N-aryl-N-hydroxy urethanes as peroxidase substrates

Article abstract of DOI:10.1007/s007060170138

N-Aryl-N-hydroxy urethanes (AHUs), which are promising N-OH mediators for oxidoreductase catalysis, are electrochemically active redox compounds with a redox potential of 0.35-0.40 V vs. SCE (0.59-0.64 V vs. NHE). Alkyl substituents in the carbamic acid residue have, as expected, only little influence on the redox potential. The highest potential was noticed for N-hydroxy-N-(4-cyanophenyl)-carbamic acid methyl ester. Recombinant Coprinus cinereus peroxidase (rCiP) catalyzes oxidation of the AHUs with apparent bimolecular constants k_ox of 2.5 · 10⁴-7.5 · 10⁴ M^-1 s^-1 at pH = 8.5 and 25°C. Structure-function connectivities of the substrates were analysed within the framework of the Marcus cross relationship and by using ab initio quantum chemical calculations. An excellent correlation of the redox potentials and the HOMO energies could be found. However, no correlation of log(k_ox) with redox potential and HOMO energy was indicated as predicted by theory. This was explained by specific docking of the substrates in the active center of rCiP.

Full text of DOI:10.1007/s007060170138

Monatshefte f uÈ r Chemie 132, 295±304 (2001)
N-Aryl-N-Hydroxy Urethanes as Peroxidase
Substrates
Juozas Kulys1 , Heinz-Josef Deussen , Kastis Krikstopaitis ,
;Ã
2
1
2
3
Palle Schneider , and Arturas Ziemys
1
Institute of Biochemistry, LT-2600 Vilnius, Lithuania
Novo Nordisk A/S, DK-2880 Bagsvñrd, Denmark
Vytautas Magnus University, Environmental Faculty, Department of Biology, LT-3500 Kaunas,
Lithuania
2
3
Summary. N-Aryl-N-hydroxy urethanes (AHUs), which are promising N±OH mediators for
oxidoreductase catalysis, are electrochemically active redox compounds with a redox potential of
0.35±0.40 V vs. SCE (0.59±0.64 V vs. NHE). Alkyl substituents in the carbamic acid residue have, as
expected, only little in¯uence on the redox potential. The highest potential was noticed for N-
hydroxy-N-(4-cyanophenyl)-carbamic acid methyl ester. Recombinant Coprinus cinereus peroxidase
4
(
rCiP) catalyzes oxidation of the AHUs with apparent bimolecular constants kox of 2.5 Á 10 ±
4
� 1 � 1
ꢀ
7
.5 Á 10 M
s
at pH ˆ 8.5 and 25 C.
Structure-function connectivities of the substrates were analysed within the framework of the
Marcus cross relationship and by using ab initio quantum chemical calculations. An excellent
correlation of the redox potentials and the HOMO energies could be found. However, no correlation
of log(kox) with redox potential and HOMO energy was indicated as predicted by theory. This was
explained by speci®c docking of the substrates in the active center of rCiP.
Keywords. Redox potential; Fungal peroxidase; Reactivity; ab initio Calculations; Substrate docking.
Introduction
A large variety of different organic redox compounds have been suggested as
mediators for peroxidase and laccase catalyzed oxidations [1]. In these reactions,
the enzyme catalyzes the oxidation of an organic molecule, the so-called mediator
(MH). The oxidized mediator then chemically oxidizes the target molecule (non-
substrate) and is in turn reduced, thus switching the mediator between its oxidized
and reduced states according to Eqs. (1)±(3). In these equations, Ered and Eox
represent the oxidoreductase in its reduced and oxidized state, MH the enzyme
Á
substrate, and M its oxidized state.
Ered ‡ oxidizer ! E
ꢁ1†
ꢁ2†
ꢁ3†
ox
Á
‡
Eox ‡ MH ! Ered ‡ M ‡ H
Á
‡
M ‡ target molecule ‡ H ! product ‡ MH
à Corresponding author

296
J. Kulys et al.
Phenols and aromatic amines are typical substrates for peroxidases and laccases
2, 3]. These enzymes catalyze the oxidation of single electron donors, e.g. phenyl-
[
enediamines, phenothiazines, and phenoxazines [4, 5]. Heterocyclic phenols, e.g.
aromatic compounds containing an N±OH group, have been suggested earlier as
mediators in peroxidase and laccase catalyzed dye and lignin bleaching as well as
lignin degradation [1]. The kinetics of the enzymatic oxidation of N-hydro-
xybenzotriazole (HBT) has been investigated [6], and the unexpectedly low
reactivity of HBT found has been attributed to the large redox potential of this
substrate. Derivatives of N-aryl hydroxamic acids, i.e. compounds containing an R±
CO±N(OH)±Ar group, represent another type of N±OH mediators, investigated only
very recently [7]. The aim of this work is to explore redox and mediator properties
of a new type of redox compounds, the N-aryl-N-hydroxy urethanes (AHUs) 1±5
[8]. The kinetics of the oxidation of 1±5 was investigated using a recombinant
Coprinus cinereus peroxidase (rCiP) with an emphasis on establishing structure-
activity relations by using ab initio quantum chemical computations and substrate
docking calculations.
Results and Discussion
The structures of the investigated AHUs 1±5 are depicted in Fig. 1. The redox
activity of compounds 1±5 was explored by cyclic voltammetry (CV). The CV
curves of 1±4 showed oxidation-reduction peaks which were almost symmetrical,
thus revealing the reversible character of the electrochemical conversion and the
rather high stability of the oxidized species (shown for 1 in Fig. 2). The calculated
redox potentials of 1±4 varied in the range between 0.33 and 0.35 V vs. SCE
(Table 1).
In contrast to 1±4, the cathodic peak of 5 was such smaller than the anodic peak
at potential scan rates of 50 and 100 mV/s (Fig. 3). A possible explanation for this
behaviour could be the low stability of the oxidized species of 5. Therefore, the
formal redox potential of 5 was calculated at a higher scan rate.
The absorption spectra of 1±4, which all contain an unsubstituted phenyl ring,
showed an absorbance at 240±250 nm in their UV/Vis spectra; the absorbance of
their oxidized products was found at 282 and 307 nm (Fig. 4). The differences of
the extinction coef®cients (Á") of oxidized and reduced compounds at 307 nm are
listed in Table 2.
The absorbance of the products showed signi®cant similarity with that of
nitrosobenzene (NB, Fig. 4). The formation of NB can be explained by single-
Fig. 1. Structures of N-aryl-N-hydroxy urethanes 1±5

Peroxidase-Catalyzed Oxidation of Urethanes
297
Fig. 2. Cyclic voltammograms of 1 at potential scan rates of 50, 100, and 200 mV Á s� 1 at pH ˆ 8:5
Table 1. Cyclic voltammetry of 1±5 in 0.05 M borate buffer solution (pH ˆ 8:5)
Scan rate/mV Á s�
1
E
pꢁa†
/mV
E
/mV
E1=2/mV
pꢁc†
1
1
1
2
2
2
3
3
3
4
4
4
5
5
5
50
100
200
50
370
378
372
366
370
372
350
366
374
370
378
380
414
422
428
312
312
306
304
304
298
290
290
284
322
322
322
±
341
345
339
335
337
335
320
328
329
346
350
351
±
100
200
50
100
200
50
100
200
50
100
200
368
370
395
399
electron oxidation of the respective substrate by the oxidized enzyme (Eq. (4))
followed by disproportionation of the formed radical (Eq. (5)) and reaction of the
intermediate with water (Eq. (6)).
Á
�
‡
Ph� NꢁOH†� CꢁO†OMe ! Ph� NꢁO †� CꢁO†OMe ‡ e ‡ H
ꢁ4†
Á
‡
2
Ph� NꢁO †� CꢁO†OMe ! Ph� NꢁOH†� CꢁO†OMe ‡ Ph� N ꢁO†� CꢁO†OMe
ꢁ
5†
‡
‡
�
Ph� N ꢁO†� CꢁO†OMe ‡ H O ! Ph� NO ‡ HOCꢁO†OMe ‡ H ‡ e
ꢁ6†
2
Compound 5 showed absorption maxima at 269 and 322 nm. The oxidation of this
substrate generated a product with a decreased absorbance at 269 nm and an
increased absorbance at 307 nm (Fig. 5).

298
J. Kulys et al.
Fig. 3. Cyclic voltammograms of 5 at potential scan rates of 50, 100, and 200 mV Á s� 1 at pH ˆ 8:5
Fig. 4. Absorbance of 2 at pH ˆ 8:5; the ®rst spectrum corresponds to 0.1 mM of 2, the others to the
mixture obtained by addition of 40 nM rCiP and 0.1 mM hydrogen peroxide (recorded at t ˆ 10, 20,
and 35 min); o: 50 mM nitrosobenzene
ꢀ
Table 2. Electrochemical and kinetic parameters of 1±5 at pH ˆ 8.5 and 25 C
E
ꢀ
max
Á"
‰rCiPŠ
kox
1=2
mM Á cm� 1
�
1
nM
M
� 1
Á s� 1
mV vs: SCE
nm
4
4
4
4
4
1
2
3
4
5
345
337
328
351
396
307
307
307
307
307
4.9
4.8
4.3
3.6
9.0
250
250
250
250
125
2.3Æ0.2 Á 10
7.5Æ0.5 Á 10
3.2Æ0.2 Á 10
4.2Æ0.2 Á 10
2.5Æ0.5 Á 10
The kinetics of the rCiP-catalyzed oxidation of 1±5 was monitored by the
absorbance increase at 307 nm. The initial rate increased linearly up to 200 mM of 1
and 3 (Fig. 6). Apparent oxidation constants (k ) were calculated as slope/[E] and
ox
are listed in Table 2. The initial oxidation rate of 2, 4, and 5 showed saturation

Peroxidase-Catalyzed Oxidation of Urethanes
299
Fig. 5. Absorbance spectrum of 5 at pH ˆ 8:5 after addition of 0.1 mM 5, 40 nM rCiP, and 0.1 mM
hydrogen peroxide; time between measurements: 10±20 min
Fig. 6. Dependence of oxidation rate on substrate concentration of 1 (o) and 3 (
50 nM, hydrogen peroxide: 0.1 mM
*
) at pH ˆ 8:5; rCiP:
2
behaviour (Fig. 7). Kinetic parameters for these substrates were calculated
according to the Michaelis-Menten equation. The apparent Km values were
3
36 mM for 2, 409 mM for 4, and 415 mM for 5.
The electrochemical investigations of 1±5 demonstrated that all compounds are
redox active. The alkyl substituents in the carbamic acid residue had only little
in¯uence on the redox potential (Table 2). The highest redox potential could be
observed for 5 which contains an electron-withdrawing cyano group in p-position.
The rate constants of the rCiP-catalyzed oxidation of 1±5 changed in the range of
4
4
� 1 � 1
2
predicts a correlation of log(k ) with the free energy of reaction [9]. However, no
.3 Á 10 to 7.5 Á 10 M s (Table 2). The Marcus theory of electron transfer
ox
satisfactory correlation could be found between log(k ) and the redox potential for
ox
all substrates.
In an attempt to explain the electrochemical behavior and enzymatic reactivity
of 1±5, ab initio calculations with respect to 1 and 4 were performed. The
calculations showed that the geometries of the reduced states of 1 and 4 are almost

300
J. Kulys et al.
Fig. 7. Dependence of oxidation rate on the concentration of 2 (*), 4 (&), and 5 (o) at pH ˆ 8:5;
rCiP: 250 (2, 4) and 125 nM (5), hydrogen peroxide: 0.1 mM
Fig. 8. Contribution of the atomic orbitals to the HOMOs of 1 and 4
planar; in the radical state, the structures became completely planar. The calculated
energies of the HOMOs were in the range of � 0.3054 au and � 0.3305 au.The
LUMOs changed in the range between of 0.0897 au and 0.1424 au. In accordance
with the molecular orbital coef®cients, the HOMO structures of 1 and 4 are mainly
derived from the N-aryl fragments (Fig. 8). The largest contributions result from N,
C6, and C9. The calculations showed a strong conjugation of the ꢁ-electrons of the
N-aryl and N±OH group; the carbonyl group contributed very little. The results
presented in Fig. 8 show that the phenyl substituent in ester groups, as e.g. in 4, does
not make an essential contribution to the HOMO. The calculations also showed that
during the oxidation the main rearrangements in geometry and electronic structure
take place in the nearest vicinity of the N±OH fragment. As a consequence, the

Peroxidase-Catalyzed Oxidation of Urethanes
301
Ê
N±O bond length becomes shorter by about 0.06±0.08 A and it is more similar to a
double bond.
The electrochemical oxidation of AHUs was associated with electron transfer
from the HOMO. Therefore, the values of redox potential were correlated with the
2
HOMO energy. The high value of the correlation coef®cient (R ˆ 0:9900)
demonstrated good coincidence between experimentally determined and calculated
parameters. In contrast, no signi®cant correlation was found between log(k ) and
ox
the HOMO energy. This fact, in combination with the bad correlation of log(kox)
with the redox potential, permits to hypothesize that the enzymatic reactivity is
determined by speci®c docking of the substrates in the active center of rCiP.
The docking of benzhydroxamic acid (BHA) in the active center of Arthromyces
ramosus peroxidase (ARP) has been investigated recently [10]. A crystal structure
of ARP and rCiP has been found to be similar [10, 11]. Therefore, a possible
docking of AHUs in the active center of rCiP was analyzed analogously to the
docking in the active center of ARP. To test whether a program can correctly predict
the experimentally determined binding mode, we ®rst investigated the binding mode
for BHA. The calculated docking completely ®tted the ARP and BHA structure if the
X-ray structure of ARP in the complex with benzhydroxamic acid was used [10].
The docking of 4 to both 1ARP (ARP structure) and 1HSR (structure in the
complex ARP and benzhydroxamic acid) produced three clusters near Pro91 and
Ile153 (Fig. 9). The mean docking energy of these clusters was � 30.2, � 28.9, and
�
1
�
27.1 kJ Á mol (Table 3). An additional cluster was found to be located at the
�
1
entrance to the heme pocket; its mean docking energy was � 27.90 kJ Á mol .
Docking of 4 to the 1HSR structure resulted in the formation of analogous clusters;
however, the docking energy was larger. In both structures the clusters are located at
the distal side of heme at the entrance to the active center. The N-aryl fragment in
the 4/1ARP complex was spaced by about 9.2 A from the water molecule, the O-
aryl fragment being closer to the heme than the N-aryl fragment. The distance of the
Ê
meta-carbon atom in the O-aryl fragment was 3.73 A from the heme and 4.72 A
Ê
Ê
Fig. 9. Docking of 4 in the ARP structure

302
J. Kulys et al.
Table 3. Docking energy of N-aryl-N-hydroxy urethanes in the active center of ARP
E1ARP
kJ Á mol�
E1HSR
kJ Á mol� 1
Location
Location
1
4
4
4
4
1
1
� 30.25
� 28.95
� 27.07
� 27.90
Pro91, Ile153
Pro91, Ile153
Pro91, Ile153
Entrance
� 29.54
� 28.99
� 27.70
� 26.90
Pro91, Ile153
Pro91, Ile153
Pro91, Ile153
Entrance
� 21.87 to � 26.77
� 22.25
Pro91, Ile153
� 21.83 to � 25.39 Pro91, Ile153
Asp98, Thr99, Lys49 � 22.08
Asp98, Thr99, Lys49
from the water molecule in the 4/1HSR complex. For 1ARP, the respective distances
Ê
were 3.73 and 3.71 A.
The calculations of the docking of 1 in 1ARP and 1HSR structures revealed that
about 95% of the substrate combines in a small pocket at Pro91 and Ile153. The
�
1
mean docking energy of these clusters changes from � 21.87 to � 26.77 kJ Á mol or
�
1
from � 21.83 to � 25.39 kJ Á mol (Table 3). A small amount of the substrate can
also combine at the residues of Asp98, Thr99, and Lys49. In this complex, the N-
Ê
aryl fragment is located at a distance of 14 A from the heme and the water molecule.
The docking results show that N-aryl fragments of 1 and 4 are located far away
from the heme. Since the HOMO of the AHUs is located mostly on this fragment,
the low reactivity of the substrates may be explained by the large electron transfer
distance. Since the redox potential of the substrates changes little, their docking in
the active center determines the electron transfer rate.
Experimental
N-Hydroxy-N-phenyl carbamic acid methyl ester (1) [12], N-hydroxy-N-phenyl carbamic acid ethyl
ester (2) [13, 14], N-hydroxy-N-phenyl carbamic acid 2-propyl ester (3) [15, 16], and N-hydroxy-N-
phenyl carbamic acid phenyl ester (4) [17] were all synthesized according to literature procedures.
1
13
They were identi®ed by H NMR and C NMR spectroscopy (Varian Mercury 400) and by their
melting points. Elemental analyses (C, H, N) were determined by PD Analytical Laboratories (Novo
Nordisk A/S, M aÊ lùv); they agreed satisfactorily with the calculated values. Melting points were
determined on a B uÈ chi 510 apparatus and are uncorrected. All chemicals used were obtained from
commercial sources (Aldrich, Fluka, Avocado). Thin-layer chromatography was performed using
Alugram SIL G/UV254 TLC plates from Macherey-Nagel.
N-Hydroxy-N-(4-cyanophenyl)-carbamic acid methyl ester (5; C H N O )
9 8 2 3
3
A solution of 3.13 g N-4-(cyanophenyl)-hydroxylamine (21.1 mmol) in 30 cm THF was synthesized
following the procedure described for the preparation of N-4-phenylhydroxylamine [7, 18]. A slurry
3
ꢀ
of 4.6 g NaHCO in 7 cm H O was added to the solution. The mixture was cooled to � 0 C, and
3
2
3
.43 cm methyl chloroformate (21.1 mmol) were added dropwise. Stirring was continued overnight;
1
3
then, 30 cm of 8% NaOH were added over a period of 45 min. The aqueous phase was separated,
3
petroleum ether (60 cm ) was added to the THF-phase, and the aqueous phase was separated again.
3
The organic phase was extracted twice with 40 cm 8% NaOH. The combined aqueous phases were
washed with CH Cl2 and neutralized with HCl under cooling. The acidi®ed H O-phase was
2
2
3
extracted 3 times with 60 cm CH Cl . The combined organic phases were dried over MgSO ,
2
2
4
®ltered, and the solvent was removed in vacuum to give crude 5. The compound was crystallized
from toluene/heptane to afford 2.69 g of white crystals (66%).

Peroxidase-Catalyzed Oxidation of Urethanes
303
ꢀ
1
Mp.: 118±121 C; H NMR (400 MHz, ꢂ, CDCl ): 7.71 (d, 2H), 7.63 (d, 2H), 3.93 (s, 3H) ppm;
3
1
3
C NMR (100 MHz, ꢂ, CDCl ): 154.16, 143.89, 132.52, 119.01, 118.54, 107.17, 54.22 ppm.
3
Recombinant fungal peroxidase from Coprinus cinereus (rCiP) was a highly puri®ed product of
Novo Nordisk A/S, Denmark; its Reinheit-Zahl (A405=A280) was 2.61. The concentration of rCiP was
5
� 1
� 1
determined spectrophotometrically at 405 nm assuming a molar absorbance of 1.08 Á 10 M Á cm
[
19]. Solutions of H O were prepared from perhydrol (30%), and concentrations were established
2
2
�
1
� 1
using a molar absorbance of 39.4 M Á cm at 240 nm [20].
The steady-state kinetic measurements of the oxidations were performed by a spectrophotometric
method. Kinetic curves and absorption spectra were registered using a computer-assisted
spectrophotometer (Gilford Instrument model 2600, Gilford Instrument Lab., USA). The substrates
were dissolved in methanol with the exception of 1 which was dissolved in buffer solution. The
amount of methanol in the ®nal solution was 0.5% (v/v). The reaction was initiated by addition of
ꢀ
H O . The measurements were carried out at 25Æ0.1 C in 0.05 M borate buffer (pH ˆ 8:5). The
2
2
kinetic data (absorbance change) were approximated by a third order polynomial function; the
relative error of approximation did not exceed 0.2%. The calculated initial steady-state rate was used
for further analysis. If the initial rate was a linear function of the substrate concentration, an apparent
bimolecular rate constant (kox) was calculated from its slope. If the rate showed a saturating
character, an apparent Michaelis constant (Km) and a catalytic constant (kcat) was calculated
according to the Michaelis-Menten equation. In this case an apparent bimolecular rate constant (kox)
was de®ned as the ratio of kcat/Km. All data are presented as mean values Æ standard deviation.
The cyclic voltammograms were recorded in a three-electrode circuit using a computer-
controlled electroanalytical system (Cypress Systems Inc., USA) and a glassy carbon electrode
(
model Cs-1087, Cypress Systems Inc., USA). A saturated calomel electrode (SCE, saturated with
KCl, mod. K-401, Radiometer, Denmark) was used as a reference. As an auxiliary electrode, a Pt-
wire (diameter 0.2 mm, length 4 cm) mounted on the end of the reference electrode was used. The
electrode potential was varied from 200 to 800 mV, the potential scan rate was 100 mV/s. The formal
redox potential (E1=2) was estimated by using the relationship E1=2 ˆ ꢁEpꢁa† ‡ Epꢁc††=2, where Epꢁa†
and E
are the anodic and cathodic peak potentials, respectively.
pꢁc†
Ab initio calculations of electronic structures and the energies of 1±5 were performed using the
Gaussian 94 W package [21]. The optimization of geometry was accomplished using HF (Hartree-Fock)
theory and the 3-21 G basis set. The optimized geometry of molecules was used for the further ab initio
HF calculations with the 6-31 G basis set. The geometry of the radicals was optimized under the ROHF
theory and 3-21 G, the energetic parameters were calculated employing the ROHF/6-31 G model.
The simulations of substrate docking in the active center of Arthromyces ramosus peroxidase
(
ARP) were performed with AutoDock 3.0 [22±24]. The crystal data of ARP (1ARP) and the ARP/
benzhydroxamic acid (1HSR) complex [25, 26] were downloaded from the Protein Data Bank. All
water molecules in the active center of ARP were removed with exception of the water molecule at
the distal side of heme.
Ê
The energy grid maps of atomic interaction were calculated with a 0.25A grid spacing and 120 grid
Ê
points forming a 30 A cubic box centered at the active site of peroxidases on the heme side exposed to
water. The space of the cubic box covered the active site of peroxidases and the space beyond. The
electrostatic interaction energy grid used a distance-dependent dielectric function of Mehler-Solmajer
[
27]. The docking was accomplished using the Lamarckian genetic algorithm. The number of
individuals in populations was set to 50. The maximum number of energy evaluations of this algorithm
is 250000; the maximum number performed was 27000. The number of top individuals guaranteed to
survive into the next generation was 1, the crossover and mutation rates were 0.02 and 0.80.
Acknowledgments
The authors express sincere thanks to Dr. A. J. Olson for the supply of AutoDock 3.0. We also wish
to thank M. Metzler and C. Ludvigsen for technical assistance.

304
J. Kulys et al.: Peroxidase-Catalyzed Oxidation of Urethanes
References
[
[
[
[
[
[
[
1] Call HP, M uÈ cke I (1997) J Biotechnol 53: 163
2] Dunford HB, Adeniran AJ (1986) Arch Biochem Biophys 251: 536
3] Hughes MF, Smith BJ, Eling TE (1992) Chem Res Toxicol 5: 340
4] Krikstopaitis K, Kulys J, Pedersen AH, Schneider P (1998) Acta Chem Scand 52: 469
5] Kulys J, Krikstopaitis K, Ziemys A (2000) J Biol Inorg Chem 5: 333
6] Krikstopaitis K, Kulys J, Palaima A (1996) Biologija 4: 33
7] Kulys J, Deussen H-J, Krikstopaitis K, Lolck R, Schneider P, Ziemys A (2000) J Chem Soc Perk
T2 (in press)
[
8] Schneider P, Deussen H-J (1999) Patent application WO9957360
9] Marcus RA, Sutin N (1985) Biochim Biophys Acta 811: 265
[
[
10] Tsukamoto K, Itakura H, Sato K, Fukuyama K, Miura S, Takahashi S, Ikezawa H, Hosoya T
1999) Biochemistry 381: 2558
(
[
11] Petersen JFW, Kadziola A, Larsen S (1994) FEBS Lett 339: 291
12] Baskakov YA, Stonov LD, Bakumenko LA, Mikhno LA, Gorskaya TV, Pokrovskaya NB (1971)
Patent GB 71-7819 19710325, CAN 78: 159291
[
[
[
[
[
[
[
[
13] Bamberger E (1919) Chem Ber 52: 1120
14] Tisue GT, Grassmann M, Lwowski W (1968) Tetrahedron 24: 999
15] Steinbrunn G, Fischer A (1967) BASF-AG Patent FR 1507608; Chem Abstr (1969) EN, 70, 57386e
16] Baskakow YA, Mel'nikov NN (1958) Khim Nauka I Prom 3: 683; Chem Abstr (1959) 7062
17] Oesper RE, Broker W (1925) J Am Chem Soc 47: 2606
18] Oxley PW, Adger BM, Sasse MJ, Forth MA (1989) Org Synth 67: 187
19] Andersen MB, Hsuanyu Y, Welinder KG, Schneider P, Dunford HB (1991) Acta Chem Scand 45:
1080
[
[
20] Nelson D, Kiesow LA (1972) Anal Biochem 49: 474
21] Frisch MJ, Trucks GW, Schlegel HB, Gill PMW, Johnson BG, Robb MA, Cheeseman JR, Keith
T, Petersson GA, Montgomery JA, Raghavachari K, Al-Laham MA, Zakrzewski VG, Ortiz JV,
Foresman JB, Cioslowski J, Stefanov BB, Nanayakkara A, Challacombe M, Peng CY, Ayala PY,
Chen W, Wong MW, Andres JL, Replogle ES, Gomperts R, Martin RL, Fox DJ, Binkley JS,
Defrees DJ, Baker J, Stewart JP, Head-Gordon M, Gonzalez C, Pople JA (1995) Gaussian 94,
Revision E. 1, Gaussian, Inc., Pittsburgh PA
[
[
[
22] Goodsell DS, Olson AJ (1990) Prot Struct Gen 8: 195
23] Morris GM, Goodsell DS, Huey R, Olson AJ (1996) J Com Aid Mol Des 10: 293
24] Morris GM, Goodsell DS, Halliday RS, Huey R, Hart WE, Belew RK, Olson AJ (1998) J Comp
Chem 19: 1639
[
25] Kunishima N, Fukuyama K, Matsubara H, Hatanaka H, Shibano Y, Amachi T (1994) J Mol Biol
235: 331
[
26] Itakura H, Oda Y, Fukuyama K (1997) FEBS Let 412: 107
27] Mehler EL, Solmajer T (1991) Protein Eng 4: 903
[
Received June 26, 2000. Accepted (revised) August 9, 2000
È
Verleger: Springer-Verlag KG, Sachsenplatz 4±6, A-1201 Wien. ± Herausgeber: Osterreichische Akademie der Wissenschaften,
È
Dr.-Ignaz-Seipel-Platz 2, A-1010 Wien, und Gesellschaft Osterreichischer Chemiker, Eschenbachgasse 9, A-1010 Wien. ±
Redaktion: W aÈ hringer Straꢀe 38, A-1090 Wien. ± Satz und Umbruch: Thomson Press Ltd., New Delhi, India. ± Offsetdruck:
MANZ CROSSMEDIA, A-1051 Wien. ± Verlagsort: Wien. ± Herstellungsort: Wien. ± Printed in Austria.