Engineering the hCRBPII Domain-Swapped Dimer into a New Class of Protein Switches

Abstract

Protein conformational switches or allosteric proteins play a key role in the regulation of many essential biological pathways. Nonetheless, the implementation of protein conformational switches in protein design applications has proven challenging, with only a few known examples that are not derivatives of naturally occurring allosteric systems. We have discovered that the domainswapped (DS) dimer of hCRBPII undergoes a large and robust conformational change upon retinal binding, making it a potentially powerful template for the design of protein conformational switches. Atomic resolution structures of the apo-and holo-forms illuminate a simple, mechanical movement involving sterically driven torsion angle flipping of two residues that drive the motion. We further demonstrate that the conformational "readout" can be altered by addition of crossdomain disulfide bonds, also visualized at atomic resolution. Finally, as a proof of principle, we have created an allosteric metal binding site in the DS dimer, where ligand binding results in a reversible 5-fold loss of metal binding affinity. The high resolution structure of the metal-bound variant illustrates a well-formed metal binding site at the interface of the two domains of the DS dimer and confirms the design strategy for allosteric regulation.

Human Cellular Retinol Binding Protein II Forms a Domain‐Swapped Trimer Representing a Novel Fold and a New Template for Protein Engineering

“…Providing larger binding surfaces, enhancing the number of binding sites for specific ligands, generating a higher local concentration of active sites, and creating new opportunities for allosteric regulation are some examples of functional diversification through domain‐swapping [6,11–16] . Domain swapping also is a useful protein engineering tool for creating new protein conformational switches, optogenetic tools, artificial enzymes, biosensors and so on as shown by many research groups including us [17–31] . Now our work in engineering the DS trimer as a novel protein engineering template can also be added to this literature.…”

“…Therefore, the mutations that provide a higher concentration of open monomer are expected to lead to increased dimer or trimer formation. We used Q108K:T51D ( M3 ), which was previously reported [20] to yield a substantial amount of domain‐swapped dimers, to test this hypothesis. The protein's monomer yield was very low (Table S2) and stability measurements showed the monomeric form of this variant to be unstable (Table S3).…”

“…Though we have identified mutational “hot spots” that give DS trimers during expression, control over the folding pathway is lacking as invariably significant amounts of DS dimer are also formed. We therefore turned to a strategy previously exploited [20] in controlling both DS dimer formation and allosteric behavior, the appropriate introduction of disulfide crosslinks that will bias the folding pathway toward DS trimer formation. We previously demonstrated that introduction of a cross‐subunit disulfide linkage, by introduction of the A28C mutation, results in new allosteric behavior in the domain‐swapped dimer and in essentially exclusive formation of DS dimer [20] (Table S1).…”

“…[6,[11][12][13][14][15][16] Domain swapping also is a useful protein engineering tool for creating new protein conformational switches, optogenetic tools, artificial enzymes, biosensors and so on as shown by many research groups including us. [17][18][19][20][21][22][23][24][25][26][27][28][29][30][31] Now our work in engineering the DS trimer as a novel protein engineering template can also be added to this literature.…”

“…The engineering of a metal-binding site to control oligomerization and create new function has been previously investigated by several research groups including us. [20,36,[42][43][44] Herein, we wanted to show how the structural differences between monomer, dimer, and trimer can be used to create a metal-binding site by minimal changes into the protein sequence. A characteristic feature of the DS trimer is the pseudo C 3 symmetry axis running through a volume of buried surface unique to the trimer structure ( Figure S4a).…”