Using machine learning to study protein–protein interactions: From the uromodulin polymer to egg zona pellucida filaments

Abstract

Neural network-based models for protein structure prediction have recently reached near-experimental accuracy and are fast becoming a powerful tool in the arsenal of biologists. As suggested by initial studies using RoseTTAFold or the ColabFold implementation of AlphaFold2, a particularly interesting future development will be the optimization of these computational methods to also routinely yield high-confidence predictions of protein-protein interactions. Here I use AlphaFold2 and ColabFold to investigate the activation and polymerization of uromodulin (UMOD)/Tamm-Horsfall protein, a zona pellucida (ZP) module-containing protein whose precursor and filamentous structures have been previously determined experimentally by X-ray crystallography and cryo-EM, respectively. Despite having no knowledge of the UMOD polymer structure (coordinates for which were neither used for model training nor as template), AlphaFold2/ColabFold are able to recapitulate a crucial conformational change underlying UMOD polymerization, as well as the general organization of protein subunits within the resulting filament. This surprising result is achieved by simply deleting from the input sequence a stretch of residues that correspond to a polymerization-inhibiting C-terminal propeptide. By mimicking in silico the activating effect of propeptide dissociation triggered by site-specific proteolysis of the protein precursor, this example has implications for the assembly of egg coat proteins and the many other molecules that also contain a ZP module. Most importantly, it shows the potential of exploiting machine learning not only to accurately predict the structures of individual proteins or complexes, but also to carry out computational experiments replicating specific molecular events.

Architecture of the vertebrate egg coat and structural basis of the ZP2 block to polyspermy

“…ZP module; (III-2) 1 copy of the hZP1 trefoil domain+ZP module, 1 copy of the hZP4 trefoil domain+ZP module; or (III-3) 1 copy of the hZP2 ZP module, 1 copy of the hZP4 trefoil domain+ZP module. Even when using 5 seeds per deep learning model, all predictions resulting from these combinations showed filamentous arrangements that were either circular (reminiscent of early prediction studies of UMOD 30 ) or straight (by disconnecting the ZP-N and ZP-C moieties of one subunit) (Figure S7A, step III).…”

“…Using this material, we crystallized a ~74 kDa native heteromeric complex consisting of an intact ZP module from the fish homolog of ZP3 (fZP3) and single copies of the ZP-N and ZP-C domains of fish ZP1 (fZP1) -a trefoil domain-containing VE subunit whose ZP module closely resembles those of ZP1, ZP2 and ZP4 proteins from other vertebrates, and whose IDL is specifically cleaved during hatching. Although the ~4 Å resolution diffraction of the crystals hampered experimental phasing, we could solve the structure by MR using a partial AlphaFold model of the fZP3/fZP1 complex, generated using a procedure developed to investigate UMOD polymerization in silico 30 (crystal form I, Figure S6A, B and Table S3). The resulting refined model was then used for MR phasing of data collected from a different crystal form of the same material, which diffracted highly anisotropically beyond 2.7 Å resolution (crystal form II, Figure S6D-H and Table S3).…”

Recent Advances in Deep Learning for Protein-Protein Interaction Analysis: A Comprehensive Review

“…Notably, Zhu et al [ 147 ] proposed PACNN + RL, a hybrid deep learning and reinforcement learning method for this task. On the other hand, Jovine [ 121 ] used AlphaFold2 and ColabFold to investigate the activation and polymerization of uromodulin, thus showcasing the practical applicability of these methods in biomedicine.…”

Evolving Concepts in Uromodulin Biology, Physiology, and Its Role in Disease: a Tale of Two Forms

“…10 There have been significant advancements into the mechanisms of polymerization, the structure of these filaments and the role of the polymerized form in the urine. [11][12][13][14][15][16][17] However, recent discoveries underscore the presence of a nonpolymerized form of uromodulin that is also released in the urine and is likely the main form released in the renal interstitium and circulation. 18 The cellular paths governing the release of these 2 forms are distinct from each other and appear to be independently regulated.…”