Probing the specificity of the serine proteases subtilisin Carlsberg and α-chymotrypsin with enantiomeric 1-acetamido boronic acids. An unexpected reversal of the normal L -stereoselectivity preference

Article abstract of DOI:10.1021/ja952816j

Enantiomeric 1-acetamido boronic acids, which are N-acetyl transition state analog inhibitor analogs of L-and D-forms of the amino acids alanine, phenylalanine, p-fluorophenylalanine, p-chlorophenylalanine, and 1-naphthylalanine, have been evaluated as inhibitors of the serine proteases subtilisin Carlsberg (SC) and α-chymotrypsin (CT). All of the boronic acids are powerful competitive inhibitors of both enzymes, with, as expected, the L-enantiomers being generally more potent than the D-enantiomers. However, a dramatic reversal of the normal stereoselectivity preference was observed in the inhibition of CT by [1-acetamido-2-(1-naphthyl)ethyl]boronic acid, with the D-enantiomer becoming a 25-fold more potent inhibitor than the L-enantiomer. Furthermore, the K₁ of 127 nM for CT inhibition by this D-enantiomer is the lowest of any of the boronic acids evaluated. Molecular modeling analyses of the possible binding modes of the inhibitors suggest that the stereoselectivity reversal is due to S₁-pocket orientations of naphthyl groups that are different from those of the aromatic side chains of the phenylalanine analogs.

Full text of DOI:10.1021/ja952816j

950
J. Am. Chem. Soc. 1996, 118, 950-958
Probing the Specificity of the Serine Proteases Subtilisin
Carlsberg and R-Chymotrypsin with Enantiomeric 1-Acetamido
Boronic Acids. An Unexpected Reversal of the Normal
“L”-Stereoselectivity Preference
Valeri Martichonok and J. Bryan Jones*
Contribution from the Lash Miller Laboratories, Department of Chemistry, UniVersity of Toronto,
80 St. George Street, Toronto, Canada M5S 1A1
ReceiVed August 15, 1995X
Abstract: Enantiomeric 1-acetamido boronic acids, which are N-acetyl transition state analog inhibitor analogs of L-
and D-forms of the amino acids alanine, phenylalanine, p-fluorophenylalanine, p-chlorophenylalanine, and
1-naphthylalanine, have been evaluated as inhibitors of the serine proteases subtilisin Carlsberg (SC) and
R-chymotrypsin (CT). All of the boronic acids are powerful competitive inhibitors of both enzymes, with, as expected,
the L-enantiomers being generally more potent than the D-enantiomers. However, a dramatic reversal of the normal
stereoselectivity preference was observed in the inhibition of CT by [1-acetamido-2-(1-naphthyl)ethyl]boronic acid,
with the D-enantiomer becoming a 25-fold more potent inhibitor than the L-enantiomer. Furthermore, the KI of 127
nM for CT inhibition by this D-enantiomer is the lowest of any of the boronic acids evaluated. Molecular modeling
analyses of the possible binding modes of the inhibitors suggest that the stereoselectivity reversal is due to S1-pocket
orientations of naphthyl groups that are different from those of the aromatic side chains of the phenylalanine analogs.
Enzymes are now widely used in synthetic organic chemistry,
able.4,5 In both SC and CT, the active site binding regions are
composed of several subsites, of which the S1 -pocket domi-
6
with their abilities to be highly stereoselective in their catalyses
being extensively exploited in asymmetric synthesis.1 However,
the factors responsible for determining the structural and
stereospecificity of enzymes toward unnatural substrates and
inhibitors remain poorly understood. We became interested in
this area2 because, in order to identify the enzymes best suited
to coping with the increasingly broad chiral synthon demands
of asymmetric synthesis, it is important that the factors
controlling substrate binding and orientation be identified.
With synthetic applications of hydrolytic enzymes being of
such widespread current interest,1 the serine proteases subtilisin
Carlsberg (SC; EC 3.4.21.14) and R-chymotrypsin (CT; EC
3.4.21.1) were selected as representative esterases for ste-
reospecificity studies. SC and CT are commercially available
enzymes that have been extensively applied synthetically3 and
for which high-resolution X-ray crystal structures are avail-
nates, particularly in the binding of hydrophobic groups. While
in their catalyses of hydrolyses of their natural protein substrates,
and of related amino acid esters, both enzymes exhibit a
dominant stereoselectivity preference for the L-amino acid
configuration, forecasting their stereoselectivities for unnatural
substrates is not straightforward. Even for amino acid esters,
reversals of stereoselectivity, i.e., to prefer D over L,7,8 can be
induced within a homologous series.9 Furthermore, the substrate
specificity of SC and CT can be modified when water is replaced
with a nonaqueous solvent,10 and some inversion of enzyme
enantioselectivity can be induced by switching solvents.11 Also,
in supercritical fluids the degree of enantioselectivity can be
tuned by changing the pressure.12
(4) (a) McPhalen, C. A.; James, M. N. G. Biochemistry 1988, 27, 6582.
(b) Bode, W.; Papamokos, E.; Musil, D. Eur. J. Biochem. 1987, 166, 673.
(5) (a) Frigerio, F.; Coda, A.; Pugliese, L.; Lionetti, C.; Menegatti, E.;
Amiconi, G.; Schnebli, H. P.; Ascenzi, P.; Bolognesi, M. J. Mol. Biol. 1992,
225, 107. (b) Tsukada, H.; Blow, D. M. M. J. Mol. Biol. 1985, 184, 703.
(c) Birktoft, J. J.; Blow, D. M. M. J. Mol. Biol. 1972, 68, 187. (d) Blevins,
R. A.; Tulinsky, A. J. Biol. Chem. 1985, 260, 4264. (e)Tulinsky, A.; Blevins,
R. A. J. Biol. Chem. 1987, 262, 7737.
X Abstract published in AdVance ACS Abstracts, January 1, 1996.
(1) (a) Enzymes in Synthetic Organic Chemistry; Wong, C.-H., White-
sides, G. M., Eds.; Pergamon: New York, 1994. (b) PreparatiVe Biotrans-
formations; Roberts, S. M., Ed.; Wiley: New York, 1993. (c) Biotrans-
formations in PreparatiVe Organic Chemistry; Faber, K., Ed.; Springer-
Verlag: Heidelberg, 1992.
(2) (a) Jones, J. B. Aldrichim. Acta 1993, 26, 105. (b) Bonneau, P. R.;
Eyer, M.; Graycar, T. P.; Estell, D. A.; Jones, J. B. Bioorg. Chem. 1993,
21, 431. (c) Jones, J. B. Can. J. Chem. 1993, 71, 1273. (d) Bonneau, P. R.;
Graycar, T. P.; Estell, D. A.; Jones, J. B. J. Am. Chem. Soc. 1991, 113,
1026. (e) Lee, T.; Jones, J. B. Tetrahedron 1995, 51, 7331.
(6) (a) Schechter, I.; Berger, A. Biochem. Biophys. Res. Commun. 1967,
27, 157. (b) Berger, A.; Schechter, I. Philos. Trans. R. Soc. London, Ser. B
1970, 257, 249.
(7) It should be noted that for the first apparent example of a CT
preference of a D- over an L-substrate,8 the D-center is actually of the
L-amino acid configuration type.
(8) Hein, G. E.; McGriff, R. B.; Niemann, C. J. Am. Chem. Soc. 1960,
82, 1830. Hein, G. E.; Niemann, C. J. Am. Chem. Soc., 1962, 84, 4487.
(See also: Jones, J. B., Beck, J. F. Tech. Chem. (N. Y.), 1976, 10, 172-
185).
(9) Schwartz, H. M.; Wu, W. -S.; Marr, P. W. Jones, J. B. J. Am. Chem.
Soc. 1978, 100, 5199.
(10) Wescott, C. R.; Klibanov, A. M. J. Am. Chem. Soc. 1993, 115, 1629.
(11) (a) Tawaki, S.; Klibanov, A. M. J. Am. Chem. Soc. 1992, 114, 1882.
(b) Orsat, B.; Drtina, G. J.; Williams, M. G.; Klibanov, A. Biotechnol.
Bioeng. 1994, 44, 1265. (c) Wescott, C. R.; Klibanov, A. M. Biochim.
Biophys. Acta 1994, 1206, 1. (d) Wescott, C. R.; Klibanov, A. M. J. Am.
Chem. Soc. 1993, 115, 10362. (e) Yennawar, H. P.; Yennawar, N. H.; Farber,
G. K. J. Am. Chem. Soc. 1995, 117, 577.
(3) (a) Delinck, D. L.; Margolin, A. L. Tetrahedron Lett. 1990, 31, 3093.
(b) Pugniere, M.; San Juan, C.; Previero, A. Tetrahedron Lett. 1990, 31,
4883. (c) Gotor, V.; Garcia, M. J.; Rebelleo, F. Tetrahedron: Asymmetry
1990, 1, 277. (d) Margolin, A. L.; Delinck, D. L.; Whalon, M. R. J. Am.
Chem. Soc. 1990. 112, 2849. (e) Chenevert, R.; Desjardins, M.; Gagnon,
R. Chem. Lett. 1990, 33. (f) Chenevert, R.; Letourneau, M.; Thiboutot, S.
Can. J. Chem. 1990, 960. (h) Frigerio, F.; Coda, A.; Pugliese, L.; Lionetti,
C.; Menegatti, E.; Amiconi, G.; Schnebli, H. P.; Ascenzi, P.; Bolognesi,
M. J. Mol. Biol. 1992, 225, 107. (i) Ricca, J. M.; Crout, D. H. J. Chem.
Soc., Perkin Trans. 1 1989, 2126. (j) Kitaguchi, H.; Fitzpatrick, P. A.; Huber,
J. E.; Klibanov, A. M. J. Am. Chem. Soc. 1989, 111, 3094. (k) Brieva, R.;
Rebelleo, F.; Gotor, V. J. Chem. Soc., Chem. Commun. 1990, 1386. (l)
Gutman, A. L.; Meyer, E.; Kalerin, E.; Polyak, F.; Sterling, J. Biotechnol.
Bioeng. 1992, 40, 760. (m) Roper, J. M.; Bauer, D. P. Synthesis 1983, 1041.
0002-7863/96/1518-0950$12.00/0 © 1996 American Chemical Society

Specificity of Subtilisin Carlsberg and R-Chymotrypsin
J. Am. Chem. Soc., Vol. 118, No. 5, 1996 951
Scheme 1
In our previous probing of enzyme specificity,13 the strategy
of evaluating the binding affinities of boronic acid transition
state analog competitive inhibitors14 was followed, coupled with
graphics analyses and molecular modeling. Boronic acids are
generally very effective, reversible, transition state inhibitors
of serine proteases14,15 and have proved well suited for the
systematic probing of the structural and electrostatic specificity
of the S1-site of SC for achiral boronic acid inhibitors.13
Accordingly, the same strategy has been adopted for probing
stereoselectivity determinants, using enantiomeric 1-acetamido
boronic acids 1a-e, whose structures mimic known N-acetyl-
amino acid ester substrates, as transition state analog inhibitors
of SC and CT. Both L- and D-enantiomers were included since,
while N-acetyl-D-amino acid esters are not serine protease
substrates, they are able to bind at the same active site locations
as their L-substrate counterparts and can be effective competitive
inhibitors.16 The achiral parent boronic acids 2a-e were also
included for reference purposes.
hexamethyldisilazane afforded the corresponding silylated amino
boronic esters, which were unstable and were treated directly
with acetic acid and acetic anhydride at -78 °C according to
the Matteson protocol.19 The formation of the 1-acetamido
boronic esters 5a-e occurred with complete inversion. Pi-
nanediol esters are very resistant to the hydrolysis, and the 5
f 1 conversions required cleavage with boron trichloride at
-78 °C. Boronic acids are notoriously difficult to characterize
in terms of elemental composition because of the ease with
which they fully or partially dehydrate to the corresponding
trimeric, or oligomeric in the case of 1a, anhydrides. Accord-
ingly, all of the target inhibitors L- and D-1a-e were further
characterized as their stable, crystalline, diethanolamine deriva-
tives 6 by reaction with diethanolamine in 2-propanol.18b,d This
derivatization also provided protection against possible autoxi-
dation20 of the acetamido boronic acids 1 by atmospheric
oxygen. Both the anhydride forms of 1 and the diethanolamine
derivatives 6 are immediately and quantitatively hydrolyzed to
the corresponding free boronic acids on solution in water.
(15) (a) Westmark, P. R.; Kelly, J. P.; Smith, B. R. J. Am. Chem. Soc.
1993, 115, 3416: Baldwin, J. E.; Smith, B. D.; Claridge, T.; Derome, A.;
Schofield, C.J. Bioorg. Med. Chem. Lett. 1991, 1, 9. (b) Elgendy, S.;
Deadman, J.; Patel, G.; Green, D.; Chino, N.; Goodwin, C. A.; Scully, M.
F.; Kakkar, V. V.; Claeson, G. Tetrahedron Lett. 1992, 33, 4209. (c) Kettner,
C.; Mersinger, L.; Knaubb, R. J. Biol. Chem. 1990, 265, 18289. (d)
Abouakil, N.; Lombardo, D. Biochim. Biophys. Acta 1989, 1004, 215. (e)
Crompton, I. E.; Cuthbert, B. K.; Lowe, C.; Waley, S. G. Biochem. J. 1988,
251, 453. (f) Goz, B.; Ganguli, C.; Troconis, M.; Wyrick, S.; Isahq, K. S.;
Katzenellenbogen, J. A. Biochem. Pharmacol. 1986, 35, 3587. (g) Kinder,
D. H.; Katzenellenbogen, J. A. J. Med. Chem. 1985, 28, 1917. (h) Kettner,
C. A.; Shenvi, A. B. J. Biol. Chem. 1984, 259, 15106. (i) Tsai, I. H.; Bender,
M. L. Arch. Biochem. Biophys. 1984, 228, 555. (j) Matteson, D. S.; Sadhu,
K. M.; Lienhard, G. F. J. Am. Chem. Soc. 1981, 103, 5241. (k) Garner, C.
W. J. Biol. Chem. 1980, 255, 5064. (l) Nakatani, H.; Uehara, Y; Hiromi,
K. J. Biochem. 1975, 78, 611.
(16) (a) Foster, R. J.; Niemann, C. J. Am. Chem. Soc. 1955, 77, 1886.
(b) Foster, R. J.; Shine, H. J.; Niemann, C. J. Am. Chem. Soc. 1955, 77,
2378. (c) Manning, D. T.; Niemann, C. J. Am. Chem. Soc. 1958, 80, 1478.
(17) (a) Matteson, D. S.; Sadhu, K. M. Organometallics 1984, 3, 614.
(b) Matteson, D. S.; Michnick, T. J.; Willett, R. D.; Patterson, C. D.
Organometallics 1989, 8, 726. (c) Matteson, D. S.; Michnick, T. J. J.
Labelled Compd. Radiopharm. 1992, 31, 567.
(18) (a) Matteson, D. S.; Majumdar, D. Organometallics 1983, 2, 1529.
(b) Tripathy, P. B.; Matteson, D. S. Synthesis 1990, 200. (c) Matteson, D.
S.; Sadhu, K. M.; Peterson, M. L. J. Am. Chem. Soc. 1986, 108, 810. (d)
Matteson, D. S.; Ray, R.; Rocks, R. R.; Tsai, D. J. S. Organometallics
1983, 2, 1536. (e) Matteson, D. S.; Majumdar, D. J. Am. Chem. Soc. 1980,
102, 7588. (f) Matteson, D. S.; Ray, R. J. Am. Chem. Soc. 1980, 102, 7590.
(g) Matteson, D. S.; Erdik, E. Organometallics 1983, 2, 1083. (h) Matteson,
D. S.; Sadhu, K. M. J. Am. Chem. Soc. 1983, 105, 2077. (i) Matteson, D.
S.; Beedle, E. C. Tetrahedron Lett. 1987, 28, 4499.
(19) Matteson, D. S.; Jesthi, P. K.; Sadhu, K. M. Organometallics 1984,
3, 1284.
(20) Korcek, S.; Watts, G. B.; Ingold, K. U. J. Chem. Soc., Perkin Trans.
2 1972, 242.
Results and Discussion
Preparation of Boronic Acids. The L-(R)- and D-(S)-1-
acetamido boronic acids used for this study were prepared, as
shown in Scheme 1 for the L-(R)-series, employing the basic
strategy developed by Matteson et al.15h,17 Matteson homolo-
gation18 of the pinanediol esters 3a-e gave the R-chloro boronic
esters 4a-e in 75-95% yields with diastereoselectivities >98%.
Treatment of the R-chloro boronic esters 4 with lithium
(12) (Kamat, S. V.; Beckman, E. J.; Russell, A. J. J. Am. Chem. Soc.
1993, 115, 8845.
(13) (a) Keller, T. H.; Seufer-Wasserthal, P.; Jones, J. B. Biochem.
Biophys. Res. Commun. 1991, 176, 401. (b) Seufer-Wasserthal, P.; Marti-
chonok, V.; Keller, T. H.; Chin, B.; Martin, R.; Jones, J. B. Bioorg. Med.
Chem. 1994, 2, 35.
(14) (a) Koehler, K. A.; Lienhard, G. E. Biochemistry 1971, 10, 2477.
(b) Rawn, J. D.; Lienhard, G. E. Biochemistry 1974, 13, 3124. (c) Phillip,
M.; Bender, M. L. Proc. Natl. Acad. Sci. U.S.A. 1971, 68, 478. (d) Antonov,
V. K.; Ivanina, T. V.; Berezin, I. V.; Martinek K. FEBS Lett. 1970, 7, 23.
(e) Wolfenden, R. Acc. Chem. Res. 1972, 5, 10.

952 J. Am. Chem. Soc., Vol. 118, No. 5, 1996
Martichonok and Jones
Table 1. Inhibition of Subtilisin Carlsberg and R-Chymotrypsin by Acetamido Boronic Acids 1a-ea
inhibition constant KI (µM)
inhibitor
R
subtilisin Carlsberg
R-chymotrypsin
OH
B
R
OH
NHAc
L-1a
H
78.5 ( 4.1
no inhibitionb
L-1b
L-1c
L-1d
L-1e
1.22 ( 0.03
3.80 ( 0.13c
0.28 ( 0.02
0.15 ( 0.01
0.88 ( 0.02
1.24 ( 0.06
1.20 ( 0.05
3.11 ( 0.09.
F
Cl
D-1a
D-1b
H
9300 ( 350
126.2 ( 5.2
no inhibitionb
79.9 ( 3.3d
D-1c
D-1d
D-1e
87.3 ( 3.2
91.3 ( 5.4
56.4 ( 3.0
46.3 ( 1.6
F
5.76 ( 0.23
0.127 ( 0.03
Cl
OH
B
R
OH
2a
2b
H
13000e
257e
not available
481f
2c
2d
2e
48e
19e
142f
52f
F
Cl
110e
18.8f
a KI values for both enzymes were determined in duplicate at pH 7.8 in 0.1 M KCl and at 25 °C. Initial rates for SC were measured at substrate
(TAME) concentrations in the range of 0.075-0.15 M, inhibitor concentrations of 10-7 to 5.0 × 10-2 M, and an enzyme concentration of 2.0 ×
10-7 M. Initial rates for CT were measured at substrate (NATEE) concentrations in the range of 1.5 × 10-4 to 3.0 × 10-3 M, inhibitor concentrations
of 10-7 to 10-1 M and an enzyme concentration of 2.0 × 10-8 M. b No inhibition was observed at the concentration of the inhibitor, 5.0 × 10-2
M. c The reported value15h is 2.1 µM. It was determined for the R-chymotrypsin-catalyzed hydrolysis of methyl hippurate at pH 7.5 and 25 °C.
d The reported value15h is 53 µM. e From reference 13b. f From reference 15e.
Inhibition Studies. The individual inhibition constants for
each L- and D-acetamido boronic acid 1a-e for SC and CT
were determined using a pH-stat method and with N-p-tosyl-
L-arginine methyl ester (TAME) as the standard substrate.2d The
boronic acids were added to the assay mixtures as their
diethanolamine derivatives 6, these being readily hydrolyzed
by water under kinetic conditions to generate the corresponding
boronic acid in situ.15f The inhibitory activities of the boronic
acids generated by in situ generation from their diethanolamine
esters or anhydride forms were identical. Each boronic acid
was found to be a competitive inhibitor, and the results are
summarized in Table 1. For comparison purposes, and to
evaluate the effect of the N-acetamido group itself, literature
values13b,15e for inhibition of SC and CT by the unsubstituted
boronic acid parents of 1a-e are also included in Table 1.
The (1R)-1-acetamido boronic acids L-1a-e are much more
potent inhibitors of SC than the unsubstituted parent boronic
acids 2a-e, respectively. The largest increase of inhibitory
power attributable to the introduction of an R-configuration
N-acetamido group is exhibited by the phenethyl derivative L-1b,
with its KI being 210-fold lower than that of 2b. However,
even the smallest, N-acetamido-induced, binding enhancement
observed, that of 125-fold for the 1-naphthylethyl inhibitor L-1e
over that of 2e, is highly significant. Electronegative substit-
uents in the para-position contribute very positively to inhibitor
binding, as exemplified by the low KI values of [2-(4-
fluorophenyl)ethyl]- and [2-(4-chlorophenyl)ethyl]boronic acids
(L-1c,d) relative to that of the phenethyl inhibitor L-1b. This
is as expected from the additional electrostatic binding contribu-
tions resulting from the interactions of para-electronegative
substituents of this type with the region of positive potential
identified at the bottom of the S1-pocket of SC.13b In contrast,
for the enantiomeric compounds of the (S)-series D-1a,b, the
effect of the N-acetamido group is minimal, and the inhibitory
properties of these D-compounds are comparable to those of
their unsubstituted parents 2a-e. Some increases of binding
efficiency (1.4-2.0-fold) when the N-acetamido is present are
observed for the acetamido boronic acids D-1a, D-1b, and D-1e,

Specificity of Subtilisin Carlsberg and R-Chymotrypsin
J. Am. Chem. Soc., Vol. 118, No. 5, 1996 953
but inhibition by D-1c and D-1d is 1.8- and 4.8-fold worse than
for the unsubstituted analogs 2c and 2d, respectively. It is
evident that the configuration of the C-1 stereocenter is much
more important than the properties of the substituent inducing
the chirality.
The inhibition pattern for CT is quite different. Firstly,
neither enantiomer of (1-acetamidoethyl)boronic acid (1a)
inhibits CT, even at boronic acid concentrations of 50 mM.
However, the CT inhibition trends for the (R)-inhibitors L-1b-d
are similar to those for SC, with once more the (acetamidophen-
ethyl)boronic acid L-1b manifesting the largest increase (127-
fold) in binding enhancement relative to the parent, unsubsti-
tuted, boronic acid 2b. Furthermore, the naphthyl acetamido
compound L-1e, with its 6-fold reduction of KI over that of 2e,
again showed the lowest degree of substituent-induced aug-
mentation of inhibition. However, in contrast to the minimal
effects on SC inhibition of the acetamido substituent in the
enantiomeric D-(S)-series, all the acetamido boronic acids
D-1b-e are much better inhibitors of CT than their unsubstituted
parents 2b-e, respectively. Even the smallest (3-fold) decrease
in KI observed for the 4-fluorophenethyl compound D-1c relative
to 2c is greater than any such trend for SC. Furthermore, unlike
the SC situation, not of D-1b-e was a worse CT inhibitor than
2b-e, respectively.
The most dramatic augmentation of inhibitory power arising
from the introduction of the acetamido group is manifest in the
inhibition of CT by the (S)-acetamido naphthyl boronic acid
D-1e, whose KI of 0.127 µM is extraordinarily low, and
represents a remarkable 148-fold increase in binding efficiency
relative to that of its progenitor 2e. Most surprising was the
fact that the D-enantiomer of 1e was a 25-fold better inhibitor
of CT than its L-counterpart, which represents a totally
unexpected reversal of the high L-fidelity generally exhibited
by this enzyme. The stereoselectivity reversal was rendered
even more perplexing by the fact that, for SC, the normal
L-over-D preference for 1e-binding was retained by a large (64-
fold) margin, and that no stereoselectivity reversals were evident
for inhibitions of SC or CT by the other enantiomeric pairs of
boronic acids 1b-d. That the aromatic rings of 1b-d possessed
CV symmetry while the naphthyl group of 1e did not offered a
possible basis for rationalizing the stereoselectivity reversal. In
their EI complexes with SC and CT, the aromatic moieties of
both L- and D-1b-e will occupy the hydrophobic S1-pocket,
and the boronic acid OH’s and the acetamido group will be
directed toward the oxyanion hole and S2-pockets, respectively.
Because of their symmetry, the orientations of the phenyl rings
of 1b-d in S1 will not be affected by rotations about the C-2-
to-aromatic σ-bonds. In contrast, for 1e, rotations of the
naphthyl group about this bond will give rise to distinct
conformations, whose interactions with the individual S1-pockets
of SC and CT could induce different enzyme inhibitor (EI)
complexes with oppositely oriented naphthyl components, and
for which binding of the D-enantiomer could become preferred.
This concept is illustrated schematically in Figure 1. Further
enlightenment on the basis of the observed stereospecificity
reversal, and of the Table 1 binding trends, was sought using
molecular modeling.
Figure 1. As a consequence of rotation about the σ-bond shown (C),
different orientations can be envisaged for naphthyl groups in the S1-
pockets of serine proteases, such as subtilisin Carlsberg and R-chy-
motrypsin, on formation of the preferred EI complexes. The left-oriented
(a) and right-oriented (b) naphthyl conformations shown for the
schematic EI complex of L-4e represent two such possibilities.
boronic acid binding to serine proteases. Boronic acids have
been shown to be capable of forming tetrahedral complexes with
either the active site serine or histidine residues.21 Generally
the trend appears to be that good substrate analogs bind to serine
and poor substrate analogs to histidine, although binding of a
single inhibitor to give both a serine and a histidine adduct has
been reported.22 However, because there is so far no X-ray
structure of a serine protease-boronic acid complex involving
histidine to use as a guide, only serine-bound boronic acid
adducts, for which good X-ray data are available, were
considered in the current molecular modeling study. Another
consideration was our inability to overcome the problem posed
by the unavailability of force field parameters for covalent B-N
bonds.
Molecular graphics analyses in conjunction with molecular
mechanics, molecular dynamics, and electrostatic calculations
were applied in analyzing the Table 1 data. The high-resolution
X-ray structures of SC4a and CT5b were energy-minimized using
the Biosym “Discover” program and the boronic acid inhibitors
L-(R)- and D-(S)-1b-e and then individually docked into the
active site, using a CT-(phenylethyl)boronic acid X-ray
structure5e as a reference guide. Each EI complex was then
subjected to energy minimization by Discover’s molecular
mechanics protocol, followed by molecular dynamics simulation
for 20 ps. Analyses of the trajectories for SC showed that the
L- and D-acetamido boronic acids 1b-d oriented themselves at
the active site in the normal5e manner. For example, for L- and
D-1b, the oxygen atoms of the two hydroxyl groups attached to
the boron atom form hydrogen bonds to Nꢀ of His64 and to the
amide hydrogens of Asn155 and of the backbone NH of Ser221
of the oxyanion hole. The phenyl ring remains in the S1-pocket
after the molecular dynamics simulation for both enantiomers.
Also, the usual L-over-D stereoselectivity preference is main-
tained. The superiority of the L-enantiomer of 1b as an SC
inhibitor is due to the amide hydrogen of the 1-acetamido group
forming a strong hydrogen bond with the carbonyl oxygen of
Ser125 in the EI complex, whereas for its D-counterpart, this
amide NH is directed toward the solvent and does not contribute
to binding. Similarly, for the CT complexes with L- and D-1b,
the phenyl ring remains in the S1-pocket, and there are strong
interactions between the oxygens of the boronic acid with the
backbone NH’s of Gly193 and Ser195 of the oxyanion hole.
Furthermore, the L-preference is again due to favorable hydrogen
bonding, this time of the acetamido NH of L-1b with the
carbonyl oxygen of Ser214, whereas the acetamido group of
D-1b is once more oriented toward the solvent.
Molecular Modeling. Formulating appropriate molecular
modeling protocols toward interpreting the Table 1 data was
not straightforward because of the sometimes fickle nature of
(21) (a) Tsilikounas, E.; Kettner, C. A.; Bachovchin, W. W. Biochemistry
1993, 32, 12651. (b) Tsilikounas, E.; Kettner, C. A.; Bachovchin, W. W.
Biochemistry 1992, 31, 12839. (c) House, K. L.; Garber, A. R.; Dunlap, R.
B.; Odom, J. D.; Hilvert, D. Biochemistry 1993, 32, 3468. (d) London, R.
E.; Gabel, S. A. J. Am. Chem. Soc 1994, 116, 2570. (e) Snow, R. J.;
Bachovchin, W. W.; Barton, R. W.; Campbell, S. J.; Coutts, S. J.; Freeman,
D. M.; Gutheil, W. G.; Kelly, T. A.; Kennedy, C. A.; et al. J. Am. Chem.
Soc 1994, 116, 10860.
Most molecular modeling attention was direct toward inter-
preting the basis for the different stereoselectivities of SC and
CT toward the enantiomeric naphthyl boronic acids L- and D-1e,
focusing particularly on ascertaining if binding of the naphthyl
(22) Zhong, S.; Haghjoo, K.; Kettner, C.; Jordan, F. J. Am. Chem. Soc.
1995, 117, 7048.

954 J. Am. Chem. Soc., Vol. 118, No. 5, 1996
Martichonok and Jones
Figure 3. Superimposed energy-minimized EI complexes of D-(S)-1e
(dark -) and L-(R)-1e (dark ‚‚‚) respectively, in the active site of CT.
The naphthyl groups of both inhibitors fit very similarly into the S1-
pocket. The oxyanions of the tetrahedral complexes derived from D-
and L-1e are located in the oxyanion hole, with the negative charges
on the boron oxygens well stabilized by hydrogen-bonding (light ‚‚‚)
with the peptide NH’s of Ser195 and Gly193 for both inhibitors. The
excellence of the fits of the naphthyl groups into S1, and of the oxyanion
hole stabilizations, accounts for the good inhibitory properties of both
1e enantiomers. D-1e is the better inhibitor as a consequence of its
ability to form an additional hydrogen bond between the CO of its
acetamido group and the Nꢀ of His57. All calculated hydrogen-bonding
distances are given in the Experimental Section.
Figure 2. Superimposed energy-minimized EI complexes of L-(R)-1e
(dark -) and D-(S)-1e (dark ‚‚‚) respectively, in the active site of SC.
Both naphthyl residues bind in S1. The oxyanions of the tetrahedral
complexes derived from L-1e and D-1e are located in the oxyanion hole,
with the negative charges on the boron oxygens well stabilized by
hydrogen-bonding (light ‚‚‚) with the peptide NH of Ser221, the side
chain -NH2 of Asn155, and the Nꢀ of His64 for L-1e, but less well for
the weaker inhibitor D-1e. In addition, a strong hydrogen bond is
indicated between the backbone CO of Ser125 and the NH of the
acetamido group of L-1e, but not for D-1e. All calculated hydrogen-
bonding distances are given in the Experimental Section.
groups in different orientations of the Figure 1 type was an
important determinant. The dimensions of the S1-pockets of
both SC and CT are such that binding of the naphthyl
substituents of 1e in the two orientations shown in Figure 1 is
allowed. In the initial dockings of L- and D-1e with SC, the
left-orientation modes of Figure 1a directed the nonpolar C-5,
-6, -7, and -8 region of the naphthalene ring toward the solvent,
while with right-orientation binding (Figure 1b) these carbon
atoms pointed inside toward the hydrophobic center of the S1-
pocket. Molecular dynamics calculations on the SC complexes
confirmed the right orientation to be favorable, and showed that
left-orientation binding did not lead to stable EI complexes.
Furthermore, the key hydrogen bond for L-stereoselectivity
between the NH of the acetamido group of L-1e and the carbonyl
oxygen of Ser125 is only present in the minimized right-
orientation complex, and is absent for the analogous SC-D-1e
complex in which the acetamido group points toward the
solvent. The normal anion hole stabilization provided by
Asn155 is not possible, and the tetrahedral intermediate anion
is now hydrogen bonded to Ser221. The superimposed L- and
D-1e complexes with SC are shown in Figure 2.
In contrast, binding of L- and D-1e into the active site of CT
shows that the left orientation of the naphthyl group is now the
more favorable. In this case, the naphthyl C-5, -6, -7, and -8
atoms are acceptably directed toward the Cys191-Cys220
disulfide bond and toward the backbone of the Gly216-Ser218
sequence (Figure 3). Moreover, the basis for the reversal of
the normal L-stereoselectivity of CT is also revealed in that
binding of D-1e becomes preferred over L-1e because of the
strong hydrogen bond D-1e forms between the carbonyl oxygen
of its acetamido group and the NH of His57. Conversely,
analysis of the molecular dynamics trajectory for the complex
of CT with L-1e provides no indication of such hydrogen bond
stabilization between the acetamido carbonyl and its nearest NH
neighbor, Ser214. Significantly, for the complexes of D-1a-d
with CT, the molecular dynamics simulations do not identify
an acetamido CO to histidine NH hydrogen bond, but instead
indicate that in these cases the acetamido groups are exposed
to the solvent. Thus, inhibitions of CT by the L-1a-d inhibitors
remain superior to those by their D-counterparts. However, the
slightly better (3-9-fold) binding to CT of D-1b-d relative to
that of 2b-d, respectively, might reflect very weak hydrogen
bond interactions of the D-1e-His57 type. The CT-L-1e
complex is further disfavored by the somewhat adverse interac-
tions of its naphthalene ring with the Cys191-Cys220 disulfide
bond and with the Gly216-Ser218 backbone. These direct the
acetamido group away from Ser214 and toward Val213. The
calculated EI structures are shown in Figure 3. Interestingly,
the calculated positions of the naphthalene rings of the 1e
enantiomers coincide more closely for the CT complexes than
they do for those of SC (Figure 2). Attempts to correlate the
calculated energies of the minimized structures, ranging from
1863 to 1914 kcal/mol for SC and from 2124 to 2179 kcal/mol
for CT, with the differences in KI values of the Table 1 inhibitors
were unsuccessful because of the variations in each complex
of the positions of the active site region water molecules
included in the simulations. With different water positions
affecting the minimizations uniquely, we were unable to make
valid comparisons of the final energies of the individual
complexes.
The indication from molecular modeling that SC and CT can
exhibit preferences for different inhibitor conformations raises
the prospect of exploiting conformer preference as a means of
tailoring enzyme stereoselectivity toward appropriate substrates,
perhaps Via strategies mimicking those applied in conforma-
tionally restricted enzyme inhibitor approaches to drug design.
However, clearly many more data are needed before reliable
guidelines for achieving such control can be formulated.
Furthermore, once again it must be noted that a possibility
remains that a change in boronic acid binding pattern, from the
serine preference considered in the current modeling studies to
one favoring histidine binding, could be at least partly respon-
sible for the the KI and stereoselectivity trends observed. In
this regard, the X-ray structures currently being determined of
the SC and CT complexes with L- and D-1b-e23 will provide
(23) Pai, E.; Stoll. V. Work in progress.

Specificity of Subtilisin Carlsberg and R-Chymotrypsin
J. Am. Chem. Soc., Vol. 118, No. 5, 1996 955
1.06 (d, J ) 10.58 Hz, 1 H), 1.25 (s, 3 H), 1.36 (s, 3 H), 1.70-2.31
(m, 5 H), 2.75 (s, 2 H), 4.24 (dd, J ) 2.02 and 8.73 Hz, 1 H), 7.36-
8.07 (m, 7 H). 13C NMR (CDCl3): δ 17.36, 23.74, 26.23, 26.82, 28.40,
35.22, 37.92, 39.25, 51.09, 77.89, 85.86, 124.59, 125.43, 125.50, 125.83,
125.90, 126.46, 128.64, 132.47, 133.92, 135.76.
valuable insights, particularly into the concept of conformational
selection by enzymes.
Experimental Section
General Methods. Unless otherwise stated, all reactions were
performed under nitrogen using oven-dried glassware. Anhydrous
reagents and solvents were prepared according to literature procedures.24
Analytical thin layer chromatography was performed on Merck plates
(silica gel F254, 0.25 mm). Compounds that were not visualized by
UV were detected by spraying with a mixture of ninhydrin (0.3 g) and
acetic acid (3 mL) in ethanol (100 mL) followed by heating. Preparative
flash column chromatography was performed using silica gel 60 (40-
63 µm), supplied by Toronto Research Chemicals Inc. Melting points
were obtained on a Fisher-Johns melting point apparatus, and are
uncorrected. Boiling points are of Kugelrohr distillations and are
uncorrected. Optical rotations were measured with a Perkin-Elmer 243
B polarimeter equipped with a thermostated cell. Infrared (IR) spectra
were determined in KBr pellets (for solids) and films (liquids) on a
Nicolet 5DX FTIR spectrophotometer. 1H and 13C NMR spectra were
recorded on a Gemini 200 (at 200 and 50 MHz, respectively)
spectrometer unless otherwise indicated. 1H NMR chemical shifts are
reported in parts per million relative to the CHCl3 peak (δ ) 7.24)
with CHCl3 as solvent, the DMSO peak (δ ) 2.49) in DMSO-d6, and
the HOD peak (δ ) 4.80) in D2O. 13C NMR chemical shifts are
reported in parts per million relative to the CHCl3 peak (δ ) 77.00)
with CHCl3 as solvent, the DMSO peak (δ ) 39.50) in DMSO-d6 and
external dioxane (δ ) 66.50) in D2O as solvent. Mass spectra were
measured on a Bell and Howell 21-490 (low resolution) or an AEI
MS3074 (high resolution) instrument. Elemental analyses were by
Galbraith Laboratories, Knoxville, TN.
A solution of CH2Cl2 (1.02 mL, 16 mmol) in dry THF (30 mL) was
cooled to -100 °C in an EtOH/liquid N2 bath and stirred magnetically
during the dropwise addition of n-BuLi (6.9 mL of a 1.6 M solution in
hexane, 11 mmol) by running the BuLi solution down the cold wall of
the reaction flask. After about half of the BuLi had been added, a
white precipitate of LiCHCl2 became evident. Twenty minutes after
all the BuLi had been added, (+)-pinanediol (1-naphthylmethyl)-
boronate (3e; 3.2 g, 10 mmol) in dry Et2O (10 mL) was added in one
portion. The solution was stirred at -100 °C for 10 min, after which
27
a portion of rigorously dried ZnCl2 (0.56 g, 4.1 mmol) was added.
The mixture, still in the cooling bath, was then allowed to warm slowly
to 20 °C and stirred overnight. The solution was concentrated by rotary
evaporation (bath temperature <30 °C), and the residue was dissolved
in Et2O (25 mL), treated with silica gel (10 g), and then triturated with
hexanes (25 mL). The resulting mixture was loaded on a short silica
gel column (length 10 cm, diameter 3.5 cm) and eluted with hexanes/
Et2O (1:1, 300 mL). The eluants were concentrated on a rotary
evaporator (bath temperature <30 °C) to give (+)-pinanediol [(1S)-1-
chloro-2-(1-naphthyl)ethyl]boronate (4e; 2.97 g, 81%) as an oil, [R]23
D
) +22.1 (c 2.13, toluene). IR (film): ν 2924, 1455, 1408, 1240 cm-1
.
1H NMR (CDCl3): δ 0.83 (s, 3 H), 1.06 (d, J ) 11.00 Hz, 1 H), 1.28
(s, 3 H), 1.35 (s, 3 H), 1.83-2.34 (m, 5 H), 3.16 (m, 2 H), 3.67 (t, J
) 8.06 Hz, 1 H), 4.35 (dd, J ) 1.91 and 8.87 Hz, 1 H), 7.28 (m, 5 H).
13C NMR (CDCl3): δ 23.62, 25.94, 26.69, 27.96, 34.82, 37.20, 37.91,
39.04, 41.62, 50.89, 78.34, 86.64, 123.44, 125.30, 125.59, 126.16,
127.48, 127.65, 128.89, 131.82, 133.93, 134.21.
Reagent grade chemicals were purchased from Aldrich. Subtilisin
Carlsberg (SC; EC 3.4.21.14), R-chymotrypsin (CT; EC 3.4.21.1), N-p-
tosyl-L-arginine methyl ester (TAME), and N-acetyl-L-tyrosine ethyl
ester (NATEE) were purchased from Sigma Chemical Co., St. Louis,
MO. The concentrations of SC and R-chymotrypsin were determined
by assaying the rates of hydrolysis of standard solutions of succinyl-
L-Ala-L-Ala-L-Pro-L-Phe-p-nitroanilide25 and p-nitrophenyl acetate,26
respectively.
A solution of lithium hexamethyldisilazane was prepared at -78
°C from hexamethyldisilazane (1.84 mL, 8.70 mmol) and n-BuLi (5.25
mL of a 1.6 M solution in hexane, 8.40 mmol) in THF (15 mL). (+)-
Pinanediol [(1S)-1-chloro-2-(1-naphthyl)ethyl]boronate (4e; 2.68 g, 7.30
mmol) was added at -78 °C to the stirred solution and the mixture
then allowed to warm to 20 °C and stirred for 10 h. The mixture was
then cooled to -78 °C, and Ac2O (2.82 mL, 25.5 mmol) and CH3-
COOH (0.53 mL, 9.3 mmol) were added dropwise with stirring. After
the mixture had been stirred overnight at 20 °C, the solvent was
removed on a rotary evaporator and EtOAc (30 mL) and H2O (30 mL)
were added. The aqueous layer was separated and extracted with
EtOAc (2 × 20 mL). The combined EtOAc layers were washed
successively with 5% aqueous NaHCO3 (25 mL), H2O (25 mL), and
brine (25 mL) and dried (MgSO4). The solution was concentrated under
vacuum, and the residue flash-chromatographed on a silica gel column
with 5% MeOH in Et2O elution. The eluents were rotary evaporated
and the product recrystallized to constant rotation from EtOAc/CHCl3
to afford (+)-pinanediol [(1R)-1-acetamido-2-(1-naphthyl)ethyl]bor-
Preparations of Inhibitors. L-Series. The same basic, Scheme 1,
procedure was used for each inhibitor, with the following being
representative.
[(1R)-1-Acetamido-2-(1-naphthyl)ethyl]boronic Acid (L-1e). A
solution of (1-naphthylmethyl)magnesium chloride, obtained from
1-naphthylmethyl chloride (8.8 g, 50 mmol) and activated Mg (1.2 g,
50 mmol) in dry Et2O (50 mL) at 20 °C, was added slowly (over 15
min) at -70 °C to a solution of trimethyl borate (5.68 mL, 50 mmol)
in dry Et2O (150 mL) and the resulting white suspension stirred for 1
h at -70 °C. The mixture was then allowed to warm to 20 °C and
stirred for a further 10 h. Aqueous H2SO4 (10%, 50 mL) was then
slowly added at 0 °C, the Et2O layer separated, and the aqueous phase
extracted with Et2O (3 × 50 mL). The combined ethereal phases were
washed with water and then extracted with 2 M aqueous KOH (3 ×
50 mL). The aqueous extracts were combined, washed with Et2O,
cooled to 0 °C, and then acidified to pH 2 with 10% aqueous H2SO4
and saturated with NaCl. Et2O was added (50 mL), the Et2O layer
separated, and the aqueous phase extracted further with Et2O (2 × 50
mL). The combined organic phases were dried (MgSO4) and filtered,
and (+)-pinanediol (4.25 g, 25 mmol) was added at 20 °C to the stirred
solution. The course of the reaction was followed by TLC, and was
generally complete in 1 h, although sometimes addition of more
pinanediol was needed to complete the reaction. The mixture was then
filtered, concentrated in Vacuo, and Kugelrohr distilled to give (+)-
pinanediol (1-naphthylmethyl)boronate (3e; 6.24 g, 39%), bp 160-
onate (5e; 1.88 g, 66%), mp 235-239 °C, [R]23 ) -55.5 (c 2.88,
D
CHCl3). IR (KBr): ν 3173, 1606, 1561, 1170 cm-1
.
1H NMR
(CDCl3): δ 0.88 (s, 3 H), 1.29 (s, 3 H), 1.42 (s, 3 H), 1.55 (d, J )
10.26 Hz, 1 H), 1.86-2.06 (m, 3 H), 1.98 (s, 3 H), 2.18-2.36 (m, 2
H), 3.09-3.20 (m, 2 H), 3.46 (q, J ) 10.53 Hz, 1 H), 4.27 (dd, J )
2.03 and 8.49 Hz, 1 H), 6.16 (br s, 1 H, variable position), 7.27-8.06
(m, 7 H). 13C NMR (DMSO-d6): δ 17.14, 23.99, 25.99, 26.09, 27.24,
29.49, 34.83, 36.72, 37.58, 39.65, 43.84, 52.23, 75.43, 81.88, 123.71,
125.68, 125.81, 126.07, 126.72, 127.17, 128.95, 131.74, 133.87, 137.04,
175.24.
To a stirred solution of BCl3 (15 mL of a 1 M solution in CH2Cl2)
in CH2Cl2 (30 mL) at -78 °C was added solid (+)-pinanediol [(1R)-
1-acetamido-2-(1-naphthyl)ethyl]boronate (5e; 1.27 g, 3.25 mmol). The
mixture was stirred at -78 °C for 1 h and the cooling bath then
removed. The mixture was evaporated under vaccum (0.1 mmHg)
while evolving BCl3 and CH2Cl2 were condensed in a trap at -78 °C.
Water (30 mL) and Et2O (70 mL) were added, and the aqueous phase
was separated and washed with Et2O (3 × 20 mL). Lyophilization of
the aqueous phase yielded crude 1e, which was freed from the boric
acid contaminant by successive treatments with MeOH (50 mL)
followed by distillation until the distillate showed no green boron color
in the flame when a drop was ignited. The remaining MeOH was
165 °C (0.2 mmHg), [R]23 ) +24.4 (c 6.52, toluene). IR (film): ν
D
2926, 1458, 1282, 1238 cm-1 1H NMR (CDCl3): δ 0.79 (s, 3 H),
.
(24) Perin, D. D.; Armarego, W. L. F.; Perrin, D. R. Purification of
Laboratory Chemicals; Pergamon Press: New York, 1980.
(25) (a) Del Mar, E. G.; Largman, C.; Brodrick, J. W.; Goekas, M. C.
Anal. Biochem. 1979, 99, 316. (b) Russell, A. J.; Thomas, P. G.; Fersht, A.
R. J. Mol. Biol. 1987, 193, 803.
(26) (a) Ke´zdy, F. J.; Kaiser, E. T. Methods Enzymol. 1970, 19, 3. (b)
Ottensen, M.; Svendsen, I. In Methods of Enzymatic Analysis, Berman, H.
V., Ed.; 1984; Vol. 5, p 159.
(27) Zinc chloride was dried at 130 °C and 0.05 mmHg with magnetic
stirring for 12 h.

956 J. Am. Chem. Soc., Vol. 118, No. 5, 1996
Martichonok and Jones
removed by distillation under vacuum, and water (30 mL) and Et2O
(30 mL) were added to the residue. The aqueous phase was separated,
washed with Et2O (3 × 20 mL), and then lyophilized to give [(1R)-
1-acetamido-2-(1-naphthyl)ethyl]boronic acid (1e; 0.36 g, 43%) as the
mmHg), [R]23 ) +31.8 (c 6.00, toluene). IR (film): ν 2928, 1282,
D
1238, 1076 cm-1 1H NMR (CDCl3): δ 0.85 (s, 3 H), 1.08 (d, J )
.
10.80 Hz, 1 H), 1.30 (s, 3 H), 1.41 (s, 3 H), 1.80-2.39 (m, 5 H), 2.37
(s, 2 H), 4.30 (dd, J ) 1.82 and 8.67 Hz, 1 H), 7.15-7.38 (m, 5 H).
13C NMR (CDCl3): δ 20.00, 24.40, 26.16, 26.83, 28.37, 35.23, 37.91,
39.26, 51.13, 77.82, 85.71, 124.90, 128.33, 129.00, 138.85. (+)-
Pinanediol [(1S)-1-chloro-2-phenylethyl]boronate (4b; 92%), recrystal-
lized from EtOH, mp 46-47 °C, [R]23D ) +25.0 (c 2.35, toluene). IR
trimeric anhydride, mp 166-171 °C, [R]23 ) -142.3 (c 1.22, CH3-
D
OH). IR (KBr): ν 3700-2800, 1631, 1248 cm-1 1H NMR (CD3-
.
OD): δ 2.13 (s, 3 H), 3.02 (m, 2 H), 3.40 (m, 1 H), 7.35-8.07 (m, 7
H). 13C NMR (CD3OD): δ 16.41, 34.68, 48.87 (br), 124.55, 126.54,
126.63, 126.93, 127.67, 128.03, 129.87, 133.18, 135.55, 137.51, 178.43.
HRMS: calcd for C42H42 B3N3O6 717.3353, found 717.3358.
(KBr): ν 2930, 1239, 1077, 1008 cm-1 1H NMR (CDCl3): δ 0.83
.
(s, 3 H), 1.06 (d, J ) 11.00 Hz, 1 H), 1.28 (s, 3 H), 1.35 (s, 3 H),
1.83-2.34 (m, 5 H), 3.16 (m, 2 H), 3.67 (t, J ) 8.06 Hz, 1 H), 4.35
(dd, J ) 1.91 and 8.87 Hz, 1 H), 7.28 (m, 5 H). 13C NMR (CDCl3):
δ 23.70, 25.98, 26.76, 28.10, 34.91, 37.98, 39.09, 40.17, 42.93, 50.93,
78.39, 86.71, 126.78, 128.41, 129.24, 138.49. (+)-Pinanediol [(1R)-
1-acetamido-2-phenylethyl]boronate (L-5b; 81%), mp 190-192 °C,
[R]23D ) -82.5 (c 4.90, CHCl3) (lit.15h mp 185-186 °C, [R]23D ) -82.4
The target L-1e was furthered characterized as its diethanolamine
derivative L-6e as follows: diethanolamine (0.21 g, 2 mmol) in
2-propanol (5 mL) was added with stirring at 20 °C to a solution of
[(1R)-1-acetamido-2-(1-naphthyl)ethyl]boronic acid (L-1e; 514 mg, 2
mmol) in i-PrOH (5 mL). The solution was stirred for a further 1 h,
the i-PrOH then rotary evaporated, and the residue dissolved in CH2-
Cl2 (20 mL). Anhydrous MgSO4 and activated charcoal were added,
and the mixture was stirred for 10 h. Filtration through Celite, followed
by concentration under vacuum, afforded a white solid which was
further purified by repeated precipitation from CH2Cl2 with Et2O to
yield diethanolamine [(1R)-1-acetamido-2-(1-naphthyl)ethyl]boronate
(c 5.00, CHCl3)). IR (KBr): ν 3183, 3072, 1609, 1161 cm-1 1H NMR
.
(CDCl3): δ 0.87 (s, 3 H), 1.28 (s, 3 H), 1.40 (s, 3 H), 1.43 (d, J )
9.40 Hz, 1 H), 1.82-2.34 (m, 5 H), 2.04 (s, 3 H), 2.72 (dd, J ) 12.29
Hz and 14.85 Hz, 1 H), 2.97 (m, 2 H), 4.24 (dd, J ) 2.04 and 8.52 Hz,
1 H), 6.30 (br s, 1 H, variable positions), 7.27 (m, 5 H). 13C NMR
(CDCl3): δ 17.75, 23.83, 25.96, 26.97, 28.91, 36.09, 36.96, 37.76,
39.64, 44.51, 51.98, 75.91, 83.09, 126.12, 128.41, 128.77, 140.46,
174.68. [(1R)-1-Acetamido-2-phenylethyl]boronic acid (L-1b; 83%),
mp 145-147 °C, [R]23 ) -185.2 (c 1.33, H2O) (lit.15h [R]22 (as
(L-6e; 228 mg, 35%), mp 240-245 °C dec, [R]23 ) -129.6 (c 0.55,
D
CH2Cl2). IR (KBr): ν 3254, 3085, 1630, 1218 cm-1
.
1H NMR
(CDCl3): δ 1.89 (s, 3 H), 2.52-2.74 (m, 3 H), 3.07-3.24 (m, 3 H),
3.34-3.83 (m, 5H), 6.76 (br t, 1 H), 7.20-8.13 (m, 8H). 13C NMR
(CDCl3): δ 23.26, 31.87, 46.65 (br), 50.67, 50.85, 63.04 (double peak),
123.88, 125.33, 125.45, 125.60, 125.76, 126.47, 128.61, 132.03, 133.89,
137.59, 172.18. Anal. Calcd for C18H23BN2O3: C, 66.28; H, 7.11;
N, 8.59. Found: C, 66.21; H, 7.15; N, 8.49.
D
D
anhydride) ) -195.8 (c 0.60, H2O). IR (KBr): ν 3700-2800, 1635,
1268 cm-1 1H NMR (D2O) δ 2.11 (s, 3 H), 2.60 (dd, J ) 12.64 and
.
15.57 Hz, 1 H), 2.88 (m, 2 H), 7.30-7.43 (m, 5H). 13C NMR (D2O,
CH3OH as internal standard, δ 49.90): δ 17.03, 37.18, 50.59 (br),
127.56, 129.90, 130.05, 141.80, 177.94. HRMS: calcd for C30H36
B3N3O6 567.2883, found 567.2859. Diethanolamine [(1R)-1-acetamido-
2-phenylethyl]boronate (L-6b; 42%), mp 177-178 °C, [R]23D ) -32.8
(c 0.96, CH2Cl2); IR (KBr): ν 3449, 3394, 1620, 1105, 1079 cm-1. 1H
NMR (CDCl3) δ 1.86 (s, 3 H), 2.71-2.86 (m, 3 H), 3.10-3.23 (m, 3
H), 3.32-3.56 (m, 1 H), 3.76-4.10 (m, 4 H), 5.87 (br s, 1 H), 7.20-
7.34 (m, 5 H), 7.64 (br s, 1 H). 13C NMR (CDCl3): δ 23.14, 31.80,
46.90 (br), 50.69, 51.01, 63.09 (double peak), 125.62, 128.16, 128.12,
141.52, 172.07. HRMS: calcd for C14H21BN2O3 276.1645, found
276.1646.
The other inhibitors in this (1R)-configuration series, L-1a-d, were
prepared on the same scale by the above procedures, Via the Scheme
1 intermediates, as follows.
[(1R)-1-Acetamidoethyl]boronic Acid (L-1a). (+)-Pinanediol meth-
ylboronate (3a; 41%, obtained using purchased methylmagnesium
bromide), bp 60-65 °C (3 mmHg), [R]23 ) +37.0 (c 3.58 CHCl3)
D
(lit.18d bp 37-41 °C (0.25 mmHg)). IR (film): ν 2931, 1280, 1078
cm-1 1H NMR (CDCl3): δ 0.26 (s, 3 H), 0.81 (s, 3 H), 1.09 (d, J )
.
10.67 Hz, 1 H), 1.26 (s, 3 H), 1.36 (s, 3 H), 1.86-2.30 (m, 5 H), 4.23
(dd, J ) 1.85 and 8.69 Hz, 1 H). 13C NMR (CDCl3): δ 23.77, 26.22,
26.87, 28.45, 35.27, 37.93, 39.33, 51.14, 77.55, 85.36 (C next to B not
seen). (+)-Pinanediol [(1S)-1-chloroethyl]boronate (L-4a; 79%), bp
[(1R)-1-Acetamido-2-(1-fluorophenyl)ethyl]boronic Acid (L-1c).
(+)-Pinanediol [(4-fluorophenyl)methyl]boronate (3c; 44%), bp 120-
122 °C (0.35 mmHg), [R]23 ) +29.5 (c 6.60 toluene). IR (film): ν
58-60 °C (0.1 mmHg), [R]23 +33.6 (c 2.35, toluene). IR (film): ν
D
D
2928, 1456, 1412, 1394, 1380, 1339, 1284, 1240, 1076, 1006 cm-1
.
2922, 1286, 830 cm-1 1H NMR (CDCl3): δ 0.80 (s, 3 H), 1.02 (d, J
.
1H NMR (CDCl3): δ 0.82 (s, 3 H), 1.14 (d, J ) 10.94 Hz, 1 H), 1.27
(s, 3 H), 1.40 (s, 3 H), 1.55 (d, J ) 7.57 Hz, 3 H), 1.83-2.34 (m, 5
H), 3.55 (q, J ) 7.51 Hz, 1 H), 4.34 (dd, J ) 1.91 and 8.83 Hz, 1 H).
13C NMR (CDCl3): δ 20.37, 23.75, 26.07, 26.23, 26.81, 28.21, 35.07,
38.06, 39.15, 51.05, 78.52, 86.73. (+)-Pinanediol [(1R)-1-acetamido-
) 10.62 Hz, 1 H), 1.25 (s, 3 H), 1.35 (s, 3 H), 1.76-2.36 (m, 5 H),
2.28 (s, 2 H), 4.25 (dd, J ) 1.91 and 8.59 Hz, 1 H), 6.90 (t, J ) 8.50
Hz, 2 H), 7.11 (dd, J ) 5.51 and 8.53 Hz, 2 H). 13C NMR (CDCl3):
δ 18.39, 23.73, 26.18, 26.82, 28.37, 35.25, 37.95, 39.27, 51.13, 77.89,
85.84, 115.00 (d, J ) 21.1 Hz), 130.24 (d, J ) 7.6 Hz), 34.37 (d, J )
2.2 Hz), 160.97 (d, J ) 242.8 Hz). (+)-Pinanediol [(1S)-1-chloro-2-
ethyl]boronate (5a; 73%), mp 196-198 °C, [R]23 ) -21.24 (c 1.13,
D
CHCl3) (lit.19 mp 197-198 °C, [R]21546 ) -25.50 (c 2.9, CHCl3)). IR
(4-fluorophenyl)ethyl]boronate (4c; 87%), mp 64-65 °C, [R]23
)
D
(KBr): 3182, 1612, 1286, 1082 cm-1 1H NMR (CDCl3): δ 0.82 (s,
.
+25.5 (c 2.19, toluene). IR (KBr): ν 2920, 1285, 832 cm-1 1H NMR
.
3 H), 1.08 (d, J ) 7.24 Hz, 3 H), 1.22 (s, 3 H), 1.34 (s, 3 H), 1.43 (d,
J ) 9.40 Hz, 1 H), 1.70-2.28 (m, 5 H), 1.99 (s, 3 H), 2.53 (q, J )
7.32 Hz, 1 H), 4.14 (dd, J ) 2.07 and 8.47 Hz, 1 H), 9.20 (br s, 1 H,
variable positions). 13C NMR (CDCl3): δ 15.87, 16.78, 23.89, 26.34,
27.08, 29.13, 36.77, 37.87, 39.28, 39.94, 52.38, 75.63, 82.76, 174.75;
[(1R)-1-acetamidoethyl]boronic acid (L-1a; 91%), mp 185-188 °C,
[R]23D ) -82.8 (c 0.39, H2O). IR (KBr): ν 3700-2800, 1630, 1540,
(CDCl3): δ 0.83 (s, 3 H), 1.04 (d, J ) 10.62 Hz, 1 H), 1.29 (s, 3 H),
1.36 (s, 3 H), 1.82-2.36 (m, 5 H), 3.13 (m, 2 H), 3.62 (t, J ) 7.98 Hz,
1 H), 4.35 (dd, J ) 2.04 and 8.71 Hz, 1 H), 6.98 (t, J ) 8.10 Hz, 2 H),
7.23 (dd, J ) 5.49 and 8.06 Hz, 2 H); 13C NMR (CDCl3): δ 23.69,
25.99, 26.75, 28.12, 34.91, 38.00, 39.10, 39.30, 42.99 (br), 50.94, 78.46,
86.81, 115.11 (d, J ) 21.2 Hz), 131.20 (d, J ) 7.9 Hz), 136.23 (d, J
) 3.0 Hz), 161.96 (d, J ) 245.3 Hz). (+)-Pinanediol [(1R)-1-
acetamido-2-(4-fluorophenyl)ethyl]boronate (5c; 83%), recrystallized
1429, 1383, 1240, 807, cm-1 1H NMR (D2O) δ 1.10 (d, J ) 7.32 Hz,
.
3 H), 2.16 (s, 3 H), 2.70 (q, J ) 7.28 Hz, 1 H). 13C NMR (D2O,
CH3CN as internal standard, δ 1.60 for CH3): δ 15.23, 16.76, 43.57
(br), 177.37. The HRMS was not obtainable, so L-1a was further
characterized as diethanolamine [(1R)-1-acetamidoethyl]boronate (6a;
from EtOAc, mp 139-140 °C, [R]23 ) -68.4 (c 4.62, CHCl3). IR
D
(KBr): ν 3184, 3076, 1608, 1160 cm-1 1H NMR (CDCl3): δ 0.82
.
(s, 3 H), 1.24 (s, 3 H), 1.32 (d, J ) 10.43 Hz, 1 H), 1.34 (s, 3 H),
1.75-2.34 (m, 5 H), 2.00 (s, 3 H), 2.66 (dd, J ) 11.76 and 14.98 Hz,
1 H), 2.90 (m, 2 H), 4.18 (dd, J ) 2.20 and 8.79 Hz, 1 H), 6.25 (br s,
1 H, variable positions), 6.95 (t, J ) 8.73 Hz, 2 H), 7.13 (dd, J ) 5.50
and 8.68 Hz, 2 H); 13C NMR (CDCl3): δ 17.92, 23.82, 25.94, 26.97,
28.92, 36.06, 36.18, 37.81, 39.64, 44.32 (br), 52.00, 76.1, 83.3, 115.12
(d, J ) 21.1 Hz), 130.23 (d, J ) 7.8 Hz), 136.01 (d, J ) 3.2 Hz),
161.50 (d, J ) 244.7 Hz), 174.65. [(1R)-1-Acetamido-2-(1-fluorophen-
49%), mp 175-176 °C, [R]23 ) -22.4 (c 0.90, CH2Cl2). IR (KBr):
D
ν 3307, 3115, 1626, 1309, 1099 cm-1 1H NMR (CDCl3) δ 1.20 (d, J
.
) 7.61 Hz, 3 H), 1.93 (s, 3 H), 2.71-2.78 (m, 2 H), 2.91-3.02 (m, 2
H), 3.24-3.41 (m, 1 H), 3.78-4.01 (m, 4 H), 5.70 (br s, 1 H), 7.43 (br
s, 1 H). 13C NMR (CDCl3): δ 15.57, 23.15, 39.70 (br), 50.67, 51.02,
62.95, 63.10, 171.10. Anal. Calcd for C8H17 BN2O3: C, 48.03; H,
8.57. Found: C, 47.81; H, 8.59.
yl)ethyl]boronic acid (L-1c; 68%), mp 132-135 °C, [R]23 ) -156.8
D
[(1R)-1-Acetamido-2-phenylethyl]boronic Acid (l-1b). (+)-Pi-
nanediol (phenylmethyl)boronate (3b; 44%), bp 110-112 °C (0.2
(c 0.91, H2O). IR (KBr): ν 3700-2800, 1630, 1222, 824 cm-1 1H
.
NMR (D2O): δ 2.10 (s, 3 H), 2.54 (dd, J ) 12.72 and 15.57 Hz, 1 H),
2.82 (m, 2 H), 7.08 (t, J ) 8.83 Hz, 2 H), 7.28 (dd, J ) 5.64 and 8.80
mmHg), [R]23 +31.6 (c 6.30 toluene) (lit.18c bp 108-110 °C (0.1
D

Specificity of Subtilisin Carlsberg and R-Chymotrypsin
J. Am. Chem. Soc., Vol. 118, No. 5, 1996 957
Hz, 2H). 13C NMR (D2O): δ 16.99, 36.30, 50.47 (br), 116.25 (d, J )
21.1 Hz), 131.47 (d, J ) 8.0 Hz), 137.29 (d, J ) 2.8 Hz), 162.50 (d,
J ) 241.9 Hz), 178.09. Anal. Calcd for C10H13BFNO3: C, 53.38; H,
5.82; N, 6.22. Found: C, 53.25; H, 5.66; N, 6.05. Diethanolamine
[(1R)-1-acetamido-2-(4-fluorophenyl)ethyl]boronate (L-6c; 56%), mp
doethyl]boronate (D-6a; 55%), mp 170-173 °C, [R]23 ) +23.0 (c
0.95, CH2Cl2). HRMS: calcd for C8H17BN2O3 200.1332, found
200.1326.
D
[(1S)-1-Acetamido-2-phenylethanyl]boronic Acid (D-1b). (-)-
Pinanediol (phenylmethyl)boronate (ent-3b; 47%), bp 110-112 °C (0.2
mmHg), [R]23 ) -30.6 (c 6.30, toluene) (lit.17a bp 110-112 °C (0.2
197-200 °C, [R]23 ) -29.6 (c 0.68, CH2Cl2). IR (KBr): ν 3381,
D
D
mmHg), [R]23 ) -30.6 (c 6.30 toluene). (-)-Pinanediol [(1R)-1-
3085, 1622, 1112, 826 cm-1 1H NMR (CDCl3): δ 1.82 (s, 3 H), 2.65-
.
D
chloro-2-phenylethyl]boronate (ent-4b; 89%), mp 46-47 °C, [R]23
)
2.82 (m, 3 H), 2.99-3.12 (m, 3 H), 3.29-3.46 (m, 1 H), 3.72-4.01
(m, 4 H), 5.85 (br d, J ) 5.13 Hz, 1 H), 6.92 (t, J ) 8.76 Hz, 2 H),
7.12 (dd, J ) 5.49 and 8.58 Hz, 2H), 7.51 (br s, 1 H); 13C NMR
(CDCl3): δ 23.12, 31.05, 46.71 (br), 50.71, 51.02, 63.12 (double peak),
114.87 (d, J ) 21.0 Hz), 129.73 (d, J ) 7.7 Hz), 137.08 (d, J ) 3.1
Hz), 161.15 (d, J ) 243.2 Hz), 172.01. Anal. Calcd for C14H20-
BFN2O3: C, 57.17; H, 6.85; B, 3.68; N, 9.52. Found: C, 56.75; H,
6.98; B, 3.47; N, 9.26.
D
-24.8 (c 2.25, toluene). (-)-Pinanediol [(1S)-1-acetamido-2-phenyl-
ethyl]boronate (ent-5b; 82%), mp 191-192 °C, [R]23D +82.0 (c 4.85,
CHCl3) (lit.17a [R]21 +82.1 (c 4.00, CHCl3). [(1S)-1-Acetamido-2-
D
phenylethyl]boronic acid (D-1b; 77%), mp 146-148 °C, [R]23D +184.8
(c 1.28, H2O) (lit.15h [R]22 as anhydride ) +195.0 (c 0.70, H2O).
D
HRMS: calcd for C30H36B3N3O6 567.2883, found 567.2899. Dietha-
nolamine [(1S)-1-acetamido-2-phenylethyl]boronate (D-6b; 48%), mp
179-180 °C, [R]23 ) +32.8 (c 1.05, CH2Cl2). HRMS: calcd for
D
[(1R)-1-Acetamido-2-(1-chlororophenyl)ethyl]boronic Acid (L-
1d). (+)-Pinanediol [(4-chlorophenyl)methyl]boronate (3d; 40%),
C14H21BN2O3 276.1645, found 276.1659.
[(1S)-(1-Acetamido-2-(1-fluorophenyl)ethyl]boronic Acid (D-1c).
(-)-Pinanediol [(4-fluorophenyl)methyl]boronate (ent-3c; 40%), bp
recrystallized from Et2O/hexanes, mp 59 °C, [R]23 ) +25.9 (c 6.06
D
toluene). IR (KBr): ν 2912, 1283, 1078, 806 cm-1
.
1H NMR
120-122 °C (0.35 mmHg), [R]23 ) -29.06 (c 6.04, toluene). (-)-
D
(CDCl3): δ 0.81 (s, 3 H), 1.01 (d, J ) 10.62 Hz, 1 H), 1.26 (s, 3 H),
1.36 (s, 3 H), 1.99-2.33 (m, 5 H), 2.29 (s, 2 H), 4.26 (d, J ) 8.57 Hz,
1 H), 7.10 (d, J ) 8.22 Hz, 2 H), 7.19 (d, J ) 8.24 Hz, 1 H). 13C
NMR (CDCl3): δ 18.42 (br), 23.76, 26.22, 26.84, 28.39, 35.24, 37.98,
39.27, 51.13, 77.95, 85.94, 128.44, 130.36, 130.71, 137.43. (+)-
Pinanediol [(1S)-1-chloro-2-(4-chlorophenyl)ethyl]boronate (L-4d; 95%),
Pinanediol [(1R)-1-chloro-2-(4-fluorophenyl)ethyl]boronate (ent-4c;
81%), mp 63-64 °C, [R]23 ) -25.15 (c 2.29, toluene). (-)-
D
Pinanediol [(1S)-1-acetamido-2-(4-fluorophenyl)ethyl]boronate (ent-5c;
79%), mp 139-140 °C, [R]23 ) +68.5 (c 4.70, CHCl3). [(1S)-(1-
D
Acetamido-2-(1-fluorophenyl)ethyl]boronic acid (D-1c; 66%), mp 132-
135 °C, [R]23 ) +156.0 (c 0.72, H2O). HRMS: calcd for
D
recrystallized from EtOH, mp 92-93 °C, [R]23 ) +23.5 (c 2.20,
D
C30H33B3F3N3O6 621.2601, found 621.2589. Diethanolamine [(1S)-1-
acetamido-2-(4-fluorophenyl)ethyl]boronate (D-6c; 62%), mp 197-200
°C, [R]23D +30.3 (c 0.74, CH2Cl2). HRMS: calcd for C14H20BFN2O3
294.1551, found 294.1548.
toluene). IR (KBr): ν 2910, 1284, 1075, 805 cm-1
.
1H NMR
(CDCl3): δ 0.84 (s, 3 H), 1.05 (d, J ) 10.62 Hz, 1 H), 1.29 (s, 3 H),
1.37 (s, 3 H), 1.83-2.34 (m, 5 H), 3.12 (m, 2 H), 3.62 (t, J ) 7.88 Hz,
1 H), 4.36 (d, J ) 8.71 Hz, 1 H), 7.20 (d, J ) 8.36 Hz, 2 H), 7.28 (d,
J ) 8.36 Hz, 2 H). 13C NMR (CDCl3): δ 23.71, 26.02, 26.77, 28.15,
34.92, 38.02, 39.11, 39.41, 42.38, 50.96, 78.51, 86.88, 128.54, 130.68,
132.65, 137.04. (+)-Pinanediol [(1R)-1-acetamido-2-(4-chlorophenyl)-
[(1S)-1-Acetamido-2-(1-chlororophenyl)ethyl]boronic Acid (D-1d).
(-)-Pinanediol [(4-chlorophenyl)methyl]boronate (ent-3d; 38%), mp
59-60 °C (Et2O/hexanes), [R]23 -26.3 (c 6.00, toluene). (-)-
D
Pinanediol [(1S)-1-chloro-2-(4-chlorophenyl)ethyl]boronate (ent-4d;
ethyl]boronate (5d; 79%), mp 152-153 °C, [R]23 ) -78.6 (c 5.36,
D
91%), mp 92 °C, [R]23 ) -24.9 (c 2.03, toluene). (-)-Pinanediol
D
CHCl3). IR (KBr): ν 3170, 3075, 1606, 1066 cm-1
.
1H NMR
[(1S)-1-acetamido-2-(4-chlorophenyl)ethyl]boronate (ent-5d; 81%), mp
152-153 °C, [R]23D ) +79.1 (c 5.15, CHCl3). [(1S)-1-Acetamido-2-
(1-chlorophenyl)ethyl]boronic acid (D- 1d; 79%), mp 138-141 °C,
(CDCl3): δ 0.85 (s, 3 H), 1.25 (s, 3 H), 1.34 (d, J ) 9.77 Hz, 1 H),
1.36 (s, 3 H), 1.77-2.31 (m, 5 H), 2.01 (s, 3 H), 2.67 (dd, J ) 11.96
and 15.02 Hz, 1 H), 2.91 (m, 2 H), 4.19 (dd, J ) 2.20 and 8.71 Hz, 1
H), 6.43 (br s, 1 H, variable positions), 7.11 (d, J ) 8.42 Hz, 2 H),
7.28 (d, J ) 8.42 Hz, 2H). 13C NMR (CDCl3): δ 17.83, 23.83, 25.99,
26.98, 28.95, 36.10, 36.39, 37.81, 39.63, 44.40, 52.00, 75.85, 83.28,
128.52, 130.17, 131.91, 138.95, 174.77. [(1R)-1-Acetamido-2-(1-
[R]23 ) +156.0 (c 0.72, H2O). Anal. Calcd for C10H13BClNO3: C,
D
49.74; H, 5.43; B, 4.48; N, 5.80. Found: C, 49.78; H, 5.11; B, 4.29;
N, 5.98. Diethanolamine [(1S)-1-acetamido-2-(4-chlorophenyl)ethyl]-
boronate (D-6d; 54%), mp 216-218 °C, [R]23D ) +25.0 (c 0.60, CH2-
Cl2). HRMS: calcd for C14H20BClN2O3 310.1256, found 310.1258.
[(1S)-1-Acetamido-2-(1-naphthyl)ethyl]boronic Acid (D-1e). (-)-
Pinanediol (1-naphthylmethyl)boronate (ent-3e; 41%), bp 160-165 °C
(0.2 mmHg), [R]23D -24.2 (c 6.78, toluene). (-)-Pinanediol [(1R)-1-
chloro-2-(1-naphthyl)ethyl]boronate (ent-4e; 81%), obtained as an oil
chlororophenyl)ethyl]boronic acid (L-1d; 79%), mp 140-143 °C, [R]23
D
) -165.8 (c 0.63, H2O). IR (KBr): ν 3700-2800, 1625, 1266, 805
cm-1 1H NMR (D2O): δ 2.11 (s, 3 H), 2.54 (dd, J ) 12.82 and 15.47
.
Hz, 1 H), 2.83 (m, 2 H), 7.26 (d, J ) 8.42 Hz, 2 H), 7.37 (d, J ) 8.42
Hz, 2H). 13C NMR (D2O): δ 17.28, 36.73, 50.12 (br), 129.44, 131.30,
132.34, 140.11, 177.66. HRMS (FAB): calcd for C30H33 B3Cl3N3O6
669.1714, found 669.1755. Diethanolamine [(1R)-1-acetamido-2-(4-
after chromatography on silica gel, [R]23 ) -22.5 (c 2.56, toluene).
D
(-)-Pinanediol [(1S)-1-acetamido-2-(1-naphthylethyl]boronate (ent-5e;
68%), mp 238-240 °C, [R]23 ) +55.7 (c 2.97, CHCl3). [(1S)-1-
D
chlorophenyl)ethyl]boronate (L-6d; 48%), mp 218-221 °C, [R]23
)
D
Acetamido-2-(1-naphthyl)ethyl]boronic acid (D-1e; 47%), mp 168-
172 °C, [R]23D ) +141.4 (c 1.35, CH3OH). HRMS: calcd for C42H42
B3N3O6 717.3353, found 717.3398. Diethanolamine [(1S)-1-acetamido-
-25.4 (c 0.68, CH2Cl2). IR (KBr): ν 3316, 3108, 1637, 1274, 806
cm-1 1H NMR (CDCl3): δ 1.84 (s, 3 H), 2.66-2.80 (m, 3 H), 3.01-
.
2-(1-naphthyl)ethyl]boronate (D-6e; 40%), mp 242-245 °C dec, [R]23
3.16 (m, 3 H), 3.31-3.40 (m, 1 H), 3.76-3.98 (m, 4 H), 5.84 (br d, J
) 5.58 Hz, 1 H), 7.11 (d, J ) 8.42 Hz, 2 H), 7.23 (d, J ) 8.46 Hz, 2
H), 7.51 (br s, 1 H). 13C NMR (CDCl3): δ 23.05, 34.18, 46.76 (br),
50.67, 51.00, 63.05 (double peak), 128.15, 129.75, 131.25, 140.09,
172.01. Anal. Calcd for C14H20BClN2O3: C, 54.14; H, 6.49; B, 3.48.
Found: C, 54.45; H, 6.42; B, 3.45.
D
+130.6 (c 0.61, CH2Cl2). HRMS: calcd for [M + H]+ C18H24BN2O3
327.1883, found 327.1867.
Computational Methods. System Setup. The reference structures
used were those of McPhalen and James4a for the subtilisin Carlsberg-
eglin C complex and of Tsukada and Blow5b for R-chymotrypsin, both
available from the Protein Data Bank28 at Brookhaven National
Laboratory.29 The setup was done with Insight II, version 2.2.0 (Biosym
Technologies, Inc., San Diego, CA). To create initial coordinates for
the minimization of SC, the inhibitor (eglin C) and the three calcium
ions were removed. In the case of CT, the dimeric structure of the
enzyme was split into its individual, independent monomers. Only one
D-Series. The inhibitors D-1a-e were prepared by the same methods
as described above for the L-series. For each target and inhibitor, the
IR and 1H and 13C NMR spectra were identical. Their other properties
were as follows.
[(1S)-1-Acetamidoethyl]boronic Acid (D-1a). (-)-Pinanediol meth-
ylboronate (ent-3a; 41%), bp 60-65 °C (3 mmHg), [R]23D ) -37.2 (c
3.12 CHCl3) (lit.18c bp 85-87 °C (5 mmHg), [R]23546 ) -45.2 (c 3.10
CHCl3)). (-)-Pinanediol [(1S)-1-chloroethyl]boronate (ent-4a; 76%),
bp 58-60 °C (0.1 mmHg), [R]23D ) -32.9 (c 2.06, toluene) (lit.18c bp
80-82 °C (0.2 mmHg)). (-)-Pinanediol [(1S)-1-acetamidoethyl]-
(28) For the subtilisin Carlsberg entry 2SEC, 1.8 Å resolution; for
R-chymotrypsin entry 4CHA, 1.68 Å resolution.
(29) (a) Bernstein, F. C.; Koetzle, T. F.; Williams, J. B.; Meyer, E. F.,
Jr.; Brice, M. D.; Rodgers, J. R.; Kennard, O.; Shimanouchi, T.; Tasumi,
M. J. Mol. Biol. 1977, 112, 535. (b) Abola, E. E.; Bernstein F. C.; Bryant,
S. H.; Koetzle, T. F.; Weng, J. In Crystallographic Databases - Information
Content, Software Systems, Scientific Applications; Allen, F. H., Bergerhoff,
G., Sievers, R., Eds.; Data Commision of the International Union of Crys-
tallography: Bonn/Cambridge/Chester, 1987; p 107.
boronate (ent-5a; 68%), mp 197-198 °C, [R]23 ) -21.24 (c 1.13,
D
CHCl3). [(1S)-1-Acetamidoethyl]boronic acid (D-1a; 89%), mp 185-
188 °C, [R]23D ) +81.3 (c 0.47, H2O). The HRMS was not obtainable
so D-1a was further characterized as diethanolamine [(1S)-1-acetami-

958 J. Am. Chem. Soc., Vol. 118, No. 5, 1996
Martichonok and Jones
of the monomers was used to create the initial coordinates for the
enzyme. Residues Gly12 and Leu13, missing due to poor X-ray
resolution, were added using Insight. Hydrogen atoms were added at
the pH (7.8) used for the kinetic measurements. This protonated all
Lys, Arg, and His residues and the N-terminus and deprotonated the
acids Glu and Asp and the C-terminus on both enzymes. In the
calculations of the boronic acid-enzyme complexes, a tetrahedral
carbon atom was used to mimic the boron atom since, as yet, no force
field parameters have been reported for boron. This approximation
was considered acceptable since only energy differences resulting from
changes remote from boron were being explored. To set up the initial
structure for the energy minimization, the boron-equivalent carbon was
covalently bound to the oxygen of the hydroxyl group of the active
site Ser221 of SC. The X-ray structures of 2-phenethylboronic acid
bound to subtilisin BPN′30 and R-lytic protease (mutant with Met192
replaced by Ala) complexed with (methoxysuccinyl-Ala-Ala-Pro-
phenylalanyl)boronic acid31 were used as the models to guide dockings
of the acetamido acids of Table 1 into the active site of SC. The two
hydroxy groups attached to the boron were oriented such that one
pointed to the oxyanion hole, formed by NH2 of Asp155 and the
backbone NH of Ser221, and the other to His64, which becomes
positively charged after the addition of the proton from Ser221. The
aromatic residue of each inhibitor was positioned in the S1-pocket
(defined by Ser125-Ala129, Ala152-Ser156, and Ile165-Tyr167) in
a manner that avoided all bad van der Waals interactions. To determine
the position of the acetamido group within the S2-pocket, the structure
of SC was superimposed on that of the R-lytic protease complexed
with (methoxysuccinyl-Ala-Ala-Pro-phenylalanyl)boronic acid. This
positioned the acetamido within S2 such that a hydrogen bond was
created between the amide hydrogen of the inhibitor and the carbonyl
oxygen of Ser125.32
A similar docking procedure was used for CT, with the reference
structure used for docking being that of CT complexed with (2-
phenylethyl)boronic acid.5e One of the hydroxyls of the boronic acid
was directed to the oxyanion hole, formed by the backbone NH’s of
Gly193 and Ser195, and the other to the positively charged His57 of
the catalytic triad. The aromatic ring of each inhibitor was positioned
in the S1-pocket (defined by Val213-Thr219, Ser190-Asp194, Ser189,
and Gly226) and the acetamido group within the S2-site such that a
hydrogen bond formed between the amide hydrogen of the inhibitor
and the carbonyl oxygen of Ser214.
Charges on the active site serine and the enzyme-bound boronic acid
for both complexes were generated by single-point MNDO33 /ESP34
calculations (MOPAC 9335 ) and scaled to fit those of the CVFF
(consistent valence force field) library. The overall negative charge
of -1 was mostly on, and distributed between, the oxygen atoms bound
to boron. The boron atom itself was assigned a charge of -0.01. This
model system was solvated in a rectangular box (49 × 47 × 49 Å3) of
water molecules. The total number of water molecules in this system
was 2318. The overall charge of the enzyme-inhibitor complex
resulting from this setup was -1.
molecular dynamics calculations. During the molecular dynamics (MD)
simulations, the whole enzyme, with the exception of a 12 Å radius
region around Ser221, was kept fixed, as were water atoms more than
15 Å away from Ser221. The molecular dynamics simulations were
performed for an initial equilibrium period of 10 ps at 400 K, and then
continued for 20 ps at 400 K, with a time step of 10 fs. The 400 K
temperature of the MD simulations ensured that there was no trapping
in local minima. The molecular dynamics trajectories were animated
using the Analysis module of Insight, and a set of the structures was
selected visually. We assumed that the aromatic ring of all inhibitors
would bind into the S1-pocket of both enzymes. In cases where the
aromatic ring of the inhibitor left the pocket during the molecular
dynamics simulations (for example, as it did for L- and D-1e in
orientation (a) of Figure 1 for SC), the structures were discarded. Each
selected structure was cooled to 300 K by initializing the molecular
dynamics at 300 K and then minimizing, first using steepest descents
until the maximum derivative was less than 5.0 kcal/Å, and then using
a conjugate gradient, until the maximum derivative was less than 0.1
kcal/Å. The results are summarized in Figures 2 and 3. For the strong
hydrogen bonds, the calculated heteroatom distances (Å) were as
follows.
Subtilisin Carlsberg. For L-(R)-1e: boron O(1) to side chain N of
Asn155, 3.09 ( 0.33; to backbone N of Ser221, 3.18 ( 0.28; boron
O(2) to Nꢀ of His64, 2.96 ( 0.31; acetamido N to CO of Ser125, 3.01
( 0.33. For D-(S)-1e: boron O(1) to side chain N of Asn155, 3.33 (
0.46; to backbone N of Ser221, 3.31 ( 0.50; boron O(2) to Nꢀ of His64,
3.86 ( 0.66.
(r)-Chymotrypsin. For D-(S)-1e: boron O(1) to backbone N of
Ser195, 3.21 ( 0.62; to backbone N of Gly193, 3.61 ( 0.56; acetamido
CO to Nꢀ of His57, 3.05 ( 0.31. For L-(R)-1e: boron O(1) to backbone
N of Ser195, 3.12 ( 0.46; to backbone N of Gly193, 3.03 ( 0.37.
Kinetic Measurements. The enzyme kinetics were performed under
steady-state conditions at 25 °C using a pH-stat.2d In order to prevent
any oxidation of the boronic acids, all measurements were done under
argon using water degassed with argon at reflux. For SC, the reference
substrate was N-p-tosyl-L-arginine methyl ester (TAME) and the
following basic procedure was employed. After adjusting the pH to
7.0 with 0.2 M NaOH, 0.01-1.00 mL of the inhibitor solution (1.0 ×
10-5 to 5.0 × 10-1 M in water) was added to the reaction mixture
containing 0.187 M TAME solution (aliquots of 0.4, 0.6, 0.8, 1.0, 2.0,
4.0, and 8.0 mL), 1 M KCl solution (1 mL), and water to bring the
final volume to 10 mL and to give final concentrations of 7.5 × 10-3
to 1.5 × 10-1 M substrate and 10-7 to 5.0 × 10-2 M inhibitor. After
equilibration for 3 min, the pH was adjusted to 7.8 with 0.2 M NaOH
and the reaction initiated by addition of 50 µL of SC stock solution
(4.0 × 10-5 M in 0.01 M phosphate buffer, pH 7.8). The rate of uptake
of 0.2 M NaOH was recorded directly into a PC.
The kinetics data for CT were determined similarly, using the
following stock solutions: 3.8 × 10-3 M N-acetyl-L-tyrosine ethyl ester
(NATEE) as a substrate (aliquots 0.4, 0.6, 0.8, 1.0, 2.0, 4.0, 8.0 mL),
1 M KCl solution (1 mL), inhibitor (1.0 × 10-5 to 5.0 × 10-1 M in
water), and enzyme (4.0 × 10-7 M in 0.001 M HCl). The final
concentration of substrate was 1.5 × 10-4 to 3.0 × 10-3 M. After
initiation of the reaction by addition of 50 µL of CT stock solution,
the rate of the reaction was monitored by uptake of 0.02 M NaOH.
KI values were determined using the Grafit program (Erithacus
Software Ltd., U.K.). All kinetic runs were performed in duplicate at
two different concentrations of boronic acids. The results are recorded
in Table 1.
Energy Minimization. The simulations were performed with the
Discover program, version 2.9.0 (Biosym Technologies, Inc., San Diego,
CA) on a Silicon Graphics 240 GTX computer, using the CVFF.36
A
nonbonded cutoff of 10 Å with a switching function between 7.5 and
9 Å was used. The nonbonded pair list was updated every 20 cycles,
and a dielectric constant of 1 was used in all calculations. The energy
of the system was minimized with respect to all 3N Cartesian
coordinates until the maximum derivative of 0.1 kcal mol-1 Å-1 was
reached. The resulting structure was used as the starting point for the
Acknowledgment. Support from the Natural Sciences and
Engineering Research Council of Canada is gratefully acknowl-
edged.
(30) Matthews, D. A.; Alden, R. A.; Birktoft, J. J.; Freer, S. T.; Kraut,
J. J. Biol. Chem. 1975, 250, 1720.
(31) Bone, R.; Silen, J. L.; Agard, D. A. Entry 1P08 in the Brookhaven
Protein Database, 2.25 Å resolution.
(32) Robertus, J. D.; Alden, R. A.; Birktoft, J. J.; Kraut, J. J.; Powers, J.
C.; Wilcox, P. E. Biochemistry 1972, 11, 2439.
(33) (a) Dewar, M. J. S.; Theil, W. J. Am. Chem. Soc. 1977, 99, 4899.
(b) Martin, J. M. L.; Francois, J. P.; Gijbels, R. J. Comput. Chem. 1991,
12, 52.
(34) Besler, B. H.; Merz, K. M. Jr.; Kollman, P. A. J. Comput. Chem.
1990, 11, 431.
1
Supporting Information Available: Figures showing H
and 13C NMR spectra for L-1b,d,e, L-6b, D-1b,c,e, and D-6a-e
(24 pages). This material is contained in libraries on microfiche,
immediately follows this article in the microfilm version of the
journal, can be ordered from the ACS, and can be downloaded
from the Internet; see any current masthead page for ordering
information and Internet access instructions.
(35) MOPAC 93.00 Manual: J. J. P. Stewart, Fujitsu Limited, Tokyo,
Japan, 1993.
(36) Hagler, A. T.; Osguthorpe, D. J.; Dauber-Osguthorpe; P.; Hempel;
J. C. Science 1985, 227, 1309.
JA952816J