Portal de Periódicos da CAPES

Molecular Dynamics Simulation of the Consequences of a PYCR1 Mutation (p.Ala189Val) in Patients with Complex Connective Tissue Disorder and Severe Intellectual Disability

2016; Elsevier BV; Volume: 137; Issue: 2 Linguagem: Inglês

10.1016/j.jid.2016.10.007

1523-1747

Hassan Vahidnezhad, Razieh Karamzadeh, Amir Hossein Saeidian, Leila Youssefian, Soheila Sotoudeh, Sirous Zeinali, Mohammad Vasei, Fatemeh Golnabi, Taghi Baghdadi, Jouni Uitto,

Skin and Cellular Biology Research

Heritable connective tissue disorders are a phenotypically highly heterogeneous group of conditions with a spectrum of manifestations affecting different organs and tissues. The heterogeneity is explained primarily by the repertoire of individual genes harboring mutations in these conditions (Uitto, 2014Uitto J. Heritable disorders of connective tissue: introduction to mini-review cluster.Matrix Biol. 2014; 33: 8-9Crossref Scopus (2) Google Scholar). As examples of the clinical heterogeneity with tissue specific connective tissue manifestations are as follows: (i) Cutis laxa (CL), characterized by redundant, loose, and sagging skin with loss of recoil. Extracutaneous manifestations include cardiac, arterial, pulmonary, and gastrointestinal abnormalities, and the clinical findings are due to mutations in the genes required for synthesis and assembly of the elastic fiber network (Uitto et al., 2013Uitto J. Li Q. Urban Z. The complexity of elastic fibre biogenesis in the skin—a perspective to the clinical heterogeneity of cutis laxa.Exp Dermatol. 2013; 22: 88-92Crossref PubMed Scopus (33) Google Scholar); (ii) Ehlers-Danlos syndrome (EDS), demonstrating hyperextensible skin with recoil and loose jointedness, sometimes associated with cardiovascular manifestations. EDS is classified into six subcategories, associated with distinct mutations primarily in the genes encoding various collagens or enzymes critical for collagen biosynthesis and assembly (Byers and Murray, 2014Byers P.H. Murray M.L. Ehlers-Danlos syndrome: a showcase of conditions that lead to understanding matrix biology.Matrix Biol. 2014; 33: 10-15Crossref PubMed Scopus (44) Google Scholar); (iii) Osteogenesis imperfecta (OI), manifesting with fragile bones with varying degree of tendency for fractures. OI is primarily caused by mutations in the genes, proα1(I) and proα2(I), encoding type I collagen, the major matrix component in cortical bone (Forlino and Marini, 2016Forlino A. Marini J.C. Osteogenesis imperfecta.Lancet. 2016; 387: 1657-1671Abstract Full Text Full Text PDF PubMed Scopus (517) Google Scholar). In this study, we have examined the molecular basis of connective tissue manifestations, including features of CL, EDS, and OI, associated with severe intellectual disability, in a consanguineous family from Iran with Persian ethnicity with five affected individuals (Figure 1a and Case Reports in Supplementary Materials online). The clinical findings included intellectual disability, triangular facies, and intrauterine growth retardation as well as wrinkled/inelastic skin and joint hypermobility, characteristic features of autosomal recessive CL due to proline synthesis defect (ARCL type 2B) (Dimopoulou et al., 2013Dimopoulou A. Fischer B. Gardeitchik T. Schroter P. Kayserili H. Schlack C. et al.Genotype-phenotype spectrum of PYCR1-related autosomal recessive cutis laxa.Mol Genet Metab. 2013; 110: 352-361Crossref PubMed Scopus (44) Google Scholar). Histopathology of skin biopsies revealed a marked reduction in elastic fibers and marked thinning and fragmentation of few visualized elastic fibers when examined by Verhoeff-von Gieson stain (Figure 1b). A mild inflammatory cell infiltration around the papillary vessels was noted. These clinical findings, with the pedigree analysis, and lack of elastin suggested the diagnosis of autosomal recessive CL (ARCL), possibly related to ALDH18A1 or PYCR1 mutations in patients with ARCL type 3A and 2B (OMIM# 219150 and 612940, respectively). Different types of CL can be caused by mutations in at least nine distinct genes (Urban and Davis, 2014Urban Z. Davis E.C. Cutis laxa: intersection of elastic fiber biogenesis, TGFbeta signaling, the secretory pathway and metabolism.Matrix Biol. 2014; 33: 16-22Crossref PubMed Scopus (36) Google Scholar). This study was approved by the Institutional Review Board of Pasteur Institute of Iran, and the patients, their parents or guardians gave their informed written consent to participate in this study and gave permission to use their photos/case history to be published. The initial approach to identify candidate genes in this family was to perform whole genome-wide single-nucleotide polymorphism-based homozygosity mapping using Illumina Infinum Exome-24 Bead Chip (240,000 single-nucleotide polymorphism markers) (Illumina, San Diego, CA), which revealed conserved homozygosity blocks along the entire autosome (Figure 1c). After elimination of potentially incidental runs of homozygosity (less than 2 Mb), the nine candidate genes related to ARCL were superimposed with homozygosity blocks that identified three potential candidate genes: GORAB, FBLN4, and PYCR1. GORAB is associated with gerodermia osteodysplastica characterized by wrinkly skin and osteoporosis, features shown by our patients (Yildirim et al., 2011Yildirim Y. Tolun A. Tuysuz B. The phenotype caused by PYCR1 mutations corresponds to geroderma osteodysplasticum rather than autosomal recessive cutis laxa type 2.Am J Med Genet A. 2011; 155A: 134-140Crossref PubMed Scopus (26) Google Scholar). However, sequencing of GORAB and FBLN4 did not reveal any pathogenic sequence variants. Mutation analysis of PYCR1 was performed by PCR amplification followed by bidirectional Sanger sequencing. Sequencing identified a homozygous c.566C>T, p.Ala189Val (accession no. NM_006907) mutation in PYCR1 segregating in the family with the affected individuals (Figure 1b). In 2009, two independent groups showed pathogenic mutations in PYCR1 in ARCL type 2B, including p.Ala189Val in a patient from Oman initially diagnosed as gerodermia osteodysplastica (Guernsey et al., 2009Guernsey D.L. Jiang H. Evans S.C. Ferguson M. Matsuoka M. Nightingale M. et al.Mutation in pyrroline-5-carboxylate reductase 1 gene in families with cutis laxa type 2.Am J Hum Genet. 2009; 85: 120-129Abstract Full Text Full Text PDF PubMed Scopus (68) Google Scholar, Reversade et al., 2009Reversade B. Escande-Beillard N. Dimopoulou A. Fischer B. Chng S.C. Li Y. et al.Mutations in PYCR1 cause cutis laxa with progeroid features.Nat Genet. 2009; 41: 1016-1021Crossref PubMed Scopus (170) Google Scholar). This finding highlights the clinical overlap between gerodermia osteodysplastica and ARCL type 2B (Yildirim et al., 2011Yildirim Y. Tolun A. Tuysuz B. The phenotype caused by PYCR1 mutations corresponds to geroderma osteodysplasticum rather than autosomal recessive cutis laxa type 2.Am J Med Genet A. 2011; 155A: 134-140Crossref PubMed Scopus (26) Google Scholar). Up to now, 34 pathogenic mutations have been reported in this gene, about half of them being missense. The PYCR1 gene encodes pyrroline-5-carboxylate reductase, a mitochondrial enzyme that catalyzes the final step of proline biosynthesis and reduces pyrroline-5-carboxylate to l-proline. Full-length PYCR1 polypeptide folds into two separate domains forming a monomeric complex structure (Figure 1d). The N-terminal domain (N) contains the catalytic site, whereas the C-terminal domain (C) is mainly involved in multimerization. The complex of PYCR1 is a pentamer, comprising five polypeptide subunits (A–E) (Figure 1e), two of which together form decameric "yo-yo-like" architecture. Each decameric complex is built from two identical interacting pentamers through their C domains (Supplementary Figure S1a online). To examine the consequences of the p.Ala189Val mutation at the molecular level, we utilized the time-course computational method, the molecular dynamics (MD) simulation, to investigate the structural impact of this mutation and its dynamic influence on multimeric stability and subunit motions of the PYCR1 complex. MD simulation provides a detailed description of atomic motions of protein as a function of time by solving Newton's equation of motions of classical physics (Karplus and McCammon, 1986Karplus M. McCammon J.A. The dynamics of proteins.Sci Am. 1986; 254: 42-51Crossref PubMed Scopus (80) Google Scholar). Crystal structures of PYCR1 complexes (Protein Data Bank entries: 2GRA and 2IZZ) were used as the initial conformations for all analyses (Meng et al., 2006Meng Z. Lou Z. Liu Z. Li M. Zhao X. Bartlam M. et al.Crystal structure of human pyrroline-5-carboxylate reductase.J Mol Biol. 2006; 359: 1364-1377Crossref PubMed Scopus (54) Google Scholar). After preparation of wild-type (Ala189) and mutant (Val189) structures, MD simulations were performed by the generalized born implicit solvent method of the NAMD 2.10 package (Phillips et al., 2005Phillips J.C. Braun R. Wang W. Gumbart J. Tajkhorshid E. Villa E. et al.Scalable molecular dynamics with NAMD.J Comput Chem. 2005; 26: 1781-1802Crossref PubMed Scopus (13144) Google Scholar) for the total production run of 20 ns, and the data were analyzed in R statistical language (for detailed information about the MD protocol, see Supplementary Materials). All computational analyses were also accomplished for one of the dimers in decameric conformation to confirm the structural impact of mutation in different multimeric states of PYCR1 (see Supplementary Materials for more information about related results). Initial structural analysis showed that p.Ala189Val mutation is located at the end of Helix-9 within the C-terminal domain of each PYCR1 subunit, which suggested the possible interference of multimeric assembly by this mutation. To verify this assumption, MD simulations in the wild-type and mutant structures in pentameric conformations were performed (thermodynamic parameters during the MD simulation were monitored and reported in Supplementary Table S1 online). During the monitoring of the 20-ns MD simulations, we observed that the B and E subunits were separated from each other in mutant conformation, which resulted in disassembly of the complex within less than 1 ns. In contrast, in this time period, the wild-type structure stayed assembled (Movie S1). For further analysis, some key structural parameters, including the average distance between the backbone atoms (root mean square deviations) of each subunit, pairwise-distance distribution of Ala189 (in wild-type) and Val189 (mutant), and the interaction energy between the adjacent subunits in wild-type and mutant PYCR1 complexes, were measured for all 4,000 structural snapshots (Figure 1g and h). Initial structural analysis including the root mean square deviation fluctuation between wild-type and mutant conformations during the MD simulation indicated that p.Ala189Val mutation affects the C-terminal rather than the N-terminal region of each monomer, thus confirming the effect of mutation on the dynamic behavior of C-terminus (Supplementary Figure S1c and d; see Supplementary Materials for full description). The pairwise-distance distribution of Ala189 and Val189 between the adjacent subunits of mutant and wild-type conformations indicated that the Val189 distance distribution between the B and E subunits was increased in comparison with the wild-type, which confirmed that the B and E subunits gradually drifted away (Figure 1i). Moreover, the interaction energy between the B and E subunits progressively diminished within the first 1,000 ps of simulation in the mutant pentameric structure, whereas it remained steady in the wild-type structure (Figure 1g and h). This was also detectable between the B and D subunits in which interaction energy between subunits decreased followed by the increasing distance of mutated residues along the trajectories in the pentameric form. To address the question of which molecular interactions are involved in disassembly of the B–E subunits, we measured the inter- and intrasubunit salt-bridge interactions and hydrogen bonds between the wild-type and mutant conformations of all snapshots. Figure 1f depicts the exclusive intersubunit salt-bridge interactions in all 2,000 sampled wild-type structures that were absent in the mutant form during the simulation. Supplementary Tables S2 and S3 online list the overall intersubunit salt-bridge and hydrogen bonds that are eliminated in p.Ala189Val mutant pentameric and decameric structures, respectively. Most of these interactions are located at the C domain responsible in the multimeric assembly of the PYCR1 complex. Moreover, this mutation affects all adjacent intersubunit interactions to weaken the multimeric assembly. In addition to the assembly of PYCR1 to pentameric and decameric complexes, it has been recently demonstrated that PYCR1, together with PYCR2, is a component of the ribonucleotide reductase small subunit B complex, and their interactions have been suggested to protect fibroblasts from oxidative stress (Kuo et al., 2016Kuo M.L. Lee M.B. Tang M. den Besten W. Hu S. Sweredoski M.J. et al.PYCR1 and PYCR2 interact and collaborate with RRM2B to protect cells from overt oxidative stress.Sci Rep. 2016; 6: 18846Crossref PubMed Scopus (50) Google Scholar). These mechanisms may be operative in connective tissues, and inactivation of this complex may explain the widespread deficiency manifesting as a complex disorder with features of CL, EDS, and OI, as encountered in our patients. In summary, in this study we report a missense mutation, p.Ala189Val, in the PYCR1 gene in a family with ARCL type 2B. In addition to CL, these patients have clinical findings consistent with EDS and OI, signifying the presence of a complex connective tissue disorder. In addition, the patients had severe intellectual disability that cosegregated with the connective tissue phenotype. A combination of computational methods demonstrated the impact of the missense mutation on structural stability, multimeric assembly, and dynamic behavior of the PYCR1 complex. We showed that Ala189 plays a pivotal role in maintaining the pentameric and decameric assembly of PYCR1, and that the p.Ala189Val mutation also leads to elimination of the critical inter- and intrasubunit interactions by affecting adjacent charged residues within the region in both pentameric and decameric conformations. These findings are predicted to result in loss of PYCR1 activity with severe phenotypic consequences. The authors state no conflicts of interest. Molecular dynamics simulation was performed using the Computing Cluster Facility of the University of Tehran, Iran. Carol Kelly assisted in manuscript preparation. eyJraWQiOiI4ZjUxYWNhY2IzYjhiNjNlNzFlYmIzYWFmYTU5NmZmYyIsImFsZyI6IlJTMjU2In0.eyJzdWIiOiI0YjAyNDhmMTBiMjdiYmM5YmJlNTdlZTEzYjg2ZGJmMCIsImtpZCI6IjhmNTFhY2FjYjNiOGI2M2U3MWViYjNhYWZhNTk2ZmZjIiwiZXhwIjoxNjc4NzQyNjgyfQ.DKvpmMdtdfLqW32BLYG8VGRoO0M6z0G657HMvj7f8UUtlDzFDhMkdRP3cFNKCaww9l8rO25q7MCtwzbMYduS47ImUgVhGkyUodH-Q-Blh_MHxrd2PdKnbr707rhk_Tzl0RS1V1nqpdTOmZJruKflhxFN2ufcEw64pKq12NRbtuIh1BwVUbg8S138_bZr81esbl3v0gxsGQKYIiTXq6779-uOzhDO1uH61cAzPq1Jbfc8u74zlLj63FoJyGiD_BnTxT0ItNOgnEwwn0CQqlr9MMoeqLl9OSXZzv3iT52Z13OCvRKLzNs3L4IFvoxzR3UVJonPLe52pMoT9iwMbvimdw Download .mp4 (11.65 MB) Help with .mp4 files Movie S1 Download .pdf (.57 MB) Help with pdf files Supplementary Data