| Random pool|| MD|| SAXS|| NMR|| Validated|
Pau Bernado; Alfredo de Biasio; Alain Ibez de Opakua; Tiago N. Cordeiro; Maider Villate; Nekane Merino; Nathalie Sibille; Moreno Lelli; Tammo Dierks; Francisco J. BlancoPublication
: p15(PAF) Is an Intrinsically Disordered Protein with Nonrandom Structural Preferences at Sites of Interaction with Other Proteins.Release date:
The PCNA-associated factor p15PAF is a 111 residue protein that acts as a regulator of DNA repair during DNA replication. The p15PAF gene is overexpressed in several types of human cancer. Here, we present an ensemble description of p15PAF built using Residual Dipolar Couplings (RDCs) measured by NMR, and SAXS data. Our RDC analysis reveals the presence of residual secondary structural elements in the PCNA-binding motif (PIP-box), the KEN-box, and two regions at the N- and C-terminal parts of the protein. The ensemble refined with RDC data is in excellent agreement with the SAXS curve of p15PAF.Calculation procedure:
Ensembles of explicit models were generated for p15PAF using Flexible-Meccano by modifying the conformational preferences at residue level. This process was guided by the 86 1H-15N RDCs measured in stretched polyacrylamide gels. To account for deviations from the random coil description of p15PAF conformational sampling, several ensembles were computed including defined local conformational propensities in continuous regions of the protein with a defined population. The standard conformational sampling coded in Flexible-Mecano was used in the non-structurally biased regions. Canonical secondary structure backbone dihedral angles were used in the partially structured regions with the exception of the KEN box (76-80), where the φ,ψ angles were taken from the corresponding region in the crystal structure of the protein Mad3 in the mitotic checkpoint complex (PDB entry 4AEZ). All the ensembles tested contained 100,000 conformers. For each conformer of each ensemble, the 1H-15N RDCs were predicted using PALES, and the per-residue averages over all conformers were computed and defined as the simulated RDC values. The quality of the ensembles built was assessed by comparing the simulated and experimental values through a figure of merit.
The ensemble with a better description to the experimental RDCs was found imposing the following structural restrictions:
- 15% of α-helix in residues 54-60.
- 15% of 310 helix in residues 67-69.
- 8% of β-strand in residues 15-24.
- 8% of β-strand in residues 94-104.
- 15% of the canonical inverse γ turn structure in W61.
- 5% of KEN-box torsion angles in residues 76-80.
The introduction of these secondary structure preferences in the indicated regions not only results in a notable agreement with the measured RDCs in those regions but also improves the correlation along the entire chain.
The quality of the final RDC-refined ensemble was validated using SAXS. For the simulation of the SAXS curve of the RDC-derived ensemble, side chains were added using SCCOMP to each of the conformers and the scattering intensity of each of them was calculated using CRYSOL 2.6. Subsequently, the individual SAXS curves were averaged to yield the ensemble scattering curve. Both curves were directly compared by scaling the theoretical curve using a momentum transfer range of 0.015 < s < 0.50 Å-1. Constant subtraction was used. A total of 4,000 structures were used to calculate the average curve, which was sufficient to attain convergence.
The experimental SAXS data were accurately reproduced by those calculated for the RDC-refined ensemble of p15PAF indicating that this ensemble is a good representation of the conformational distribution of p15PAF in solution.