Precise Temporal Profiling of Signaling Complexes in Primary Cells Using SWATH Mass Spectrometry

Abstract

Spatiotemporal organization of protein interactions in cell signaling is a fundamental process that drives cellular functions. Given differential protein expression across tissues and developmental stages, the architecture and dynamics of signaling interaction proteomes is, likely, highly context dependent. However, current interaction information has been almost exclusively obtained from transformed cells. In this study, we applied an advanced and robust workflow combining mouse genetics and affinity purification (AP)-SWATH mass spectrometry to profile the dynamics of 53 high-confidence protein interactions in primary T cells, using the scaffold protein GRB2 as a model. The workflow also provided a sufficient level of robustness to pinpoint differential interaction dynamics between two similar, but functionally distinct, primary T cell populations. Altogether, we demonstrated that precise and reproducible quantitative measurements of protein interaction dynamics can be achieved in primary cells isolated from mammalian tissues, allowing resolution of the tissue-specific context of cell-signaling events.

Keywords: DIA; GRB2; SWATH; interactome; primary T cells; targeted mass spectrometry.

Graphical abstract

Figure 1

AP-SWATH Workflow Schematic for Mapping the Composition and Dynamics of the GRB2 Interactome in Primary T Cells (Left) Overview of affinity purification (AP) of primary mouse T cells isolated from WT mice (GRB2^WT) and from knockin mice expressing endogenous GRB2 tagged with a One-STrEP-tag (OST) (GRB2^OST). T cells were isolated before or after stimulation for 0.5, 2, 5, and 10 min with anti-CD3 and anti-CD4 antibodies followed by affinity purification of GRB2 protein complexes using Strep-Tactin Sepharose beads. (Upper right) Overview of MS analysis. Affinity-purified samples were acquired in both DIA/SWATH and DDA/shotgun modes (for a detailed description, see Aebersold and Mann, 2016, Gillet et al., 2016). Data were used to build a high-confidence GRB2 interactor-specific assay library. Visualization of AP-SWATH data, quantification, and statistics analysis were performed in Skyline. (Lower right) Clustering of interaction dynamics and network-based analysis were performed using the Garuda platform, a community-driven software platform that supports reproducibility of computational analysis from complex high-dimensional data. See also Figures S1 and S2.

Figure 2

The GRB2 Interactome in Peripheral CD4⁺ T Cells The advanced workflow applied in this study enabled the identification of 53 high-confidence direct or indirect GRB2 interactors in resting and activated CD4⁺ T cells. Proteins were classified according to their function or protein family (see key). GRB2 interactors that have been previously identified in T cells from public databases are border painted in red. Components of modules are defined by the degree of interconnectivities of preys, subcellular localizations, and described functionalities. GAP, GTPase-activating protein; GEF, guanine-exchange factor. See also Figure S3 and Data S1.

Figure 3

Precise and Reproducible Quantification of Proteins Interacting with GRB2 in Peripheral CD4⁺ T Cells (A) The heatmap shows log₂ fold change for 53 high-confidence GRB2 interactors upon stimulation of peripheral CD4⁺ T cells. Recruited, stable, and dissociated proteins are in red, black, and green, respectively. Statistically significant changes with adjusted p values < 0.05, for at least one time point, are indicated by the asterisks at the right of the map. Statistics were calculated in Skyline using MSstats. (B) Dynamics of interactions were clustered using the Garuda software platform and normalized as the percentage of the maximal value for each kinetic. Representative clusters (i.e., cluster 1 and 7) are shown. (C) Distribution of percent CV. Median CV is indicated in parentheses for each stimulation time point. Four biological replicates for each stimulation time point were used to evaluate biological reproducibility at protein level. (D) Kinetics for 12 clustered and representative GRB2 interactors. ^∗Adjusted p < 0.05; ^∗∗adjusted p < 0.01; ^∗∗∗adjusted p < 0.001. Error bars represent SD. See also Figures S4 and S5 and Data S1.

Figure 4

Quantitative Comparison between the Dynamics of GRB2 Protein Complexes in Developing versus Mature Primary T Cells (A) The GRB2 protein-protein interaction network displays the association of time-dependent proteins following anti-CD3 and anti-CD4 stimulation in mature/peripheral CD4⁺ T cells (P) and developing/thymocytes (T). Nodes represent 15 proteins, including GRB2, that were consistently detected and reproducibly quantified in both peripheral CD4⁺ T cells and thymocytes. Rectangles inside the nodes show the normalized (Norm.) log₂ fold change for each time point as the percentage of the maximal (red) or minimal value (green). The significantly different GRB2 protein interactions between the two cell types, as determined by MSstats, are indicated by an asterisk (^∗adjusted p < 0.05). (B) Quantitative temporal profiling of three representative GRB2 interactors upon activation of peripheral T cells (blue) or thymic T cell (orange). The significantly different GRB2-PTPRA interaction, as determined by MSstats, between the two cell types at 0.5 min post-stimulation is indicated (^∗∗adjusted p < 0.01). Error bars represent SD. See also Figure S6 and Data S1.