Proteomimetic Strategy for the Modulation of Intrinsically Disordered Protein MYC

Abstract

The difficulty in developing specific ligands for protein receptors is directly correlated to the presence of unique binding sites on the protein surface. Conformationally dynamic proteins increase the level of difficulty in ligand design, and the challenge is further exacerbated for proteins that are intrinsically disordered. Intrinsically disordered proteins (or IDPs) do not adopt a fixed three-dimensional shape until they bind their target; an absence of organized binding sites underscores the difficulty in developing synthetic ligands for these proteins. We hypothesized that one avenue for the development of binders for a disordered region would be to trap one of its thermodynamically accessible conformations in a receptor. Here, we show the application of this approach to MYC, which represents a critical therapeutic target but has not yielded small-molecule inhibitors due to its conformationally dynamic nature. MYC adopts a helical configuration when it binds to its cellular partner MAX. We rationally designed a proteomimetic scaffold to trap this conformation. We show that MYC can be directly engaged in both biochemical and cellular assays. Overall, this work demonstrates a general method to capture and trap intrinsically disordered proteins with a propensity to adopt α-helical conformations.

Figure 1

(A) Proposed strategy to trap disordered MYC^406–436 sequence with a rationally designed cross-linked helix dimer receptor (CHD). (B) (i) GCN4 leucine zipper can populate either a dimeric or trimeric fold—with the dimer serving as a host for the third strand (PDB codes: 4DMD and 4DME). (ii) A 5-helix bundle can accommodate C-peptide from HIV gp41 to form a 6-helix bundle.

Figure 2

Rational design, biophysical characterization, and stability of MAX-derived cross-linked coiled coils. (A) Helical wheel diagrams depicting the native MYC/MAX dimerization interface (left), a proposed trimer where dimeric MAX may host a MYC sequence (middle), and designed trimeric interaction of MYC with MAX-derived constrained helix dimer (right). (B) Peptide sequences of designed compounds. We utilized CCBuilder, a computational tool that predicts coiled coil stabilities using the BUDE (Bristol University Docking Engine) force field. The BUDE energies of CHDs in complex with MYC^406–421 are listed. (C) Binding affinities of fluorescein-labeled CHDs for His₆-MYC^353–437 were analyzed in fluorescence polarization (FP) assays. *indicates maximum concentration of His₆-MYC^353–437 possible in the assay due to the observed aggregation at higher values. (D) FP assay to assess binding between the fluorescein-labeled CHD^Max-1 and His₆-MAX, or the MYC/MAX complex. (E) Schematic representation of the proposed model for the interactions of CHD^Max-1 with MYC alone and the MYC/MAX complex.

Figure 3

(A) The conformation of the CHDs was evaluated with CD in 0.1× PBS buffer at 20 μM peptide concentration. (B) Titration of 1:1 MYC^406–421 and CHD^Max-1 indicates complex formation (100 μM peptide concentration). (C) The CHD^Max-1 binding site on MYC was analyzed with a photo-cross-linking reaction.

Figure 4

(A) Proteolytic stability of CHD^Max-1 in 25% FBS. Error bars are mean ± SD of biological replicates. (B) Flow cytometry quantification of fluorescently labeled CHD^Max-1 or DMSO in T24 cells. (C) Live-cell fluorescence imaging of Hoechst-stained T24 cells incubated with fluorescein-labeled CHD^Max-1 or DMSO for 4 h. Scale bar represents 10 μm. (D) Western blot analysis on endogenous MYC protein after biotin-CHD^Max-1 pull-down in T24 lysate. (E) Treatment of T24 cells with increasing concentrations of CHD^Max-1 for 24 h resulted in increasing protein levels of MYC and MAX but not MAD1. (F) Relative protein levels of MYC/MAX/MAD1 normalized to β-actin from Western blot in D.