Analyzing Context-Specific Protein Complexes

Identifying Protein-Protein Interactions with Split-BioID

  • Fig. 1: The AP-MS approach. The protein of interest (POI) can be part of complex C. Beads coupled to a specific antibody or to a suitable chemical are respectively used to pull down the endogenous or tagged POI from cell lysates. Interacting proteins (belonging to complex 1 or 2) are co-purified and can be identified by MS analysis.Fig. 1: The AP-MS approach. The protein of interest (POI) can be part of complex C. Beads coupled to a specific antibody or to a suitable chemical are respectively used to pull down the endogenous or tagged POI from cell lysates. Interacting proteins (belonging to complex 1 or 2) are co-purified and can be identified by MS analysis.
  • Fig. 1: The AP-MS approach. The protein of interest (POI) can be part of complex C. Beads coupled to a specific antibody or to a suitable chemical are respectively used to pull down the endogenous or tagged POI from cell lysates. Interacting proteins (belonging to complex 1 or 2) are co-purified and can be identified by MS analysis.
  • Fig. 2: The BioID-MS approach. The protein of interest is fused to a promiscuous biotin ligase (BioID). In living cells, proximal proteins to the fusion proteins are biotinylated. After cell lysis, biotinylated proteins (belonging to complex 1 or 2) are recovered on streptavidin (S)-coupled beads under denaturing conditions, and can later be identified by MS analysis.
  • Fig. 3: Split-BioID, a conditional proteomics approach. The protein of interest and another protein belonging to complex 1 (protein B) are each fused with complementary fragments of the biotin ligase. Upon interaction of the POI and protein B, the two fragments reassemble an active biotin ligase that selectively labels additional proteins belonging to complex 1. Biotin ligase activity is not reactivated in the context of complex 2. MS analysis is then used to specifically probe the composition of complex 1.

Most cellular functions are performed by proteins. Proteins seldom act alone but are typically part of large macromolecular assemblies that can remodel according to the function to be exerted. The protein-protein interactions (PPI) involved in such complexes are often deregulated in disease and represent potential target for therapeutics [1]. Hence, defining and understanding protein-protein interaction networks have been a subject of huge interest in biomedical research. However, the dynamic nature of PPI has often hampered progress in this field. We present a novel conditional proteomics method that allows probing the composition of spatiotemporally defined protein complexes [2].

Classical Methods for PPI Analysis

The gold standard method for identifying PPI has been the affinity purification coupled to mass spectrometry (AP-MS) method (fig. 1). In this approach, the protein of interest is pulled-down from cell lysates using a specific antibody coupled to beads. If no suitable antibody is available, a “tagged” fusion protein can be expressed. The tag is a usually short and easy to purify protein sequence that is appended to the protein of interest and thus provides a handle for subsequent pull-down experiments. Once the native or tagged protein has been immobilized, beads are extensively washed to remove contaminants. Interacting partners will remain bound to the beads by virtue of their PPI with the protein of interest and can be comprehensively identified by mass spectrometry analysis. While AP-MS has been widely used for small to large-scale identification of protein networks, it suffers from several limitations. First, cells must be lysed before the pull-down step. Hence any PPI that depends on the exquisite subcellular organization of the cell may be lost. Second, due to the time and multiple washing steps of the procedure, AP-MS is not performing well at identifying transient or weak interactions.

Proximity-Dependent Proteomics

Recently, proximity-dependent labeling approaches were developed to address limitations of AP-MS. Here the protein of interest is fused to an enzyme that, when expressed in cells, can modify proximal proteins hence labeling them for further analysis.

Two such enzymes APEX2 [3] and BioID [4] are used to biotinylate proteins in a proximity-dependent manner. Both enzymes catalyze the formation of a short-lived reactive intermediate (a biotin-phenoxyl radical for APEX2 and biotinyl-AMP for BioID) that is released in the cytosol where it can react with neighboring proteins. The labeling range of BioID has been estimated to 10 nm [5]. Marked with biotin, proximal proteins can then be easily isolated on streptavidin-coupled beads for subsequent identification by mass spectrometry (fig. 2). Compared to AP-MS, BioID-MS offers several key advantages. First, proximity-dependent labeling occurs in intact living cells. Hence interactions that happen on a specific cellular feature, such as endomembranes, or on a highly insoluble structures can be analyzed. Second, as opposed to AP that aims at purifying intact protein complexes, the pull down step of BioID harvest biotinylated proteins. Whether these are still interacting or not with the protein of interest is irrelevant. This unique feature of BioID combined with the exceptional affinity of streptavidin for biotin allows performing very harsh washing steps without loosing weak interacting proteins. Side-side comparison of AP- and BioID-MS confirmed that BioID performs better for identifying low expressed and transient interacting proteins [6]. A disadvantage of BioID over AP-MS is the need for a fusion protein. In those cases when a protein of interest looses its physiological activity when appended to a fusion partner, AP-MS performed with a specific antibody is still the method of choice.

Split-BioID a Conditional Proteomics Approach

Proteins are often part of several distinct complexes with small to large overlapping compositions corresponding to different maturation steps, subcellular localizations, or functions to be exerted. AP- and BioID-MS are powerful complementary methods for the study of PPI. However, while both allow a general sampling of all potential partners a given protein may have, they do not give any information on the context of these interactions. Assigning PPI to specific functional units from proteomics data generated by AP- or BioID-MS can thus be challenging. To address this long-standing hurdle, we developed a conditional proteomics approach that allows focusing the analysis on context-specific interactions [2].  The underlying concept of our assay, termed split-BioID, was to engineer a protein-fragment complementation assay (PCA) based on BioID. In PCAs, an enzyme with a measureable activity is split into two inactive fragments that can re-assemble into an active enzyme when brought in close proximity. We identified two PCA-suitable fragments of the BioID enzyme BirA* that did not show appreciable activity by themselves but re-assembled into an active biotin ligase when appended to two interacting proteins. Applying the assay on various examples, we showed that split-BioID has a considerably higher resolution than BioID. Indeed, as split-BioID is selectively activated upon the interaction of two proteins of interest, only proteins that assemble around this pair of interacting proteins are labeled (fig. 3). At the same time split-BioID conserved the unique features of BioID: interactions are captured in their native cellular environment independent of their strength.

Conclusions

Altogether we have developed a unique method in the toolbox for PPI analysis. Split-BioID is only activated when and where two given proteins interact. It thus specifically analyzes spatiotemporally defined complexes. Split-BioID is a simple and readily available assay that can be applied by scientists from many different fields of life science. With its ability to analyze context-specific protein assemblies, we expect that it will help understanding how cells react to their environment.

References
[1] Scott, D. E., Bayly, A. R., Abell, C. & Skidmore, J.: Small molecules, big targets: drug discovery faces the protein-protein interaction challenge. Nat Rev Drug Discov 15, 533-550, doi:10.1038/nrd.2016.29 (2016).
[2] Schopp, I. M. et al.: Split-BioID a conditional proteomics approach to monitor the composition of spatiotemporally defined protein complexes. Nat Commun 8, 15690, doi:10.1038/ncomms15690 (2017).
[3] Lam, S. S. et al.: Directed evolution of APEX2 for electron microscopy and proximity labeling. Nat Methods 12, 51-54, doi:10.1038/nmeth.3179 (2015).
[4] Roux, K. J., Kim, D. I., Raida, M. & Burke, B.: A promiscuous biotin ligase fusion protein identifies proximal and interacting proteins in mammalian cells. J Cell Biol 196, 801-810 (2012) doi: 10.1083/jcb.201112098
[5] Kim, D. I. et al.: Probing nuclear pore complex architecture with proximity-dependent biotinylation. Proc Natl Acad Sci U S A 111, E2453-2461, doi:10.1073/pnas.1406459111 (2014).
[6] Lambert, J. P., Tucholska, M., Go, C., Knight, J. D. & Gingras, A. C.: Proximity biotinylation and affinity purification are complementary approaches for the interactome mapping of chromatin-associated protein complexes. J Proteomics 118, 81-94, doi:10.1016/j.jprot.2014.09.011 (2015).

Author
J. Béthune

Contact
Dr. Julien Béthune

Biochemistry Center
Heidelberg University
Heidelberg, Germany
julien.bethune@bzh.uni-heidelberg.de

More articles

Overview of protein-protein interaction analysis

Contact

Universität Heidelberg
Im Neuenheimer Feld 328
69120 Heidelberg
Deutschland
Phone: 06221 54 4749

Register now!

The latest information directly via newsletter.

To prevent automated spam submissions leave this field empty.