Exploring the complete mutational space of the LDL receptor LA5 domain using molecular dynamics: linking SNPs with disease phenotypes in familial hypercholesterolemia

Human Molecular Genetics, 2016, Vol. 25, No. 6 1233–1246 doi: 10.1093/hmg/ddw004 Advance Access Publication Date: 10 January 2016 Association Studies Article A S S O C I AT I O N S T U D I E S A R T I C L E Exploring the complete mutational space of the LDL Vladimir Espinosa Angarica1,2,†, Modesto Orozco3,4,5 and Javier Sancho1,2,6, * 1 Departamento de Bioquímica y Biología Molecular y Celular, Facultad de Ciencias, Universidad de Zaragoza, Pedro Cerbuna 12, 50009 Zaragoza, Spain, 2Biocomputation and Complex Systems Physics Institute (BIFI), Joint Unit BIFI-IQFR (CSIC), Universidad de Zaragoza, Mariano Esquillor, Edificio I + D, 50018 Zaragoza, Spain, 3Institut de Recerca Biomèdica (IRB Barcelona), Baldiri Reixac 10-12, 08028 Barcelona, Spain, 4Departament de Bioquímica i Biologia Molecular, Universitat de Barcelona, Diagonal 643, 08028 Barcelona, Spain, 5Joint BSC-CRG-IRB Program in Computational Biology, Baldiri Reixac 10-12, 08028 Barcelona, Spain, and 6Aragon Institute for Health Research (IIS Aragón), Universidad de Zaragoza, Pedro Cerbuna 12, 50009 Zaragoza, Spain *To whom correspondence should be addressed at: Biocomputation and Complex Systems Physics Institute (BIFI), Mariano Esquillor, Edificio I + D, 50018 Zaragoza, Spain. Tel: +34 976761286; Fax: +34 976762123; Email: Abstract Familial hypercholesterolemia (FH), a genetic disorder with a prevalence of 0.2%, represents a high-risk factor to develop cardiovascular and cerebrovascular diseases. The majority and most severe FH cases are associated to mutations in the receptor for low-density lipoproteins receptor (LDL-r), but the molecular basis explaining the connection between mutation and phenotype is often unknown, which hinders early diagnosis and treatment of the disease. We have used atomistic simulations to explore the complete SNP mutational space (227 mutants) of the LA5 repeat, the key domain for interacting with LDL that is coded in the exon concentrating the highest number of mutations. Four clusters of mutants of different stability have been identified. The majority of the 50 FH known mutations (33) appear distributed in the unstable clusters, i.e. loss of conformational stability explains two-third of FH phenotypes. However, one-third of FH phenotypes (17 mutations) do not destabilize the LR5 repeat. Combining our simulations with available structural data from different laboratories, we have defined a consensusbinding site for the interaction of the LA5 repeat with LDL-r partner proteins and have found that most (16) of the 17 stable FH mutations occur at binding site residues. Thus, LA5-associated FH arises from mutations that cause either the loss of stability or a decrease in domain’s-binding affinity. Based on this finding, we propose the likely phenotype of each possible SNP in the LA5 repeat and outline a procedure to make a full computational diagnosis for FH. † Present address: Luxembourg Centre for Systems Biomedicine (LCSB), University of Luxembourg. 6, Avenue du Swing, L-4366 Esch-sur-Alzette, Luxembourg. Received: September 29, 2015. Revised: December 19, 2015. Accepted: January 5, 2016 © The Author 2016. Published by Oxford University Press. This is an Open Access article distributed under the terms of the Creative Commons Attribution Non-Commercial License (http://creativecommons.org/ licenses/by-nc/4.0/), which permits non-commercial re-use, distribution, and reproduction in any medium, provided the original work is properly cited. For commercial re-use, please contact 1233 receptor LA5 domain using molecular dynamics: linking SNPs with disease phenotypes in familial hypercholesterolemia 1234 | Human Molecular Genetics, 2016, Vol. 25, No. 6 Introduction Results Global conformational instability of the LA5 domain mutant variants The 227 mutants resulting from SNPs in the LA5 domain (Supplementary Material, Fig. S1 and Table S2) were generated in silico using SCWRL (51), which was also used to find the best rotamer conformations for the side chains of the substituted residues. All mutants were minimized and equilibrated in explicit water and MD simulations were extended for a total aggregated time of 6 µs, including preparation and production trajectories (see the ‘Materials and Methods’ section). For each mutant 20 nslong MD simulations were used to explore its conformational evolution upon mutation. Based on previous work (49), this time span was considered enough to permit significant relaxation from the initial wild-type-like structure. A parameter commonly used to evaluate the structural similarity of two proteins, or of two protein conformations along an MD trajectory, is the root mean square deviation (RMSD) or average distance between the atoms of the superimposed structures. However, some The low-density lipoprotein (LDL) receptor (LDL-r) belongs and gives name to an ancient family of membrane receptors, including very-low-density lipoprotein receptor (VLDL-r), ApoER2, LRP1, LRP2 and LRP6 (1), that appeared early with the onset of the first metazoans and play important roles in multiple biological processes, through binding to a diverse set of partners (2,3). The receptors of the LDL-r family contain a common set of structural constituents that, from C-terminal to N-terminal, include (a) a cytoplasmic region that embodies NPxY and PPPSP motifs, (b) a single-transmembrane segment anchoring the cytoplasmic and extracellular sections to the cell membrane and (c) an extracellular region formed by an epidermal growth factor (EGF)-like domain composed by several EGF repeats and a β-propeller domain, followed by a ligand-binding region consisting of a variable number of small cysteine-rich domains (1,3), known as LDL-r type A domains (LA domains). The ligand-binding region of the human LDL-r (4) has been widely studied to uncover the mechanism of endocytic LDL internalization and release. These studies have provided a wealth of structural data corresponding to individual domains, domain pairs, the complete extracellular region (5–8) and even a low-resolution structure of the LDL–LDL-r complex (9). Domains LA1–7 (10) and, most importantly, domains LA4–5 (2,11), are key for binding of VLDL and LDL particles (12). LA domains are 40-residue, small autonomous folding units containing little secondary structure and lacking an extensive hydrophobic core, which are mainly stabilized by three disulfide bridges and a coordinated calcium ion (2,13–16). The seven LA domains are connected through small peptide linkers that provide great flexibility to the region (8). Familial hypercholesterolemia (FH) is a genetic disorder associated to abnormally high levels of LDLs in the blood, which constitute a significant-risk factor for cardiovascular and cerebrovascular diseases (17–20), and has a prevalence of 1:500 in heterozygosis in human populations (21,22). Although FH can be caused by defects in several proteins linked to cholesterol internalization and metabolism in cells—e.g. Apo (...truncated)