Revealing the functionality of hypothetical protein KPN00728 from Klebsiella pneumonia e MGH78578: molecular dynamics simulation approaches
© Choi et al; licensee BioMed Central Ltd. 2011
Published: 30 November 2011
Previously, the hypothetical protein, KPN00728 from Klebsiella pneumoniae MGH78578 was the Succinate dehydrogenase (SDH) chain C subunit via structural prediction and molecular docking simulation studies. However, due to limitation in docking simulation, an in-depth understanding of how SDH interaction occurs across the transmembrane of mitochondria could not be provided.
In this present study, molecular dynamics (MD) simulation of KPN00728 and SDH chain D in a membrane was performed in order to gain a deeper insight into its molecular role as SDH. Structural stability was successfully obtained in the calculation for area per lipid, tail order parameter, thickness of lipid and secondary structural properties. Interestingly, water molecules were found to be highly possible in mediating the interaction between Ubiquinone (UQ) and SDH chain C via interaction with Ser27 and Arg31 residues as compared with earlier docking study. Polar residues such as Asp95 and Glu101 (KPN00728), Asp15 and Glu78 (SDH chain D) might have contributed in the creation of a polar environment which is essential for electron transport chain in Krebs cycle.
As a conclusion, a part from the structural stability comparability, the dynamic of the interacting residues and hydrogen bonding analysis had further proved that the interaction of KPN00728 as SDH is preserved and well agreed with our postulation earlier.
In the genome map of an organism, there are genes which code for hypothetical proteins. They contribute about 20 to 40% of total proteins . The only information can be obtained on hypothetical protein is from their nucleotide and amino acid sequences as rather few experimental data is found for this category of proteins. Despite many years of investigation, the annotations of these proteins have yet to progress significantly. Hence, these hypothetical proteins provide large research opportunities to scientists to elucidate their structures and functions especially those from pathogens .
Approximately 20% of 4776 protein coding genes of Klebsiella pneumoniae MGH78578 pathogen are classified as hypothetical proteins . K. pneumoniae is an opportunistic pathogen which affects patients with weakened immune system and/or underlying diseases . Elucidating the structures and functions of these hypothetical proteins will help to give insight to the possible roles and mechanisms of these proteins in relation to the pathogenesis or survivability of the pathogen. In addition to this, new functions may also emerge from protein complexes. All the information obtained can be a stimulant for further drug discovery efforts in the future.
Previously, via homology modeling and docking studies, we postulated that hypothetical protein KPN00728 (gi: 152969292) is the chain C subunit of Succinate dehydrogenase (SDH) . In both eukaryotic and prokaryotic organisms, SDH plays an important role in the aerobic respiratory chain specifically in the Krebs cycle which occurs in the transmembrane (TM) region of mitochondria. Our previous study showed that KPN00728 has a missing region containing conserved amino acid residues important for Ubiquinone (UQ), the natural ligand of SDH and heme group binding. Secondary structure and TM topology analyses showed that KPN00728 adopts SDH (subunit C)-like structure. Evolutionary relationship across 7 other Enterobacteriaceae was analyzed and showed that they are highly conserved. Molecular docking simulation on the other hand showed that UQ docked well onto the built model (consisting of KPN00728 and the annotated SDH chain D-KPN00729). Formation of hydrogen bonds between UQ and Ser27, Arg31 (from KPN00728) and Tyr83 (from KPN00729) further reinforced that KPN00728 hypothetical protein together with KPN00729 preserved the functionality of UQ binding. This observation strongly supported the possibility that KPN00728 is indeed chain C of SDH.
Although docking simulations enabled us to understand the preferred orientation of UQ when bound to the built model to form a stable complex, there were however, limitations. In docking simulation, rigidity of the built protein model and target of docking location are defined by the user. Hence this decreases the degree of freedom of both interacting components during the simulation. Furthermore, results from docking can only provide a single snapshot of the ligand orientation which does not represent a global, real-time picture of the dynamics of the interactions. Therefore, in this present study, molecular dynamics simulation was employed to obtain an in-depth understanding of the structure and function of KPN00728 as chain C of SDH across a successfully built model of the membrane environment of mitochondria.
Results and discussion
Membrane structure and selection on type of membrane
Stability of the system
Dynamic behaviour of Lipid membrane
The dynamics of the lipid membrane in this simulation was investigated to ascertain that our model is fully hydrated and comparable to experimental results as well as to previous simulations [7–9]. The properties investigated include lipid hydration, area per lipid, thickness of the membrane and order parameter of the hydrocarbon chains. It is clear from the results described below that our membrane adopted fully hydrated bilayer membrane behaviour.
Comparison of various POPC membrane protein systems with different hydration level
Number of lipid
Number of water
Alamethicin helices in a bilayer and in solution: molecular dynamics simulations 
Molecular dynamics study of the internal water molecules in vasopressin and oxytocin receptors 
Combined monte carlo and molecular dynamics simulation of fully hydrated dioleyl and palmitoyl-oleyl phosphatidylcholine lipid bilayers 
Comparison of area per lipid in simulation and experimental value in previous studies
Area per lipid (Å2)
Lipid Models for United-Atom Molecular Dynamics Simulations of Proteins 
Structure of fully hydrated fluid phase lipid bilayers with monounsaturated chains 
Molecular characterization of gel and liquid-crystalline structures of fully hydrated POPC and POPE bilayers. 
Phosphatidylcholine acyl unsaturation modulates the decrease in interfacial elasticity induced by cholesterol 
Performance of the general amber force field in modeling aqueous POPC membrane bilayers 
Structural information from multilamellar liposomes at full hydration: full q-range fitting with high quality X-ray data. 
Where θ is the angle between the CD bond and the bilayer normal (bilayer molecular axis), and the angle brackets indicate that values are averaged over all equivalent atoms and over time.
Dynamics of succinate dehydrogenase
Distance between Ubiquinone and Heme interacting
MD result (Å)
Docking result (Å)
TYR83@OH and UQ@O1
2.68 ± 0.49
ARG31@NH1 and UQ@O2
4.40 ± 1.26
SER27@OG and UQ@O3
8.83 ± 2.84
HIS84@ND and HEME@FE
7.07 ± 0.43
HIS71@ND and HEME@FE
6.43 ± 0.79
Solvation effect on the UQ binding
Radial distribution function was done between the ubiquinone and protein interacting residues
Potential interacting residues
Intensity g (r)
Distance, r (Å)
Average number of particles
Hydrogen acceptor at ubiquinone
Interacting residues at built model
RDF of Ser27@OG with OW also showed the highest intensity of 3.55 at 1.66 Å with an average number of water particles of ~2.77. This result also indicated that Ser27@OG might be able to form hydrogen bond with water molecules as the distance is within the H-bond cut-off value. Although Arg31@NH1 and Arg31@NH2 were also postulated to act as H-bond donors during interaction with UQ, the distance between Arg31@NH1 and UQ@O2 is far apart and the possibility of a H-bond formed between them is low. Thus, we suspect that water might play a role between them by forming water-mediating H-bond.
In RDF calculation, the intensity for both NH1 and NH2 groups from Arg31 with OW were low. The possibility of finding water molecules at 3.2 Å and 2.15 Å were as low as 0.19 and 0.14, respectively. Based on these results, the possibility of water to appear around both NH1 and NH2 from Arg31 is low. To further prove water molecules might be responsible in creating the drift of UQ from the interacting residues, which eventually eliminate the hydrogen bonding between UQ and the binding site residues, H-bond analysis between the interacting residues and UQ with water was performed.
A 5 Å shell was set around the binding site residues from SDH and UQ (Table 5). The analyses showed a minimum of one water mediated H-bond was found in more than 90.0% of the simulation time around Ser27@OG and UQ@O3. In addition, 39.4% of the trajectory appeared to have two water-mediating H-bonds. For UQ@O3, more than 55.0% of the trajectory consist at least two water-mediating H-bonds between UQ@O3 and protein. Only 0.74% and 0.68% of the trajectories with no water-mediating H-bond between Ser27@OG and UQ@O3. There was at least one water-mediating hydrogen bond appeared between them during the simulation. Water molecules that went into the binding site create a polar environment in the binding site which agreed well to the condition for electron transfer process in the Krebs cycle. However, we were not able to find any static water molecule which might be responsible for the interaction between UQ and Ser27@OG. All the waters appeared around the binding pocket and the longest occupancies were not more than 2 ps. This corresponded well with the RDF analysis.
Hydrogen bond analysis between those interacting residues and UQ with water within 5Å of the interacting atom
No. of hydrogen bond (HB)
5Å around SER27@OG
5Å around UQ@O3
5Å around ARG31@NH1
5Å around UQ@O2
No. of trajectory
No of trajectory
No. of trajectory
No. of trajectory
Functional implication derived from MD simulations
Oxidation of succinate to fumarate and reduction of UQ in the mitochondrial respiratory chain are carried out by SDH. Both processes utilized protons, H+. Due to SDH electroneutrality, it does not generate a proton motive force during catalysis. However, it formed a complex electron relay system which generate chemical energy through create proton gradient environment across the TM. Thus, the polar environment at the UQ binding site is very important in creating such a proton motive force during the catalysis of SDH . Water molecules and the polarity of the interacting amino acids residues might have contributed in creating a polar environment. In the crystal structure of the template, Asp95 and Glu101 at chain C of SDH, Gln78 from chain D of SDH are located at the fringe of the water channel. These residues operate as a proton wire connecting the cytoplasm to the UQ binding site. Mutations studies were done in all these residues in order to eliminate the potential H-bond formation in water channel and altered the H-bonding network by manipulating the side chain . Substitution of Asp95 to Glu95 on chain C extended the side chain which might lead to the interruption of H-bonding network in the proposed water channel. Substitution of Glu101 to Leu101 (chain C) and Gln78 to Leu78 (chain D) had created a hydrophobic environment which inhibited the formation of H-bonds. The reduction of SDH turnover rate was observed in all these generated SDH variants . It has been demonstrated that in a pH8 environment where the H+ concentration in the cytoplasm decreased 90%, the enzyme turnover rate had decreased markedly . However, all these mutations did not suppress the growth of the K. pneumoniae entirely. Hence, they proposed the existence of an alternative proton wire or pathway which involves the Asp15 residue from chain D of SDH .
UQ has two carbonyl and methoxy groups and one hydrophobic carbon tail. In our docking simulation, the positions of the carbonyl and methoxy groups of UQ were facing toward the protein. Strong H-bond was observed between carbonyl O4 and Tyr83 from chain D. On the other hand, the O1 carbonyl from UQ drifted toward the entrance of the water channel and was surrounded by water molecules. This solvation effect was most probably caused by 2 electron lone pairs of the carbonyl group. As for the hydrophobic carbon tail of UQ, it remained at the inner side of the entrance by avoiding the water molecules as shown in Figure 12. No significant changes in the orientation of the carbon tail were observed.
In our present study, MD simulation was used to give further insight into the functionality of our built model of KPN00728 hypothetical protein from Klebsiella pneumoniae MGH78578 as chain D of SDH. This was achieved by investigating the dynamics of its interaction with UQ and chain D of SDH across a transmembrane environment which was successfully established in this study. The stability of the simulation correlated well with major experimental parameters which are important for dynamic study of binding interaction of UQ and SDH. Both Ser27 and Arg31 had failed to demonstrate the possibility of forming H-bond with UQ. However, interestingly, analysis on simulation trajectories indicated that water-mediating H-bond did indeed exist and was found sandwiched between Arg31@NH1 and UQ@O2. Water molecules also appeared to be around Ser27. The occurrence of these water molecules around the binding site of UQ indicated that they might be responsible for the interaction involving binding of UQ to SDH. Examination of the structural properties at the binding site revealed that polar residues such as Asp95 and Glu101 (KPN00728), Asp15 and Glu78 (chain D SDH) were conserved and located at the entrance of the channel believed to be a water channel. The polarity of these residues might create a proton motive force which is responsible in transferring protons from the water channel or cytoplasm. The observation of this MD study had provided conclusive evidence that KPN00728 is indeed part of SDH.
Setup of the simulation system
A total of 29153 single point charge (SPC) waters were added into the simulation box with the thickness of ~27 Å away from the lipid headgroup. Three counterions Cl- were added to compensate for the net charge of the system, resulting in the system to be comprised of 111826 atoms. A total of 32,632 minimization steps were performed, starting with steepest descent (SD) and ended with conjugate gradient (CG), to remove unfavourable contacts. Subsequently, the system was subjected to equilibration in two phases. NVT equilibration was done to equilibrate the temperature of the entire system using Berendsen temperature coupling for 200 ps , whereby the protein complex was under position restraint condition. Then, NPT equilibration of the system on the protein complex was done for 2 ns under restraint condition. Nose-Hoover thermostat  was used to produce a correct kinetic ensemble and to allow molecular fluctuations within the system for more natural dynamics simulation. Semi-isotropic pressure coupling was applied. Upon completion of the two equilibration phases, the system was well equilibrated at the desired temperature and pressure. This was followed by the production run without any restraint. A total of 18 ns production MD was performed using an NPT ensemble.
Molecular dynamics simulation of the built model/membrane protein was performed using GROMACS v4.0.5 package  under NPT ensemble at a pressure of 1 bar and temperature of 300K. The GROMOS96 53a force field was used for both built model and lipid bilayer system . All bond lengths were constrained to their equilibrium value using SETTLE algorithm for water  and LINCS algorithm for the bonds between heavy atom and hydrogen atoms in protein, lipids and peptides . Integration time step of 2 fs was used and the neighbour list to calculate non-bonded interaction was updated every 10 time steps during the entire simulation time. A cut-off of 12 Å for Coulombic and van der Waals interactions was applied. Correction of long range electrostatics was done using Particle Mesh Ewald method (PME)  with a fourth-order spline interpolation and Fourier grid spacing of 0.12 nm. Periodic boundary condition in all directions was applied in the simulation.
During NVT equilibration simulation, Berendsen temperature coupling method  was used with a temperature coupling constant (τT) of 0.1 ps. Each group (peptide, lipids, solvent/ions) was coupled to a separate temperature bath. Subsequently, in NPT equilibration and production simulation, pressures were applied independently using Parrinello-Rahman pressure coupling approach [33, 34] in the x and y directions. To pack the lipids around the peptide and accelerate equilibration, a weak pressure coupling of 1.0 bar is given in the x and y directions with a pressure coupling constant (τP) of 5.0 ps.
This work was supported by the USM-RU grant (1001/PBIOLOGI/815014) Sy Bing, Choi gratefully acknowledge Universiti Sains Malaysia for the support of USM fellowship.
This article has been published as part of BMC Bioinformatics Volume 12 Supplement 13, 2011: Tenth International Conference on Bioinformatics – First ISCB Asia Joint Conference 2011 (InCoB/ISCB-Asia 2011): Bioinformatics. The full contents of the supplement are available online at http://www.biomedcentral.com/1471-2105/12?issue=S13.
- Galperin MY: Conserved 'hypothetical' proteins: new hints and new puzzles. Comparative and Functional Genomics 2001, 2(1):14–18. 10.1002/cfg.66PubMed CentralView ArticlePubMedGoogle Scholar
- Philalay JS, Palermo CO, Hauge KA, Rustad TR, Cangelosi GA: Genes required for intrinsic multidrug resistance in Mycobacterium avium . Antimicrobials Agents and Chemotherapy 2004, 48(9):3412–3418. 10.1128/AAC.48.9.3412-3418.2004View ArticleGoogle Scholar
- Lubec G, Afjehi-Sadat L, Yang JW, John JPP: Searching for hypothetical proteins: theory and practice based upon original data and literature. Prog Neurobiol 2005, 77(1–2):90–127. 10.1016/j.pneurobio.2005.10.001View ArticlePubMedGoogle Scholar
- Kawai T: Hypermucoviscosity: an extremely sticky phenotype of Klebsiella pneumoniae associated with emerging destructive tissue abscess syndrome. Clinical Infectious Diseases 2006, 42(10):1359–1361. 10.1086/503429View ArticlePubMedGoogle Scholar
- Choi SB, Normi YM, Wahab HA: Why hypothetical protein KPN00728 of Klebsiella pneumoniae should be classified as chain C of succinate dehydrogenase? Protein J 2009, 28: 415–427. 10.1007/s10930-009-9209-9PubMed CentralView ArticlePubMedGoogle Scholar
- Alberts B: Molecular biology of the cell. 4th edition. New York: Garland Science; 2002.Google Scholar
- Kucerka N, Tristram-Nagle S, Nagle JF: Structure of fully hydrated fluid phase lipid bilayers with monounsaturated chains. Journal of Membrane Biology 2005, 208(3):193–202.View ArticlePubMedGoogle Scholar
- Kukol A: Lipid models for united-atom molecular dynamics simulations of proteins. Journal of Chemical Theory and Computation 2009, 5(3):615–626. 10.1021/ct8003468View ArticleGoogle Scholar
- Leekumjorn S, Sum AK: Molecular characterization of gel and liquid-crystalline structures of fully hydrated POPC and POPE bilayers. The Journal of Physical Chemistry B 2007, 111(21):6026–6033. 10.1021/jp0686339View ArticlePubMedGoogle Scholar
- Tieleman DP, Sansom MS, Berendsen HJ: Alamethicin helices in a bilayer and in solution: molecular dynamics simulations. Biophysical Journal 1999, 76(1 Pt 1):40–49.PubMed CentralView ArticlePubMedGoogle Scholar
- Slusarz MJ, Slusarz R, Ciarkowski J: Molecular dynamics study of the internal water molecules in vasopressin and oxytocin receptors. Protein & Peptide Letters 2009, 16(4):342–350. 10.2174/092986609787848072View ArticleGoogle Scholar
- Chiu SW, Jakobsson E, Subramaniam S, Scott HL: Combined monte carlo and molecular dynamics simulation of fully hydrated dioleyl and palmitoyl-oleyl phosphatidylcholine lipid bilayers. Biophysical Journal 1999, 77(5):2462–2469. 10.1016/S0006-3495(99)77082-7PubMed CentralView ArticlePubMedGoogle Scholar
- Jojart B, Martinek TA: Performance of the general amber force field in modeling aqueous POPC membrane bilayers. J Comput Chem 2007, 28(12):2051–2058. 10.1002/jcc.20748View ArticlePubMedGoogle Scholar
- Smaby JM, Momsen MM, Brockman HL, Brown RE: Phosphatidylcholine acyl unsaturation modulates the decrease in interfacial elasticity induced by cholesterol. Biophysical Journal 1997, 73(3):1492–1505. 10.1016/S0006-3495(97)78181-5PubMed CentralView ArticlePubMedGoogle Scholar
- Pabst G, Rappolt M, Amenitsch H, Laggner P: Structural information from multilamellar liposomes at full hydration: full q-range fitting with high quality x-ray data. Physical Review E 2000, 62(3):4000–4009. 10.1103/PhysRevE.62.4000View ArticleGoogle Scholar
- Stryer L: Biochemistry. 4th edition. New York: W.H. Freeman; 1995.Google Scholar
- Lantzch G, Binder H, Heerklotz H, Wendling M, Klose G: Surface areas and packing constraints in POPC C (12)EO (n) membranes. A time-resolved fluorescence study. Biophysical Chemistry 1996, 58(3):289–302. 10.1016/0301-4622(95)00108-5View ArticlePubMedGoogle Scholar
- Tieleman DP, Marrink SJ, Berendsen HJ: A computer perspective of membranes: molecular dynamics studies of lipid bilayer systems. Biochimica et Biophysica Acta 1997, 1331(3):235–270.View ArticlePubMedGoogle Scholar
- Horne WS, Price JL, Keck JL, Gellman SH: Helix bundle quaternary structure from alpha/beta-peptide foldamers. Journal of American Chemical Society 2007, 129(14):4178–4180. 10.1021/ja070396fView ArticleGoogle Scholar
- Brändén C-I, Tooze J: Introduction to protein structure. New York: Garland Pub; 1991.Google Scholar
- Petsko GA, Ringe D: Protein structure and function. London, Sunderland, MA,Oxford: New Science Press; 2004.Google Scholar
- Jarosch R: The alpha-helix, an overlooked molecular motor. Protoplasma 2005, 227(1):37–46. 10.1007/s00709-005-0136-0View ArticlePubMedGoogle Scholar
- Cheng VW, Johnson A, Rothery RA, Weiner JH: Alternative sites for proton entry from the cytoplasm to the quinone binding site in Escherichia coli succinate dehydrogenase. Biochemistry-Us 2008, 47(35):9107–9116. 10.1021/bi801008eView ArticleGoogle Scholar
- Van Der Spoel D, Lindahl E, Hess B, Groenhof G, Mark AE, Berendsen HJ: GROMACS: fast, flexible, and free. J Comput Chem 2005, 26(16):1701–1718. 10.1002/jcc.20291View ArticlePubMedGoogle Scholar
- Schuttelkopf AW, van Aalten DMF: PRODRG: a tool for high-throughput crystallography of protein-ligand complexes. Acta Crystallographica, Section D, Biological Crystallography 2004, 60: 1355–1363. 10.1107/S0907444904011679View ArticleGoogle Scholar
- Gasteiger J, Marsili M: Iterative partial equalization of orbital electronegativity - a rapid access to atomic charges. Tetrahedron 1980, 36(22):3219–3228. 10.1016/0040-4020(80)80168-2View ArticleGoogle Scholar
- Berendsen HJC, Postma JPM, Vangunsteren WF, Dinola A, Haak JR: Molecular-Dynamics with Coupling to an External Bath. J Chem Phys 1984, 81(8):3684–3690. 10.1063/1.448118View ArticleGoogle Scholar
- Hoover WG: Canonical dynamics: equilibrium phase-space distributions. Physical Review A 1985, 31(3):1695–1697. 10.1103/PhysRevA.31.1695View ArticlePubMedGoogle Scholar
- Oostenbrink C, Villa A, Mark AE, van Gnsteren WF: A biomolecular force field based on the free enthalpy of hydration and solvation: the GROMOS force-field parameter sets 53A5 and 53A6. J Comput Chem 2004, 25(13):1656–1676. 10.1002/jcc.20090View ArticlePubMedGoogle Scholar
- Miyamoto S, Kollman PA: Settle - an analytical version of the shake and rattle algorithm for rigid water models. J Comput Chem 1992, 13(8):952–962. 10.1002/jcc.540130805View ArticleGoogle Scholar
- Hess B, Bekker H, Berendsen HJC, Fraaije JGEM: LINCS: A linear constraint solver for molecular simulations. J Comput Chem 1997, 18(12):1463–1472. 10.1002/(SICI)1096-987X(199709)18:12<1463::AID-JCC4>3.0.CO;2-HView ArticleGoogle Scholar
- Darden T, York D, Pedersen L: Particle Mesh Ewald - an N.Log(N) method for Ewald sums in large systems. J Chem Phys 1993, 98(12):10089–10092. 10.1063/1.464397View ArticleGoogle Scholar
- Bussi G, Donadio D, Parrinello M: Canonical sampling through velocity rescaling. J Chem Phys 2007, 126(1):14101–14107. 10.1063/1.2408420View ArticleGoogle Scholar
- Parrinello M, Rahman A: Polymorphic transitions in single-crystals - a new molecular-dynamics method. J Appl Phys 1981, 52(12):7182–7190. 10.1063/1.328693View ArticleGoogle Scholar
This article is published under license to BioMed Central Ltd. This is an open access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.