Interactomics can be defined as the large-scale study of biomolecular relationships. It is often simply characterised as the concurrent investigation of many protein-protein interactions; one of life’s most important biomolecular relationships. Not all approaches for studying protein-protein interactions are equal, and this presentation will tease apart some of the benefits and drawbacks of these different approaches when assessed within the broader aims of interactomics.
Some approaches for studying protein-protein interactions are targeted towards specific interactions of interest. These approaches are capable of providing deep mechanistic insights. Our recent investigation of H2A.B, a histone variant which plays a role in oncogenesis in Hodgkin lymphoma, provides a case in point. After uncovering a series of novel post-translational modifications on the N-terminus of this protein, we employed biotinylated peptide pull-downs in conjunction with SILAC to identify protein-protein interactions mediated by these modifications. This uncovered a remarkable new function for H2A.B: the genomic targeting of the chromatin remodelling complex SWI/SNF is regulated by asymmetric dimethylation of H2A.BR8.
Targeted approaches for characterising protein-protein interactions do, however, feature limitations. It has been shown that when interactions uncovered in such studies are collated on an interactome scale, systemic biases towards interactions involving highly studied and/or highly expressed proteins become apparent [1]. Our findings involving H2A.B illustrate how such biases can emerge.
To avoid such biases, approaches for studying protein-protein interactions that are less targeted and more exploratory can be employed. These approaches include proteome scale affinity purification followed by mass spectrometry (AP-MS), yeast two-hybrid (Y2H) systems, co-fractionation mass spectrometry (CF-MS), and cross-linking mass spectrometry (XL-MS) of complex samples. The bird’s eye views of the interactome provided by these approaches do, however, come with a caveat: there is strikingly little overlap between the interactions identified across these different approaches [2,3].
To gain insight into this phenomenon, we assessed the congruence of protein-protein interactions identified from concurrent XL-MS and CF-MS analyses of Saccharomyces cerevisiae protein complexes. This helped test the assay complementarity hypothesis, which suggests that different interactome mapping techniques characterise different subsets of protein-protein interactions, thus explaining the small overlaps between interactome maps [2,4,5]. Our experiments reveal that protein-protein interactions identified from identical samples using XL-MS and CF-MS do indeed produce strikingly little overlap, supporting the idea of assay complementarity.
Together these examples speak to the relative merits of targeted versus exploratory studies of protein-protein interactions, and identify possible avenues towards fulfilling the promise of systems-level interactomics.