
Available online at www.sciencedirect.com ScienceDirect Getting to know the neighborhood: using proximity- dependent biotinylation to characterize protein complexes and map organelles 1 1 2 Anne-Claude Gingras , Kento T Abe and Brian Raught The use of proximity-dependent biotinylation approaches Introduction combined with mass spectrometry (e.g. BioID and APEX) has Biological functions are generally performed by proteins revolutionized the study of protein–protein interactions and as components of complexes, organelles or other assem- organellar proteomics. These powerful techniques are based on blies. While the human genome project has provided a > the fusion of an enzyme (e.g. a biotin ligase or peroxidase) to a ‘parts list’ of the 20 000 polypeptides putatively ‘bait’ protein of interest, which is then expressed in a relevant encoded in the genome, how these proteins are organized biological setting. Additionof enzyme substrate enables covalent to perform the intricate functions required for life remains biotin labeling of proteins in the vicinity of the bait in vivo. These incompletely understood. To fill these gaps, a wide approaches thus allow for the capture and identification of variety of biochemical techniques have been developed ‘neighborhood’ proteins in the context of a living cell, and provide to characterize protein localization and identify interact- data that are complementary to more established techniques ing protein partners, including organelle purification and such as fractionation or affinity purification. As compared to immunoprecipitation coupled with mass spectrometry. standard affinity-based purification approaches, proximity- dependent biotinylation (PDB) can help to: first, identify Biochemical fractionation techniques are commonly interactions with and amongst membrane proteins, and other employed for the characterization of large protein com- polypeptide classes that are less amenable to study by standard plexes and defining organelle composition (Figure 1a). pulldown techniques; second, enrich for transient and/or low In organellar proteomics, fractionation approaches such affinity interactions that are not readily captured using affinity as density gradients are often coupled with mass spec- purification approaches; third, avoid post-lysis artefacts trometric characterization of the fractions (reviewed in associated with standard biochemical purification experiments Ref. [1]). The incorporation of quantitative proteomics and; fourth, provide deep insight into the organization of has enabled robust definition of the composition of membrane-less organelles and other subcellular structures that membrane-bound subcellular structures in diverse bio- cannot be easily isolated or purified. Given the increasing use of logical settings (e.g. [2,3]). Fractionation approaches these techniques to answer a variety of different types of can similarly be employed to purify biochemically biological questions, it is important to understand how best to stable protein complexes based on co-fractionation design PDB–MS experiments, what type of data they generate, using for example size or charge-based separation tech- and how to analyze and interpret the results. niques [4–6]. Both organellar fractionation and com- plexome profiling can also be coupled with perturba- Addresses tions such as a drug/hormone treatment or protein 1 Lunenfeld-Tanenbaum Research Institute, Sinai Health System, and knockdown/knockout [6,7]. Department of Molecular Genetics, University of Toronto, 600 University Ave, Rm 992, Toronto, ON, M5G1X5, Canada 2 Complementing these global profiling techniques, com- Princess Margaret Cancer Centre, University Health Network, and Department of Medical Biophysics, University of Toronto, 101 College mon approaches for the identification of protein–protein Street, 9-701A, Toronto, ON, M5G 1L7, Canada interactions typically rely on the in-frame fusion of an epitope tag that can be used for purification of a ‘bait’ Corresponding author: Gingras, Anne-Claude ([email protected]) polypeptide of interest, or bait-specific affinity reagents such as antibodies (Figure 1b). In these ‘affinity purifica- Current Opinion in Chemical Biology 2019, 48:44–54 tion coupled with mass spectrometry’ (AP–MS) techni- This review comes from a themed issue on Omics ques, endogenous proteins that interact with the bait can be captured, identified and quantified. AP–MS is widely Edited by Ileana M Cristea and Kathryn S Lilley used to identify interaction partners for a bait polypeptide For a complete overview see the Issue and the Editorial of interest, and the same approach conducted in an Available online 17th November 2018 iterative manner can be used to characterize protein https://doi.org/10.1016/j.cbpa.2018.10.017 complexes, as we discussed previously [8]. 1367-5931/ã 2018 Elsevier Ltd. All rights reserved. While these approaches are extremely powerful, they also share some important limitations. For example, the orga- nelles or complexes under study must first be efficiently extracted during cell or tissue lysis, and then maintain Current Opinion in Chemical Biology 2019, 48:44–54 www.sciencedirect.com Making sense of proximity-dependent biotinylation results Gingras, Abe and Raught 45 Figure 1 (a) Classical interaction and organellar proteomics (b) Affinity purification (AP) gentle cell lysate gentle cell lysis expressing bait cell lysis density organelle enrichment and fractionation gradient ultra- centrifugation various epitope protein multimers bait tag cell lysate anti-epitope beads 280 A [protein] liquid wash away chromatography protein complex purify interacting non-interacting fractionation proteins proteins (c) Proximity-dependent biotinylation (PDB) (d) Mass spectrometry analysis BioID K K K K K K K K trypsin K (protease) K K K peptide generation K biotin K bait K 3-24 hr K HPLC MS1 MS2 surface-exposed K biotin lysine residues ligase diminishing biotinylation strength intensity intensity intensity time m/z m/z Y Y APEX liquid chromatography - mass spectrometry Y Y (LC-MS) Y Y Y Y Y Y positive biotin- interactors Y bait Y phenol H2O2 Y 30 mins 1 min Y peroxidase surface-exposed tyrosine residues reconstruct elution score against controls profiles ANKRD28 PPP6R3 PPP6R2 PPP6R1 STK3 STK4 harsh HDAC6 lysis PPP2R1A NUDC HSPA4 streptavidin- ECH1 conjugated ACACB beads PCA plot dot plot in vivo protein purify biotinylation proximity-interactors data visualization Current Opinion in Chemical Biology Organellar and interaction proteomics approaches. (a) Schematic of classical organellar (top) and protein complex (bottom) fractionation approaches. A cell is first lysed, and the prepared lysate separated based on biophysical properties of its constituents. For comprehensive characterization, the entire elution profile is analyzed by quantitative MS (see d). (b) Standard affinity purification (AP). A bait is often fused to an epitope tag (here, the common 3x-FLAG tag is shown) and expressed in a relevant setting. After cell lysis under mild conditions, the epitope tag is captured on an affinity support, resulting in the purification of the bait and its stable interaction partners. (c) In proximity-dependent biotinylation www.sciencedirect.com Current Opinion in Chemical Biology 2019, 48:44–54 46 Omics their integrity throughout subsequent purification, frac- disrupt and efficiently solubilize membranes, organelles tionation and washing steps. This poses significant chal- and protein complexes. Biotin-tagged proteins are then lenges for the characterization of membrane-less orga- captured, most often using streptavidin linked to a solid- nelles or protein complexes localized to poorly soluble phase support, and identified by mass spectrometry subcellular compartments such as membranes, chroma- (Figure 1c). tin, nuclear lamina or the cytoskeleton. Many protein– protein interactions can be disrupted under the deter- To date, two main classes of enzymes have been gent/buffer conditions required to solubilize proteins employed for proximity-dependent biotinylation: biotin from these and other compartments. Similarly, maintain- ligases (BirA* [11 ], BioID2 [12], BASU [13], TurboID ing the integrity of membrane-less organelles can be [14 ], miniTurbo [14 ]) and peroxidases (APEX [15 ], difficult, and how accurately the resulting purified pro- APEX2 [16], HRP [17 ,18,19]). Biotin ligases activate ducts reflect the original structures in vivo can be difficult biotin to the reactive biotinoyl-AMP intermediate, but to assess. A second issue with standard biochemical have been modified from the wild type enzymes such that approaches is that strong interactors are much easier to they are unable to catalyze the direct transfer of activated capture using these methods than weaker or more tran- biotin to a specific amino acid sequence in a substrate sient binding partners. Weak protein–protein interactions protein. As such, these abortive enzymes simply release a play critically important roles in biology [9], but they are ‘cloud’ of activated biotin, which can react with epsilon most likely significantly underrepresented in current amine groups on lysine residues in nearby polypeptides protein interaction databases. A third limitation to these (Figure 1d). Peroxidases catalyze tyrosine attack by a approaches is post-lysis artefacts. Proteins not normally phenolic compound coupled to biotin (e.g. biotin-phe- localized at the same intracellular location can be mixed nol), when activated by peroxide. A major difference together upon lysis. Interacting partners
Details
-
File Typepdf
-
Upload Time-
-
Content LanguagesEnglish
-
Upload UserAnonymous/Not logged-in
-
File Pages11 Page
-
File Size-