Stereoselective Biocatalytic Synthesis
FULL PAPER
Computational methods
rameters. This protocol was repeated with both starting structures and
the resulting free energy differences were averaged.
Docking calculations and molecular mechanics: Models for both enantio-
mers of cyanohydrin 5 were built and optimised using the program Sybyl
v6.5 (Tripos Inc.). Partial atomic charges for these compounds were cal-
culated using the RESP protocol.[42] For the hydroxynitrile lyase from
Hevea brasiliensis (HbHNL), protein coordinates were taken from the re-
spective atomic resolution X-ray crystal structure (PDB-entry: 1qj4).[43]
For the enzyme from Prunus amygdalus (PaHNL5), a homology model
of isoenzyme no. 5 was used as in previous modelling and engineering
studies.[34,35] This model is based on the crystal structure of isoenzyme no.
1 (PaHNL1, PDB-entry: 1ju2)[44] which shares about 75% sequence iden-
tity. In both protein models, Asp, Glu, Arg and Lys residues were treated
as charged. Protonation and tautomerisation states of His residues were
chosen that resulted in sensible hydrogen bonding networks. Hydrogen
atoms were added to the structure, followed by a geometry optimisation
using AMBER,[45] applying harmonic restraints on the positions of all
heavy atoms. Only polar hydrogen atoms of the protein and the ligands
were retained for the docking simulations.
[1] L. Rosenthaler, Biochem. Z. 1908, 14, 238–253.
[2] F. Effenberger in Stereoselective Biocatalysis (Ed.: R. N. Patel),
Marcel Dekker, New York, 2000, pp. 321–342.
[3] J. Brussee, A. van der Gen in Stereoselective Biocatalysis (Ed.: R. N.
Patel), Marcel Dekker, New York, 2000, pp. 289–320.
[4] R. J. H. Gregory, Chem. Rev. 1999, 99, 3649–3682.
[5] M. North, Synlett 1993, 807–820.
[6] M. H. Fechter, H. Griengl in Enzyme Catalysis in Organic Synthesis,
Vol. 2 (Eds.: K. Drauz, H. Waldmann), Wiley-VCH, Weinheim,
2002, 2nd ed., pp. 974–989.
[7] M. North, Tetrahedron: Asymmetry 2003, 14, 147–176.
[8] M. H. Fechter, H. Griengl, Food Technol. Biotechnol. 2004, 42, 287–
294.
[9] R. F. G. Frçhlich, A. A. Zabelinskaja-Mackova, M. H. Fechter, H.
Griengl, Tetrahedron: Asymmetry 2003, 14, 355–362.
[10] P. Poechlauer, W. Skranc, M. Wubbolts in Asymmetric catalysis on
industrial scale (Eds.: H. U. Blaser, E. Schmidt), Wiley-VCH, Wein-
heim, 2004, pp. 151–164.
[11] H. Bühler, A. Bayer, F. Effenberger, Chem. Eur. J. 2000, 6, 2564–
2571.
[12] F. Effenberger, B. Hoersch, F. Weingart, T. Ziegler, S. Kuehner, Tet-
rahedron Lett. 1991, 32, 2605–2608.
[13] H. Griengl, N. Klempier, P. Pçchlauer, M. Schmidt, N. Shi, A. A.
Zabelinskaja-Mackova, Tetrahedron 1998, 54, 14477–14486.
[14] E. Kiljunen, L. T. Kanerva, Tetrahedron: Asymmetry 1997, 8, 1551–
1557.
[15] F. Effenberger, S. Heid, Tetrahedron: Asymmetry 1995, 6, 2945–
2952.
[16] R. J. H. Gregory, S. M. Roberts, J. V. Barkley, S. J. Coles, M. B.
Hursthouse, D. E. Hibbs, Tetrahedron Lett. 1999, 40, 7407–7411.
[17] F. Effenberger, J. Roos, C. Kobler, H. Bühler, Can. J. Chem. 2002,
80, 671–679.
[18] F. Effenberger, J. Roos, C. Kobler, Angew. Chem. 2002, 114, 1957–
1959; Angew. Chem. Int. Ed. 2002, 41, 1876–1879.
[19] A. Nahrstedt, J. D. Kant, V. Wray, Phytochemistry 1982, 21, 101–
105.
[20] M. Avi, M. H. Fechter, K. Gruber, F. Belaj, P. Poechlauer, H.
Griengl, Tetrahedron 2004, 60, 10411–10418.
[21] L. Tan, C. Chen, W. Chen, L. Frey, A. O. King, R. D. Tillyer, F. Xu,
D. Zhao, E. J. J. Grabowski, P. J. Reider, P. O’Shea, P. Dagneau, X.
Wang, Tetrahedron 2002, 58, 7403–7410.
[22] R. W. Friesen, C. Brideau, C. C. Chan, S. Charleson, D. Deschenes,
D. Dube, D. Ethier, R. Fortin, J. Y. Gauthier, Y. Girard, R. Gordon,
G. M. Greig, D. Riendeau, C. Savoie, Z. Wang, E. Wong, D. Visco,
L. J. Xu, R. N. Young, Bioorg. Med. Chem. Lett. 1998, 8, 2777–2782.
[23] G. Cardillo, M. A. Hashem, C. Tomasini, J. Chem. Soc. Perkin Trans.
1 1990, 1487–1488.
The two enantiomers were docked to the active sites of these enzymes
using a Monte-Carlo simulated-annealing approach as implemented in
the program AutoDock v3.0.[46] The structure of the protein was kept
rigid, whereas the ligands possessed translational, rotational and three
torsional degrees of freedom. In order to probe the conformational space
of the five-membered rings, an open structure was created by breaking
the bond between C4 and C5. During the simulations, a constraint was
applied which kept the respective distance between 1.4 to 1.7 . For each
ligand, 100 randomly chosen initial structures were subjected to 150
Monte Carlo steps starting at an RT value of 1000 and using a cooling
factor of 0.95 per cycle. In each cycle, a maximum of 10000 accepted or
rejected moves were allowed. The resulting structures were clustered
with a root-mean-square tolerance of 1.5 . The structure of the lowest-
energy representative of each cluster was further optimised using pro-
grams from the AMBER package.[45]
During minimization, only the bound ligand and amino acid residues
within 12 of the ligand were allowed to move. Non-bonded interactions
were truncated to a sphere of 10 radius. A distance dependent dielec-
tric function was employed (er =r). The protocol for the geometry optimi-
sation consisted of three steps: a) an energy minimisation run employing
weak harmonic restraints (1 kcalmolꢀ1 ꢀ2) on the ligand, b) a 100 ps mo-
lecular dynamics simulation (T=298 K, Dt=2 fs) with snapshots taken
every 10 ps, and c) an unrestrained minimisation of the ten snapshot
structures. The structure with the lowest energy was then used in the
analysis.
Thermodynamic integration: In order to estimate the difference of bind-
ing free energy of the two enantiomers thermodynamic integration calcu-
lations were performed for the complexes with HbHNL. The two states
between which the free energy difference was calculated, were the S- and
R enantiomer of 5 bound in the active site of the enzyme. Inversion of
the absolute configuration was accomplished by perturbing the sulfur
atom to methylene group and the C5-methylene group to a sulfur atom
during a molecular dynamics simulation. The calculations were per-
formed with the program GIBBS from the AMBER package.[45] Starting
structures were the modeled complexes (see above) plus a 30 water
cap around the active site. All atoms except those from residues in the vi-
cinity of the ligand (within 12 ), the water cap and the ligand itself
were constrained. The time step for integration was set to 2 fs and all
bonds involving hydrogen atoms were constrained using the SHAKE al-
gorithm.[47] Each simulation was preceded by 3000 steps of geometry op-
timisation of the water cap, a 20 ps dynamics run during which the tem-
perature was continuously raised from 10 to 298 K and a 80 ps equilibra-
tion run at 298 K. In the “forward” direction the coupling parameter l
was decremented in 20 steps from 1 (representing the complex with the
R enantiomer) to 0 (representing in this case the complex with the S en-
antiomer) during a 210 ps molecular dynamics simulation (21 windows
with 2 ps of equilibration and 8 ps of data collection) at 298 K. After
that, the system was equilibrated for 20 ps at the final state (l=0), fol-
lowed by a “backward” perturbation run (l=0!1) using the same pa-
[24] T. Sugai, H. Kakeya, H. Ohta, J. Org. Chem. 1990, 55, 4643–4647.
[25] A. De Raadt, H. Griengl, H. Weber, Chem. Eur. J. 2001, 7, 27–31.
[26] S. Fçrster, J. Roos, F. Effenberger, H. Wajant, A. Sprauer, Angew.
Chem. 1996, 108, 493–494; Angew. Chem. Int. Ed. Engl. 1996, 35,
437–439.
[27] J. Albrecht, I. Jansen, M. R. Kula, Biotechnol. Appl. Biochem. 1993,
17, 191–203.
[28] S. Nanda, Y. Kato, Y. Asano, Tetrahedron 2005, 61, 10908–10916.
[29] A. de Raadt, H. Griengl, Curr. Opin. Biotechnol. 2002, 13, 537–542.
[30] Z. Li, J. B. van Beilen, W. A. Duetz, A. Schmid, A. de Raadt, H.
Griengl, B. Witholt, Curr. Opin. Chem. Biol. 2002, 6, 136–144.
[31] A. De Raadt, B. Fetz, H. Griengl, M. F. Klingler, I. Kopper, B.
Krenn, D. F. Munzer, R. G. Ott, P. Plachota, H. J. Weber, G. Brau-
negg, W. Mosler, R. Saf, Eur. J. Org. Chem. 2000, 3835–3847.
[32] K. Gruber, G. Gartler, B. Krammer, H. Schwab, C. Kratky, J. Biol.
Chem. 2004, 279, 20501–20510.
[33] I. Dreveny, C. Kratky, K. Gruber, Protein Sci. 2002, 11, 292–300.
Chem. Eur. J. 2007, 13, 3369 – 3376
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
3375