In silico and in vitro elucidation of BH3 binding specificity toward Bcl-2

Abstract

Interactions between Bcl-2-like proteins and BH3 domains play a key role in the regulation of apoptosis. Despite the overall structural similarity of their interaction with helical BH3 domains, Bcl-2-like proteins exhibit an intricate spectrum of binding specificities whose underlying basis is not well understood. Here, we characterize these interactions using Rosetta FlexPepBind, a protocol for the prediction of peptide binding specificity that evaluates the binding potential of different peptides based on structural models of the corresponding peptide-receptor complexes. For two prominent players, Bcl-xL and Mcl-1, we obtain good agreement with a large set of experimental SPOT array measurements and recapitulate the binding specificity of peptides derived by yeast display in a previous study. We extend our approach to a third member of this family, Bcl-2: we test our blind prediction of the binding of 180 BIM-derived peptides with a corresponding experimental SPOT array. Both prediction and experiment reveal a Bcl-2 binding specificity pattern that resembles that of Bcl-xL. Finally, we extend this application to accurately predict the specificity pattern of additional human BH3-only derived peptides. This study characterizes the distinct patterns of binding specificity of BH3-only derived peptides for the Bcl-2 like proteins Bcl-xL, Mcl-1, and Bcl-2 and provides insight into the structural basis of determinants of specificity.

Fig. 1. BIM peptide in complex with Mcl1

The amphiphilic BIM BH3-only helical peptide (cartoon representation, rainbow coloring) binds Mcl1 (surface representation, gray) in a hydrophobic groove. 10 peptide interface positions that were mutated in the different datasets used in this study are indicated in stick representation, accompanied with their position nomenclature.

Fig. 2. Predicted binding values for BIM-derived peptides to Bcl-xL and Mcl-1 are in good agreement with experimental results from SPOT arrays

A.+B. Correlation of calculated score (X-axis) to the experimental peptide binding signal (Y-axis; −log(SPOT intensity) measured in BLU) for (A) Bcl-xL and (B) Mcl-1, for TRAIN1 (blue circles) and TRAIN2 probed with 1µM receptor (red crosses) (see Materials and Methods for description; the experimental data for TRAIN2 are also shown in Fig. 5A). The −7.5 −log (BLU) threshold (see Fig. S1), and the selected score thresholds are indicated with lines. C.+D. ROC curves for the discrimination of binders and non binders (threshold for binding defined as −log(BLU)=−7.5) for (C) Bcl-xL and (D) Mcl-1. Black circles indicate the selected thresholds. The high Area Under the Curve (AUC) values demonstrate good discrimination between BH3 Bim-derived peptides that bind / do not bind to different Bcl-2 binding proteins.

Fig. 3. Binding specificity and promiscuity at positions 3d and 4a

Binding score estimation for peptide binding to Bcl-xL (top) and Mcl-1 (bottom) recapitulates distinct specificity-determining residues identified using SPOT experiments . Bcl-xL binding peptides show a restricted range of amino acids at position 4a (blue diamonds), while many residues are tolerated at position 3d (red squares). Mcl-1 binding peptides in contrast show specific preferences for position 3d, but little discrimination at position 4a.

Fig. 4. Specificity prediction for yeast display-derived peptides

FlexPepBind is able to discriminate between Bcl-xL binders/non-binders (Bcl-xL specific binders all obtain good scores for binding to Bcl-xL, but worse scores for binding to Mcl-1; blue squares in upper left quadrant), as well as detect sequences that bind both proteins (peptides that bind to both proteins mostly obtain good scores for both; green triangles in lower left quadrant). However, it performs badly in detecting the Mcl-1 specific peptides, which all obtain unfavorable scores both for Bcl-xL and Mcl-1 binding (red circles in upper right quadrant–should be located in lower right quadrant).

Fig. 5. Computational and experimental characterization of Bcl-xL and Bcl-2 BIM-derived peptide-binding specificities

Sequences of 180 point mutants of the BIM BH3 peptide (TRAIN2) were evaluated for binding to Bcl-2 (left panels) and Bcl-xL (right panels). Experimental and predicted SPOT arrays are shown in the upper and lower panels, respectively. The predicted SPOT intensities are similar to the experimental SPOT arrays (r=0.72 for Bcl-xL and 0.67 for Bcl-2). The experimental specificity profiles of Bcl-xL and Bcl-2 are very similar (r=0.88; see Text for comparison). The array image in Fig. 5A was taken from .