The intervening domain from MeCP2 enhances the DNA affinity of the methyl binding domain and provides an independent DNA interaction site

Abstract

Methyl-CpG binding protein 2 (MeCP2) preferentially interacts with methylated DNA and it is involved in epigenetic regulation and chromatin remodelling. Mutations in MeCP2 are linked to Rett syndrome, the leading cause of intellectual retardation in girls and causing mental, motor and growth impairment. Unstructured regions in MeCP2 provide the plasticity for establishing interactions with multiple binding partners. We present a biophysical characterization of the methyl binding domain (MBD) from MeCP2 reporting the contribution of flanking domains to its structural stability and dsDNA interaction. The flanking disordered intervening domain (ID) increased the structural stability of MBD, modified its dsDNA binding profile from an entropically-driven moderate-affinity binding to an overwhelmingly enthalpically-driven high-affinity binding. Additionally, ID provided an additional site for simultaneously and autonomously binding an independent dsDNA molecule, which is a key feature linked to the chromatin remodelling and looping activity of MeCP2, as well as its ability to interact with nucleosomes replacing histone H1. The dsDNA interaction is characterized by an unusually large heat capacity linked to a cluster of water molecules trapped within the binding interface. The dynamics of disordered regions together with extrinsic factors are key determinants of MeCP2 global structural properties and functional capabilities.

Figure 1. Structural features and thermal stability of the MeCP2 MBD.

(A,B) Solution structure of MBD (pdb 1qk9) and crystallographic structure of MBD bound to mCpG-dsDNA (pdb 3c2i) showing the tryptophan 104 within the folding core. (C) Fluorescence spectra recorded at different temperatures showing the large quantum yield of the single tryptophan at low temperature and the substantial dynamic quenching by water molecules upon unfolding. (D) Fluorescence thermal denaturations for MBD, NTD-MBD and NTD-MBD-ID in the absence of dsDNA and in the presence of unmethylated and mCpG-dsDNA. All unfolding traces could be fitted considering a two-state unfolding model (not shown for clarity purposes). Figures have been created with PyMol (http://www.pymol.org/).

Figure 2. MBD interaction with dsDNA.

(A,B) Calorimetric titrations of MBD interacting with dsDNA in Tris 50 mM, pH 7, 20 °C. The upper plots show the thermogram (thermal power as a function of time), whereas the lower plots show the binding isotherm (normalized heats as a function of the dsDNA/protein molar ratio). (C) Experiments at different temperatures provided an estimation of the binding heat capacity. (D) Experiments using buffers with different ionization enthalpy provided an estimation of the buffer-independent binding enthalpy and the net number of exchanged protons upon MBD-dsDNA complex formation.

Figure 3. NTD-MBD interaction with dsDNA.

(A,B) Calorimetric titrations of NTD-MBD interacting with dsDNA in Tris 50 mM, pH 7, 20 °C. The upper plots show the thermogram (thermal power as a function of time), whereas the lower plots show the binding isotherm (normalized heats as a function of the dsDNA/protein molar ratio). (C) Experiments at different temperatures provided an estimation of the binding heat capacity. (D) Experiments using buffers with different ionization enthalpy provided an estimation of the buffer-independent binding enthalpy and the net number of exchanged protons upon complex formation.

Figure 4. NTD-MBD-ID interaction with dsDNA.

(A,B) Calorimetric titrations of NTD-MBD-ID interacting with dsDNA in Tris 50 mM, pH 7, 20 °C. The upper plots show the thermogram (thermal power as a function of time), whereas the lower plots show the binding isotherm (normalized heats as a function of the dsDNA/protein molar ratio). (C) Experiments at different temperatures provided an estimation of the binding heat capacities. (D) Experiments using buffers with different ionization enthalpy provided an estimation of the buffer-independent binding enthalpy and the net number of exchanged protons upon complex formation.

Figure 5. Conformational and hydration contributions to the MBD-dsDNA interaction.

(A) Structural alignment of the structure of MBD in solution determined by NMR and the structure of MBD bound to mCpG-dsDNA determined by X-ray crystallography. (B) Network of hydrogen-bonding water molecules trapped in the MBD-dsDNA interaction interface. Water molecules at less than 4 Å simultaneously from both MBD and dsDNA are shown as spheres. Methyl-cytidines are shown in orange. (C) Buried hydrogen-bonding water molecules in the MBD-dsDNA complex. Water molecules with less than 2 Ų of SASA are shown as spheres (SASA values were calculated with Surface Racer70). Most of the water molecules shown are completely buried (SASA = 0 Ų). Methyl-cytidines are shown in orange.