The Relationship Between the Three-Dimensional (3D) Structures of BF Molecules and MHC-Related Marek's Disease Resistance in Chickens.
Y-C Jin,P Wei,S Cui,L Huang
DOI: https://doi.org/10.1111/sji.12278
2015-01-01
Scandinavian Journal of Immunology
Abstract:To the Editor, The association between 3D structures of the BF molecules (with or without exon 7) and MHC-related MD resistance in chickens is not known. Because these associations have direct relevance to our research relating that the exon 7-deprived variant of the BF was detected in MD-susceptible haplotypes, but not in MD-resistant haplotypes, we set out to address this question. Structural analysis of an amino acid sequence of BF that is from the B21 haplotype and another amino acid sequence of BF that is from the B21 haplotype by only removing the amino acid sequence of exon 7 of BF showed the overall structure of the both is similar, but an in-depth analysis revealed that the structure of BF without exon 7 contains a distinctive narrow peptide-binding groove, which is significantly different from the structure of BF with exon 7 with a larger central cavity in groove. The larger central cavity in groove confers substantial conformational flexibility to the crucial residue Arg9, which allows remodelling of key peptide-binding sites and allows peptides with conspicuously different sequences to be bound. This might explain that the exon 7-deprived variant of the BF was detected in MD-susceptible haplotypes, but not in MD-resistant haplotypes 1-3. The amino acid sequence of BF2*2110 (GenBank accession number: NP_001026509) and BF2*2120 (the same sequence of BF2*2110 only removing the amino acid sequence of exon 7) was directly uploaded to the I-TASSER web to acquire BF2*2110 and BF2*2120 protein models (named model A1 and model B1, respectively) handled by energy minimization and FG-MD molecular dynamic simulation and then model A1 and model B1, respectively, to were uploaded SAVES for protein model evaluation. A psf/pdb file (named mymem1) of 50Å pre-equilibrium plasma membrane in X and Y directions, respectively, was generated to acquire and build bilayer of BF2*2110 and BF2*2120 proteins using membrane plug-in tools of VMD then the bilayer molecular was put into a pre-equilibration water box at specified location and size for solvate and then the mymem1 formation was balanced under the conditions of constant temperature (310 K) and pressure (the methods in accordance with Grubmüller and Seeliger). The post-equilibrium mymem1 and pdb file of model A1 or model B1 was merged and loaded into VMD to show the protein with new cartoon and the membrane with line and dragged on the protein that interacts with the membrane to ensure that the 280–302 amino acid residues of BF molecule are in transmembrane position and then the file was saved as model A1 withmem.pdb or model B1 withmem.pdb. The model A1withmem.pdb or modelB1withmem.pdb was directly uploaded to Dawning 5000A supercomputer of China to perform the molecular dynamic simulation by NAMD software and BF2*2110 and BF2*2120 protein models were ultimately acquired (named model A and model B, respectively). Five prediction models for ProSeq A and ProSeq B, respectively, arranged by their order in the score set were returned by the I-TASSER web. Only the highest scoring models A1 and B1 as initial structure were uploaded, respectively, to SAVES for protein model evaluation. As main feedback, Ramachandran plot of models A1 and B1 (Figs. 1 and 2) reflects that models A1 and B1 obtained a good evaluation result and were in the acceptable range: 83.6% and 86.8% of the amino acids fall within the most credible regions, 10.5% and 8.3% of the amino acids fall within the acceptable regions and 3.3% and 1.9% of the amino acids fall within the totally wrong regions. Film thickness ultimately stabilizes at around 4.8 nm, which means it has a low energy level and a steady state. After the post-equilibrium mymem1 and pdb file of model A1 or model B1 was merged, it was uploaded to the supercomputer for processing. In the course of energy minimization, the total energy of the membrane-protein system entered a rapid decline and eventually converged. In the subsequent molecular dynamics simulations, RMSD (root-mean-square deviation) of all atoms and main-chain Cα atoms gradually drops and tends to stabilization, which shows that whole membrane-protein system becomes gradually stable. Finally, we obtained model A and model B. Structural analysis showed the overall structure of both model A and model B is similar, but an in-depth analysis of the two structures revealed that model B (the structure of BF without exon 7) contains a distinctive narrow peptide-binding groove (1 310 Å3), which is significantly different from model A (the structure of BF with exon 7) with a large central cavity in groove (1 550 Å3). Model A and model B were enlarged to better see the maximum distance of peptide-binding groove of those in chimera (Fig 3), and the results showed that there are considerable differences between them: 14.35 Å and 12.40 Å, respectively. The larger central cavity in groove confers substantial conformational flexibility to the crucial residue Arg9, which allows remodelling of key peptide-binding sites and allows peptides with conspicuously different sequences to be bound. This might explain that the exon 7-deprived variant of the BF was detected in MD-susceptible haplotypes, but not in MD-resistant haplotypes. This research was supported by grants from the Nature Science Foundation of China (30460099, 39860055 and 30860206), Guangxi Provincial Programs for Science and Technology Development (10100005-10), Guangxi University Key Project (2005ZD01) and Nature Science Foundation of Hunan, P. R. China (13JJ6055). The authors have no financial conflict of interests.