Protein Complexes / Protein-Protein Interactions
In a living cell or a living environment proteins do usually not function in isolation, but they heavily interact with all biomolecule in their surroundings. Depending on these interactions they can alter their activity and carry out different biological functions. Protein complexes and protein interaction networks are mediators of essentially all biological functions, such as signal transduction processes, cellular transport, mechanical stability or active cellular movement.
Currently, the state-of-the-art proteomic workflow to study the composition of protein complexes is affinity purification followed by mass spectrometry (AP-MS). Affinity purification refers to the capture of biological material via specific enrichment with a ligand coupled to a solid support. Many types of ligands can be used, including DNA and RNA molecules, chemicals, lipids, peptides, or proteins. However, in the context of protein-protein interactions the most widely used ligands for affinity purification are protein antibodies targeting either endogenous or epitope-tagged proteins, for example through fusions to GFP or with short peptides (FLAG, MYC or haemagglutinin). Conducting affinity purification using antibodies is often referred to as immunopurification or immunoprecipitation (IP). The general idea is to immobilize a protein antibody on agarose or magnetic beads and to use this coupled ligand to capture target protein(s) from a soluble phase in which the native protein complexes are solubilized and stabilized. Once purified, proteins can be processed and directly analyzed by MS. Depending of the type of ligand, support matrix, and the stringency of the solid support washes (post purification), a certain amount of contaminating proteins “false positives” will be recovered and detected. Hence identification of such “false positive” interaction partners through thorough control and background samples is of immense importance in AP-MS studies. BayBioMS supports and scientifically consults customers in planning and execution of different kinds of IP workflows.
One alternative approach to AP-MS is “in-vivo proximity labelling”, for example through expression of a protein of interest fused to either a promiscuous biotin-ligase derived from bacteria (the BioID method) or to a peroxidase enzyme capable of activating biotin–phenol (the APEX method). Once activated, the biotin is rapidly and covalently conjugated to nearby Lys (in the case of BioID) or to Tyr (in the case of APEX) residues. This facilitates the subsequent enrichment of potential interacting proteins using a streptavidin pull-down. A second AP-MS alternative is based on variations of the “protein correlation profiling” (PCP) approach. These techniques use either chromatography or density gradient centrifugation to separate native protein complexes according to size, density, shape, charge and/or hydrophobicity. Protein elution or protein gradient profiles can then be generated for each protein individually and putative interacting proteins are finally identified by computational clustering on the basis of similarities in their elution profiles.
The typical outcome of an AP-MS analysis is a list of confidently identified complex-associated proteins. For each protein in this list a protein ratio to a specified control condition gets reported as well as a p-value that estimates the statistical significance of the detected protein ratio (only possible if replicate measurements have been carried out). A particular challenge in AP-MS data sets is the handling of missing values (i.e. a protein interaction partner is completely missing in the un-tagged control sample, but present in the sample with the tagged bait protein). This issue can be addressed using various strategies, for example by background intensity computation or by the re-filling of the missing values with some defined minimal intensity value, however, this always bears the risk of identification of false-positive interaction partners.