Computational Studies of Subtilisin-Catalyzed Transesterification of Sucrose: Importance of Entropic Effects

Computational Studies of Subtilisin-Catalyzed Transesterification of Sucrose: Importance of Entropic Effects
of 4
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
  ChemBioChem  2002  , No. 9 ¹ 2002 WILEY-VCH Verlag GmbH&Co. KGaA, Weinheim 1439-4227/02/03/09 $ 20.00+.50/0  907 Computational Studies of Subtilisin-Catalyzed Transesterification of Sucrose: Importance of EntropicEffects. Gloria Fuentes, [a, b] M. A¬ngeles Cruces, [b] FranciscoJ. Plou, [b] Antonio Ballesteros, [b] andChandraS. Verma* [a] KEYWORDS: carbohydrates  ¥  computer studies  ¥  entropy  ¥  enzymes  ¥ regioselectivity Introduction Regioselective acylation of sugars at their hydroxy groups iscurrently of great importance for the food, chemical, andpharmaceutical industries. [1] Enzymes are very useful for catalysisof these processes owing to their exquisite chemo-, stereo-, andregioselectivity, [2] and subtilisins are among the most researched.One of these specific transformations is the transesterification of sucrose with fatty acid esters with a high selectivity for the 1  -position (Figure 1; experimentally, 80±90% of the products arethe 1  -product while 2±5% are minor species that are largelymonoesters whose acylation positions could not be assigned;there is no available thermodynamic data on these reactions). [3] Figure 1.  Schematic representation of the subtilisin active site (catalytic triad and oxyanion hole) and the centres of reactions (the 6OH, 6  OH, and 1  OH sites) inthe sucrose moiety. The torsions across which conformational sampling wascarried out are indicated by dark lines. Experimental Section Surfaceplasmonresonancefluorescencespectroscopy(SPFS): TheSPFS binding experiments were carried out with a self-assembledset-up. [16] The CB3(IV) fragment of collagen type IV and the twosynthetic heterotrimers were fluorescence-labeled with Cy5 dye(Amersham Pharmacia, Uppsala, Sweden) and the purified human  1  1 integrin (Chemicon Inc., Temecula, Ca) was embedded into adimyristoyl phosphatidylethanolamine/phosphatidylcholine bilayercoated onto a gold surface through use of a hydrophilic lamininpeptide layer. The binding experiments were performed at 25  C intris(hydroxymethyl)aminomethane (Tris)¥HCl (50 m M , pH 7.4), NaCl(150 m M ), MgCl 2 ¥6H 2 O (2 m M ), and MnCl 2 ¥2H 2 O (1 m M ). Isothermal titration calorimetry (ITC) : ITC measurements wereperformed at 4  C on MicroCal VP-ITC equipment by titration of asolution (0.5   M ) of    1  1 integrin Fos/Jun construct in Tris¥HCl(50 m M , pH 7.4), NaCl (150 m M ), MgCl 2 ¥6H 2 O (2 m M ), and MnCl 2 ¥2H 2 O(1 m M ) with a solution of each heterotrimer (20   M ) in the samebuffer. The data were fitted by using a nonlinear least squaresMarquardt algorithm with three independent variables,  K  bind ,   H  bind ,and the stoichiometry  N   (here fixed as 1); [17, 18] a single set of identicalsites was used as the fitting model. Modeling Experiments : Complex models were created by homol-ogy modeling with the program MAIN. [19] Energy minimization wasperformed starting from the X-ray structure 1QC5 [14] of the   1  1Idomain and a model for the heterotrimers based on the structure of the homotrimer in 1DZI. [12] In addition to the atomic force field,constraints were used to tether homologous C   positions of thecomplex model to the 1DZI structure. This study was supported by SFB 563 of the Technical University of Munich. Elisabeth Weyher is gratefully acknowledged for theexcellent technical assistance. [1] B. Trueb, B. Grobli, M. Spiess, B. F. Odermatt, K. H. Winterhalter,  J. Biol.Chem.  1982 ,  257  , 5239±5245.[2] P. Vandenberg, A. Kern, A. Ries, L. Luckenbill-Edds, K. Mann, K. K¸hn,  J. Cell Biol.  1991 ,  113 , 1475±1483.[3] P. D. Yurchenco, J. C. Schittny,  FASEB J.  1990 ,  4 , 1577±1590.[4] R. O. Hynes,  Cell   1987 ,  48 , 549±554.[5] D.S. Tuckwell, M.J. Humphries,  Crit. Rev. Oncol. Hematol.  1993 ,  15 , 149±171.[6] M. D. Pierschbacher, E. Ruoslahti,  Nature  1984 ,  309 , 30±33.[7] K. K¸hn, J. Eble,  Trends Cell Biol.  1994 ,  4 , 256±261.[8] J. A. Eble, R. Golbik, K. Mann, K. K¸hn,  EMBO J.  1993 ,  12 , 4795±4802.[9] R. Golbik, J. A. Eble, A. Ries, K. K¸hn,  J. Mol. Biol.  2000 ,  297  , 501±509.[10] B. Sacca¡, L. Moroder,  J. Pept. Sci.  2002 ,  8 , 192±204.[11] J. Ottl, D. Gabriel, G. Murphy, V. Kn‰uper, Y. Tominaga, H. Nagase, M.Krˆger, H. Tschesche, W. Bode, L. Moroder,  Chem. Biol.  2000 ,  7  , 119±132.[12] J. Emsley, C. G. Knight, R. W. Farndale, M. J. Barnes, R. C. Liddington,  Cell  2000 ,  101 , 47±56.[13] C. G. Knight, L. F. Morton, A. R. Peachey, D. S. Tuckwell, R. W. Farndale, M. J.Barnes,  J. Biol. Chem.  2000 ,  275 , 35±40.[14] R. L. Rich, C. C. S. Deivanayagam, R. T. Owens, M. Carson, A. Hook, D.Moore, V. W. C. Yang, S. V. L. Narayana, M. Hook,  J. Biol. Chem.  1999 ,  274 ,24906±24913.[15] S. Kumar, B. Ma, C. J. Tsai, R. Nussinov,  Proteins  2000 ,  38 , 368±383.[16] E. K. Schmidt, T. Liebermann, M. Kreiter, A. Jonczyk, R. Naumann, A.Offenhausser, E. Neumann, A. Kukol, A. Maelicke, W. Knoll,  Biosensors &Bioelectronics  1998 ,  13 , 585±591.[17] I. Jelesarov, H. R. Bosshard,  J. Mol. Recognit.  1999 ,  12 , 3±18.[18] J. E. Ladbury, B. Z. Chowdhry,  Chem. Biol.  1996 ,  3 , 791±801.[19] D. Turk, Thesis, Technische Universit‰t M¸nchen,  1992 .Received: May 14, 2002 [Z418][a]  Dr. C. S. Verma, G. FuentesStructural Biology Laboratory Department of Chemistry University of York, York, YO10 5DD (UK)Fax: (   44)1904-410519E-mail:  [b]  G. Fuentes, Dr. M. A¬ . Cruces, Dr. F. J. Plou, Prof. A. BallesterosDepartamento de Biocata¬ lisisInstituto de Cata¬ lisisCSIC, Cantoblanco, 28049 Madrid (Spain)  908  ¹ 2002 WILEY-VCH Verlag GmbH&Co. KGaA, Weinheim 1439-4227/02/03/09 $ 20.00+.50/0 ChemBioChem  2002  , No. 9 Regioselective acylation of carbohydrates leads to a range of starting compounds for the controlled synthesis of a variety of polymers with diverse properties. [4] An understanding of struc-ture±function relationships is therefore desirable. We presenthere the results of a computational study that examines theorigins of this selectivity in the case of sucrose acylation.Previous studies [5] modeled transition-state structures and foundthat the 1  OH-adduct was preferred to adducts formed at theother OH groups as the 1  OH-adduct was the least stericallyconstrained. We present the results of a systematic study of thisacylation carried out by computation of the free energies of interaction of the possible transition-state adducts and find thatin the 1  OH-adduct the sucrose moiety is the most solventexposed. This exposure confers a higher mobility upon thesucrose, which translates into a larger entropic stabilization.Computation of the enthalpic components of interactionsbetween a protein and its ligands is relatively straightforwardand routine. However, computation of entropies has in the pastbeen more difficult, limited by the complexity of the problemand the intensive computations required. [6, 7] We apply a recentlydeveloped method that avoids the need for intensive compu-tations and yet seems to capture most of the entropic change ina ligand±protein complex and is thereby attractive for high-throughput computational screenings for lead compounds. [8] Observations The conformers with the lowest free energies (as a result of theprotein±ligand interactions and the vibrational entropic con-tribution from the librational modes of the ligand in its ™proteincage∫ [8] ) were chosen for analysis. Since we examined thetransition states, it is appropriate that the most stable state isone that maximizes the protein±ligand interactions. Our find-ings are discussed below.If only the interaction enthalpies are taken into account(Table 1), reaction at the 1  OH-position is marginally favorable,with a stability relative to reaction at 6OH smaller than k  T  (0.7 kcalmol  1 at 318 K). In contrast, when we include theentropic components in the free energy, we see that reaction at1  OH is clearly more favorable than at the other two sites (byabout 4±5 kcalmol  1 ).Van der Waals interactions dominate and account for about94% of the net interaction enthalpies. Sucrose accounts for 60±65% of these interactions while laurate contributes the rest. Thisdomination by van der Waals interactions is not surprisingconsidering that the surface of the ligands is 68% apolar (33%is formed by sucrose and the rest by laurate).Table 1 reveals that both enthalpy and entropy contributesimilar amounts to the net free energy. The discriminating factorbetween the 1  -adduct and the 6  -adduct is enthalpic, while it isentropic for the 6-adduct. In the 6  -adduct, the electrostatic andvan der Waals interactions of the lauroyl moiety are destabilizedby approximately 1±4k  T  . In contrast, one of the sucrose rings of the 6  -adduct is more buried than in the 1  -adduct and makeselectrostatic interactions that are more stable than in the 1  -adduct by about 2k  T   but are offset by van der Waals interactionsthat are less stable by around 1k  T   (Table 2). In the 6-adduct, theentropic destabilization relative to the 1  -adduct arises largelyfrom the sucrose. The sucrose in the 1  -adduct is slightly moreexposed than in the other two adducts and the increase in therelative librational degrees of freedom of the sucrose moietyenhance the stability of the 1  -adduct entropically by about 4k  T  .In all three adducts, the oxyanion hole caused by the anioniclaurate oxygen moiety is stabilized by the side-chain aminegroup of Asn155 and the backbone amide group of Ser221. Inaddition, the sucrose makes three hydrogen bonds to theprotein backbone atoms and one to the side-chain amine groupof Asn155 in the 1  -adduct. Five hydrogen bonds from thesucrose are made to the protein backbone and one to the sidechain of Thr220 in the 6  -adduct. Four hydrogen bonds are madeto the protein backbone and one to the side chain of Thr220 inthe 6-adduct.The sucrose moiety in both 1  - and 6  -adducts is localized inthe S1 pocket of the enzyme, while in the 6-adduct the glucosemoiety lies in the S1 pocket and the fructose moiety is in the S3pocket (Figure 2). The laurate is located near the S2 and S2  subsites in the 1  -, 6  -, and 6-adducts, respectively.Graphical inspection of the structures suggests two mutations(possible through site-directed mutagenesis) at sites that areexposed and hence unlikely to cause too much perturbation tothe protein structure and stability, and yet may lead to enhancedreactivity at the 1  -adduct: a) Gly127 could be mutated to anyneutral residue (Ser/Thr/Asn/Gln) to introduce hydrogen bondsbetween this amino acid side chain and the hydroxy groups atpositions 4 or 6 in the glucose moiety or the 6  -position in thefructose moiety; b) Ser155Asn mutation would increase thelength of the side chain and introduce an additional hydrogenbond between this side chain and the hydroxy group at the6-position of the sucrose. These mutations would, of course, be Table 1.  Components of free energy for the lowest-energy conformers. [a] Regioisomer Interaction energy[kcalmol  1 ] T   S  ( T   318 K)[kcalmol  1 ]  G [kcalmol  1 ]6OH   62.98 57.41   120.39(  6.45,  56.52)6  OH   60.64 61.12   121.76(  2.90,  57.74)1  OH   63.46 61.96   125.42(  6.04,  57.41)[a] Electrostatic and van der Waals components of interaction are given inparentheses. Table 2.  Contributions from the different ligands to the overall free energy of the protein. [a] Regioisomer Interaction energy[kcalmol  1 ]Electrostatics[kcalmol  1 ]Van der Waals[kcalmol  1 ]Entropy ( T   S )[kcalmol  1 ]nlau   24.0   3.9   21.1 21.06OH sucr   39.0   2.6   36.4 34.7nlau   20.8   1.0   19.8 21.76  OH sucr   39.9   1.9   38.0 34.1nlau   24.2   3.5   20.7 21.11  OH sucr   39.3   2.5   36.7 36.7[a] The entropy values apply at 318 K.  ChemBioChem  2002  , No. 9 ¹ 2002 WILEY-VCH Verlag GmbH&Co. KGaA, Weinheim 1439-4227/02/03/09 $ 20.00+.50/0  909 subject to some complexity as a result of enthalpy±entropycompensations. [14] One could also mutate destabilizing residues.Val176 is the only amino acid that destabilized the interactionsenthalpically (more than 0.1 kcalmol  1 ). This residue is quiteburied in the protein and so any mutation at this site could affectthe stability and folding of the protein. We have not come acrossany reports in the literature that report the effects of muta-genesis on this reaction.If minimization was discontinued after the gradient of potential was smaller than 10  2 kcalmol  1 ä  1 (as opposed tothe 10  4 kcalmol  1 ä  1 cut-off that we actually used for thisstudy), then the computational results indicate that the reactionis most favorable at 6  OH (interaction energies are  61.6,  63.1,and  61.4 kcalmol  1 for 6OH, 6  OH, and 1  OH, respectively). Thisresult is contrary to experimental observations and suggests thatextensive minimizations are necessary to reach minima thatallow for appropriate, more realistic discrimination between theproducts. However, the discrimination achieved when minimi-zation was discontinued earlier was made on purely enthalpicgrounds because in order to compute the entropic contributionsminimizations need to be performed until the gradient of potentials is smaller than 0.0001 kcalmol  1 ä  1 .In summary, we found that entropic factors are important indetermination of the selectivity of the enzyme subtilisin inacylation of sucrose. An appropriate choice of simple andrealistic models that take into account the enthalpy of inter-action of the ligands and the librational entropy of ligands intheir binding pockets can successfully identify the experimen-tally observed product of the transesterification of sucrose bysubtilisin as the lowest energy isomer. Methods The enzyme Subtilisin Carlsberg (PDB entry 1cse.pdb,resolved at 1.2 ä) [9] was used as the starting model.Structures that mimic the transition states for reactionat each of the three primary sites on the sucrose(1  OH, 6  OH, and 6OH, see Figure 1) were constructedby using the Quanta software (Accelrys Inc. SanDiego). The initial orientations of the ligands wereguided by examination of the structures of lipasesinhibited with alkylic compounds, [10] while the sucrosewas placed such that it was involved in hydrogenbonds with the catalytic triad and with the oxyanionhole (as shown in Figure 1). The Charmm force field [11] was used, with appropriate parameters [12] (electro-statics were modulated through a constant dielectricof 9, which approximately corresponds to the solventmixture conditions used in such reactions, for exam-ple, in the transesterification of disaccharides by usinghydrolases [13] ), for all calculations. Conformationalsearches were carried out across 5 torsions (high-lighted in Figure 1), by variation of each between the g  ,  g  , and  t   states and minimization of theconformer with decreasing constraints. [12] Minimiza-tions were continued until the gradient of potentialenergy was less than or equal to 0.0001 kcalmol  1 ä  1 .We excluded rotations across torsions that woulddisrupt the oxyanion-hole geometry (the anionicoxygen moiety in Figure 1 is stabilized by the back-boneamide group of Ser221 and theside chain aminogroup of Asn155; the protons on the ring nitrogens of the cationicHis64 are hydrogen bonded to the two oxygens atoms that arebonded to the tetrahedral carbon). This process was followed bycomputation of the vibrational modes of the bound ligand. [8] We thank the Spanish Ministry of Science & Technology and theBBSRC, UK, for support. [1] a) T. Watanabe,  Foods Food Ingredients J. Japan  1999 ,  180 , 18; b) K. Hill, O.Rhode,  Fett/Lipid   1999 ,  101 , 25±33; c) S. Okabe, M. Saganuma, Y. Tada, Y.Ochiai, E. Sueoka, H. Kohya, A.Shibata, M. Takahashi, M. Mizutani, T.Matsuzaki, H. Fujiki,  Jpn. J. Cancer Res.  1999 ,  90 , 669±676.[2] R. D. Schmid, R. Verger,  Angew. Chem.  1998 ,  110 , 1694±1720;  Angew.Chem. Int. Ed.  1998 ,  37  , 1608±1633.[3] a) G. Carrea, S. Riva, F. Secundo, B. Danieli,  J. Chem. Soc. Perk. Trans.  1, 1989 , 1057±1061; b) M. A. Cruces, C. Otero, M. Bernabe, M. Martin-Lomas, A. Ballesteros,  Ann. NY Acad. Sci.  1992 ,  672 , 436±443; c) T. Polat,H. G. Bazin, R. J. Linhardt,  J. Carbohydr. Chem.  1997 ,  16 , 1319±1325.[4] O.-J. Park, D.-Y. Kim, J. S. Dordick,  Biotechnol. Bioeng.  2000 ,  70 , 208±216.[5] J. O. Rich, B. A. Bedell, J. S. Dordick,  Biotechnol. Bioeng.  1995 ,  45 , 426±434.[6] R. J. Kazlauskas,  Curr. Opin. Chem. Biol.  2000 ,  4 , 81±88.[7] a) W. Wang, O. Donini, C. M. Reyes, P. A. Kollman,  Annu. Rev. Biophys.Biomol. Struct.  2001 ,  211 ±243; b) K. B. Ljungberg, J. Marelius, D. Musil, P.Svenson, B. Norden, J. Aqvist,  Eur. J. Pharm. Sci.  2001 ,  12 , 441±446; c) P. A.Kollman,  Chem. Rev.  1993 ,  93 , 2395±2417.[8] S. Fischer, J. C. Smith, C. S. Verma,  J. Phys. Chem. B  2001 ,  105 , 8050±8055.[9] W. Bode, E. Papamocos, D. Musil,  Eur. J. Biochem.  1996 ,  176 , 673±692.[10] J. Uppenberg, N. Ohrner, M. Norin, K. Hult, G. J. Kleywegt, S. Patkar, V.Waagen, T. Anthonsen, T. A. Jones,  Biochemistry   1995 ,  34 , 16838±16851.[11] B. R. Brooks, R. E. Bruccoleri, B. D. Olafson, D. J. States, S. Swaminathan, M.Karplus,  J. Comput. Chem.  1993 ,  4 , 187±217. Figure 2.  The minimum vibrational free energy structures for reactions at the three subsites onsucrose. The orientations of the ligands relative to the various subsites of subtilisin are shown.  910  ¹ 2002 WILEY-VCH Verlag GmbH&Co. KGaA, Weinheim 1439-4227/02/03/09 $ 20.00+.50/0 ChemBioChem  2002  , No. 9 [12] F. J. Plou, D. Kowlessur, J. P. G. Malthouse, G. W. Mellor, M. J. Hartshorn, S.Pinitglang, H. Patel, C. M. Topham, E. W. Thomas, C. Verma, K. Brockle-hurst,  J. Mol. Biol.  1996 ,  257  , 1088±1111.[13] M. Ferrer, M. A. Cruces, F. J. Plou, M. Bernabe, A. Ballesteros,  Tetrahedron 2000 ,  56 , 4053.[14] C. T. Calderone, D. H. Williams,  J. Am. Chem. Soc .  2001 ,  123 , 6262±6267.Received: March 25, 2002Revised: July 4, 2002 [Z386]
Similar documents
View more...
Related Search
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks

We need your sign to support Project to invent "SMART AND CONTROLLABLE REFLECTIVE BALLOONS" to cover the Sun and Save Our Earth.

More details...

Sign Now!

We are very appreciated for your Prompt Action!