Tumor-Suppressor p53TAD1-60 Forms a Fuzzy Complex with Metastasis-Associated S100A4: Structural Insights and Dynamics by an NMR/MD Approach

Abstract

Conformationally flexible protein complexes represent a major challenge for structural and dynamical studies. We present herein a method based on a hybrid NMR/MD approach to characterize the complex formed between the disordered p53TAD^1-60 and the metastasis-associated S100A4. Disorder-to-order transitions of both TAD1 and TAD2 subdomains upon interaction is detected. Still, p53TAD^1-60 remains highly flexible in the bound form, with residues L26, M40, and W53 being anchored to identical hydrophobic pockets of the S100A4 monomer chains. In the resulting "fuzzy" complex, the clamp-like binding of p53TAD^1-60 relies on specific hydrophobic anchors and on the existence of extended flexible segments. Our results demonstrate that structural and dynamical NMR parameters (cumulative Δδ, SSP, temperature coefficients, relaxation time, hetNOE) combined with MD simulations can be used to build a structural model even if, due to high flexibility, the classical solution structure calculation is not possible.

Keywords: NMR structures; S100A4; fuzzy complexes; molecular dynamics simulations; p53; protein folding.

Figure 1

Top: amino acid sequence of p53TAD^1–60, boxes represent the regions undergoing structural changes. A) ¹H,¹⁵N HSQC spectra at 313 K and 700 MHz of ¹⁵N‐labeled p53TAD^1–60 (black) and p53TAD^1–60 in complex with unlabeled, Ca²⁺‐loaded S100A4 (red, with assignment). B) Binding information: cumulative Δδ chemical shift changes of p53TAD^1–60 resonances upon S100A4 binding. Residues broadened below the detection limit are represented by an asterisk. C) Structural information: secondary structure propensities (SSP) for free (black) and complexed (red) p53TAD^1–60. D) Backbone dynamics results: reduced spectral density mapping J(ω_N) vs J(0) for free (black) and complexed p53TAD^1–60 (red).

Figure 2

A) Structure of the MD‐derived complexes. Cluster mid‐structures are shown for both Model A (dark gray) and Model B (light gray) accounting for 90 % of all snapshots of the last 300 ns of the simulations. For clarity, the S100A4 dimer is only shown in a single copy in green (monomer I) and cyan (monomer II). B) Calculated B‐factors from the MD models. The boxes indicate the position of the main helices. C) Examples of the H‐bonds formed during the MD simulations between the C‐terminal 95–101 segments of S100A4 (chains A (green) and B (cyan) are shown explicitly) and p53TAD^1–60. Interacting and anchoring residues of p53TAD^1–60 are also highlighted, D) Anchoring positions. The position of the three residues that undergo the largest chemical‐shift change on complexation: L26, M40 and W53 are labeled.

Figure 3

A) p53TAD structures (from Table 1, in color) in various complexes overlaid independently on the MD‐derived conformational ensembles of the present work (Model A in light gray, Model B in dark). B) Secondary structures of p53TAD in the above complexes. The length of the investigated fragment is shown in light gray, blue parts denote the helical regions. C) Overlaid structures of p53TAD^1–60 from the of the NMR/MD‐derived p53TAD^1–60–S100A4 complex (Model A in light gray, Model B in dark) and the crystallographically determined p53TAD^17–56 segment of the p53TAD^17–56–S100A4Δ8 complex in red, showing similar anchor points on the surface of the S100A4 dimer (shown in cyan/green in a single copy and based on the MD simulation).