Proximity Dependent Labeling Describes Role of Abelson Interactor 1 in Centrosome Biology

Abstract

Abelson interactor 1 (Abi-1) is a signaling adaptor protein dysregulated in cancer. Abi-1 is established to be involved in cytoskeletal reorganization regulated by Ras to Rac GTPase signal transduction, and our group recently showed that murine Abi-1 deficiency causes myeloproliferative neoplasm-like phenotype linked to activation of Src family kinases, STAT3, and NF-kappaB (Chorzalska et al, Blood, 2018). To resolve molecular mechanisms by which Abi-1 dysregulation destabilizes inflammatory homeostasis leading to cancer, we used proximity dependent biotin labeling followed by mass spectrometry (MS) using TurboID linked to Abi-1. TurboID is a biotin ligase from E. coli catalytically enhanced so that it rapidly converts biotin to active biotinoyl-5’-AMP which then binds to local proteins, enabling enrichment of biotinylated proteins from lysates by streptavidin and identification by MS (Brannon, Nat Biotech, 2018) (A). We used a retroviral transduction system to generate stable mouse embryonic fibroblast (MEF, NIH/3T3) cell lines expressing TurboID or TurboID-Linker-Abi-1 (B). We performed three experiments of labeling followed by MS using TurboID control, TurboID-Abi-1, and wildtype cell lines. Label-free quantification MS and data processing was performed using a Q Exactive Plus LC-MS/MS and Proteome Discoverer 2.3. Probable Abi-1 interacting proteins were determined by comparing assigned protein peak areas (PA) and peptide spectral matches (PSM) between streptavidin-enriched lysates from MEFs expressing TurboID-Abi-1 vs. TurboID, and TurboID-Abi-1 vs. wildtype (C). Valid statistical thresholds were determined by retrospective analysis of known interactor enrichment assisted by in-house software “Interactomatiker”. We identified 4,082 proteins, 2,997 in TurboID-Abi-1 lysates. Retrospective analysis of MS data determined a statistical threshold defining most probable Abi-1 interactors as: TurboID-Abi-1 vs. controls PA ratio ≥ 1.5, FDR ≤ 0.05, and PSM ratio ≥ 3.5, FDR ≤ 0.001, with a minimum average of 1 PSM in TurboID-Abi-1 MS injections. 74 proteins met this statistical threshold, 20 of which are known Abi-1 interactors (D). Pathway analyses of most probable interactors identified expected Abi-1 role in cytoskeletal reorganization (E), but also indicate a new role in centrosome biology by interaction with Nde1, Haus5, Haus7, Cep131, and Pcm1 (F). Additionally, Abi-1 deficient or over-expressing cells show increased or decreased proliferation by EdU incorporation assay (G). TurboID using Abi-1, assessed by expected enrichments, identified new Abi-1 roles which may provide mechanistic context to it's role in cancer. In addition to affecting centrosome biology, less stringent filtering strategy also indicates possible roles in TNFR signaling and protein trafficking. Using a discovery-based method to assess adaptor protein interactions, we identified signaling nodes likely affected in cancers associated with dysregulated Abi-1 activity.