A new soaking procedure for X-ray crystallographic structural determination of protein–peptide complexes

Abstract

Interactions between a protein and a peptide motif of its protein partner are prevalent in nature. Often, a protein also has multiple interaction partners. X-ray protein crystallography is commonly used to examine these interactions in terms of bond distances and angles as well as to describe hotspots within protein complexes. However, the crystallization process presents a significant bottleneck in structure determination since it often requires notably time-consuming screening procedures, which involve testing a broad range of crystallization conditions via a trial-and-error approach. This difficulty is also increased as each protein–peptide complex does not necessarily crystallize under the same conditions. Here, a new co-crystallization/peptide-soaking method is presented which circumvents the need to return to the initial lengthy crystal screening and optimization processes for each consequent new complex. The 14-3-3σ protein, which has multiple interacting partners with specific peptidic motifs, was used as a case study. It was found that co-crystals of 14-3-3σ and a low-affinity peptide from one of its partners, c-Jun, could easily be soaked with another interacting peptide to quickly and easily generate new structures at high resolution. Not only does this significantly reduce the production time, but new 14-3-3–peptide structures that were previously not accessible with the 14-3-3σ isoform, despite screening hundreds of other different conditions, were now also able to be resolved. The findings achieved in this study may be considered as a supporting and practical guide to potentially enable the acceleration of the crystallization process of any protein–peptide system.

A Systematic Approach to the Discovery of Protein-Protein Interaction Stabilizers

“…The screening was performed on five different peptide targets displaying three conceptually distinct 14-3-3 interaction motifs (Figure 1C): truncated (ERα), 14,40 turned (FOXO1), 37 and linear (C-RAF, USP8, SOS1). 32,35,38,41 14-3-3σ (100 nM) was screened in complex with the 5 client phospho-peptides at a concentration twice their respective KD values (Figure S2). This condition provided a consistent presence of the 14-3-3σ/phospho-peptide composite interface that the fragments would engage.…”

“…The screening was performed on five different peptide targets displaying three conceptually distinct 14-3-3 interaction motifs (Figure 1C): truncated (ERα), 14,40 turned (FOXO1), 37 and linear (C-RAF, USP8, SOS1). 32,35,38,41…”

“…30 The Protein Data Bank contains dozens of crystallographic structures of 14-3-3 with bound phospho-peptides derived from many of its binding partners, as well as a few examples of CryoEM structures of full length proteins. 13,33–35 This wealth of structural information allows for direct visualization of the various 14-3-3/client binding interfaces which could be capitalized on for the discovery of selective fragment stabilizers and the development of potent lead compounds through structure-guided chemical optimization. For our screens, we utilized the phospho-peptide mimetics of 14-3-3 PPI partners which bind 14-3-3 in a similar fashion to the unstructured regions of the full-length proteins, but offer greater synthetic flexibility and simplified crystallography.…”

A Systematic Approach to the Discovery of Protein–Protein Interaction Stabilizers

“…The cysteine forms a reversible covalent bond with the fragment thiol through disulfide exchange; the amount of bound fragment is measured by MS. A fragment stabilizer is expected to show a higher “% tethering” in the presence of the 14-3-3σ/client phosphopeptide complex than 14-3-3σ alone due to cooperativity between the fragment and the peptide (Figure B). The screening was performed on five different peptide targets displaying three conceptually distinct 14-3-3 interaction motifs (Figure C): truncated (ERα), , turned (FOXO1), and linear (C-RAF, USP8, SOS1). ,,, …”

“…Of the 7 isoforms found in mammalian cells, 14-3-3σ is the only one that harbors a cysteine residue proximal to the client binding groove, providing an additional degree of isoform specificity (Figure A) . The Protein Data Bank contains dozens of crystallographic structures of 14-3-3 with bound phosphopeptides derived from many of its binding partners, as well as a few examples of CryoEM structures of full length proteins. ,− This wealth of structural information allows for direct visualization of the various 14-3-3/client binding interfaces which could be capitalized on for the discovery of selective fragment stabilizers and the development of potent lead compounds through structure-guided chemical optimization. For our screens, we utilized the phosphopeptide mimetics of 14-3-3 PPI partners which bind 14-3-3 in a similar fashion to the unstructured regions of the full-length proteins but offer greater synthetic flexibility and simplified crystallography. , …”