Submitter Withdrawn 2014 International Biophysics Congress

Attempting homology modeling, energy optimization and structural validation of a difficult target SERAC1: The protein involved in an inborn error of metabolism, MEGDEL syndrome (#321)

Surajit Debnath 1 , Soma Addya 2
  1. Women‘s Polytechnic, Government of Tripura, Agartala, Tripura, India
  2. Department of Health Services, Government of West-Bengal, Pashchim Medinipur, West-Bengal, India

Membrane proteins are weakly resolved in X-ray/NMR with low crystallographic symmetry & several Ramachandran outliers 1 2.Serin Activesite Containing1 (SERAC1) is thought to be involved in intracellular cholesterol trafficking essential for mitochondrial function and is localized at endoplasmic reticulum (ER) & ER-mitochondria interface3.Its mutations are hypothesized to inhibit phosphatidylglycerol remodeling causing recessive metabolic disorder MEGDEL Syndrome (OMIM 614739).Catalytic mechanism of  transacylation-acylation and its role in MEGDEL Syndrome  is not fully understood.

 Crystal structure of SERAC1 is not resolved by X-ray/NMR. However it is a difficult target for usual homology modeling,(i)because it is a large enzyme (654 aa),(ii) it lacks suitable templates,& so initial stringent modeling algorithms failed to generate any perspective model & (iii)because it is a single pass membrane protein. 

 Remote Template based Multiple Model workflow using several algorithms (verified in rounds of CASP) successfully generated stable 3D model with adequate qualitative stringency for further study.

ClustalW identified 3 models with >75% coverage of the target NCBI Ref Seq. NP_116250.3 generated by 7 modeling programs.Cycles of steepest descent/conjugate gradient in various combinations (GROMOS96) on each model directed to parameters (Molprobity) that approach benchmark values of the remote templates used.We identified single model  with energetically stable ( -17906.334 KJ/mole) & geometrically acceptable constraints (highest Ramachandran favored residues 91.56% & lowest outliers of 15 or 2.29%).Surface electrostatic potential (AMBER 99) was in the range +/-78.209 -kT/e.Final model has lowest Cβ deviations >0.25Å (1.12%) & bad backbone angles (0.03%) compared to other models & adequately approaches the benchmark parameters of 2.00 Å template.

  1. M.S. Johnson, N. Srinivasan, R. Sowdhamini, T.L. Blundell. Crit Rev Biochem Mol Biol.;29(1):1–68, 1994.
  2. Lucy R. Forrest, Christopher L. Tang, and Barry Honig. Biophysical Journal, Volume 91 ,July , 508–517. 2006.
  3. S.B. Wortmann, F.M. Vaz, T. Gardeitchik, L.E. Vissers, G.H. Renkema, J.H. Schuurs- Hoeijmakers, W. Kulik, M. Lammens, C. Christin, L.A. Kluijtmans,R.J. Rodenburg, L.G. Nijtmans, A. Grunewald, C. Klein, J.M. Gerhold, T. Kozicz, P.M. van Hasselt, M. Harakalova and A.P. de Brouwer.Nat. Genet. 44:797-802, 2012.