-marks are annotated with the protein name followed by “_amino acid position” of the phospho-mark.

Figure 5.. Recruitment and functional interrogation of the glutamate transporter SLC1A1

(A) Enrichment pattern heatmap of all members of the SLC family identified in the CD19-APEX time course. Each protein’s enrichment level across different time points is normalized to its maximum enrichment signal. (B) Enrichment profile of the glutamate transporter SLC1A1. Error bars represent mean ± SEM of two biological replicates. (C) Glutamate uptake kinetics in parental Raji cells upon activation. (D) Western blot of SLC1A1 knockout Raji cells. “P” denotes parental Raji cells. (E) Glutamate uptake in parental and SLC1A1 knockout Raji cells upon activation for 2 min. (F) Enrichment pattern heatmap of NADPH oxidase 2 (CYBB) and glutamate dehydrogenase 2 GLUD2. Each protein’s enrichment level across different time points is normalized to its maximum enrichment signal. (G) Glutathione levels in parental and SLC1A1 knockout Raji cells in resting and activated cells, measured 48 h after activation. (H) ROS levels in parental and SLC1A1 knockout Raji cells in resting and activated cells, measured 48 h after activation. For (C), (E), (G), and (H), error bars represent mean ± SEM of three independent experiments. Statistical analysis was performed in GraphPad Prism using an unpaired t test. *p ≤ 0.05; **p ≤ 0.01; ***p ≤ 0.001; ****p ≤ 0.0001; ns, not significant. In all panels, phospho-marks are annotated with the protein name followed by “_amino acid position” of the phospho-mark.