Expanding the utility of proteases in synthesis: Broadening the substrate acceptance in non-coded amide bond formation using chemically modified mutants of subtilisin

Article abstract of DOI:10.1016/S0957-4166(01)00024-6

The strategy of combined site directed mutagenesis and chemical modification creates chemically modified mutants (CMMs) with greatly broadened substrate specificities. We have previously reported that the CMMs of subtilisin Bacillus lentus (SBL) are efficient catalysts for the coupling of both L- and D-amino acids. We now report that these powerful catalysts also allow amide bond formation between a variety of non-coded carboxylic acids, including β-alanine and β-amino homologues of phenylalanine, with both L- and D-amino acid nucleophiles. As a guide to enzyme efficiency, a hydrolysis assay indicating pH change has been employed. CMMs selected by this screen furnished higher yields of coupling products compared to the wild-type enzyme (WT). Furthermore, both WT and CMM enzymes allow highly stereoselective aminolysis of a meso diester with an amino acid amine. These results highlight the utility of CMMs in the efficient formation of non-coded amides as potential peptide isosteres.

Full text of DOI:10.1016/S0957-4166(01)00024-6

TETRAHEDRON:
ASYMMETRY
Pergamon
Tetrahedron: Asymmetry 12 (2001) 249–261
Expanding the utility of proteases in synthesis: broadening the
substrate acceptance in non-coded amide bond formation using
chemically modified mutants of subtilisin
Kanjai Khumtaveeporn,a Astrid Ullmann,a Kazutsugu Matsumoto,b Benjamin G. Davisc,* and
J. Bryan Jonesa,*
aDepartment of Chemistry, University of Toronto, 80 St. George St., Toronto, Ontario, Canada M5S 3H6
bDepartment of Applied Chemistry and Biotechnology, Faculty of Engineering, Fukui University, Bunkyo 3-9-1,
Fukui 910-8507, Japan
cDepartment of Chemistry, University of Durham, South Road, Durham DH1 3LE, UK
Received 19 November 2000; accepted 10 January 2001
Abstract—The strategy of combined site directed mutagenesis and chemical modification creates chemically modified mutants
(CMMs) with greatly broadened substrate specificities. We have previously reported that the CMMs of subtilisin Bacillus lentus
(SBL) are efficient catalysts for the coupling of both
amide bond formation between a variety of non-coded carboxylic acids, including b-alanine and b-amino homologues of
phenylalanine, with both - and -amino acid nucleophiles. As a guide to enzyme efficiency, a hydrolysis assay indicating pH
L- and D-amino acids. We now report that these powerful catalysts also allow
L
D
change has been employed. CMMs selected by this screen furnished higher yields of coupling products compared to the wild-type
enzyme (WT). Furthermore, both WT and CMM enzymes allow highly stereoselective aminolysis of a meso diester with an amino
acid amine. These results highlight the utility of CMMs in the efficient formation of non-coded amides as potential peptide
isosteres. © 2001 Published by Elsevier Science Ltd.
1. Introduction
vantageous solvent specificity.3 Alternatively, substrate
engineering or mimicry4–6 has been employed, but by
definition requires the synthesis of specifically designed
substrates that typically are not readily available.
Impressive protein engineering methods, such as site
directed mutagenesis7,8 or forced evolution9–11 have also
successfully altered the specificities of proteases for use
in synthesis. However, these methods do not routinely
allow the creation of large numbers of catalysts in
sufficient quantities for synthesis.
The use of enzymes as biocatalysts for transformations
in both forward and reverse hydrolyses is now an
attractive and widely accepted synthetic method.1,2
Hydrolytic enzymes operate with high efficiency under
mild conditions that minimise undesirable side reac-
tions and the need for extensive protection regimes. In
particular, serine proteases allow the ready construction
of a variety of acyl linkages including peptide ligation,
esterification and amide bond formation. For example,
their use allows racemisation-free peptide ligation with-
out the need for side chain protection. However, despite
these demonstrated advantages their natural chemo-,
regio- and stereospecificities can sometimes confine
their use in synthesis to a limited range of substrates.
Several strategies have been employed in an attempt to
overcome these restrictions. As alternative catalysts
with typically broader specificities, lipases have been
employed but their lower activities in aqueous solvents
negate this broader substrate specificity with a disad-
To overcome these problems, biocatalysts with broad
specificities for amide synthesis that are readily created,
that may be used in simple solvent systems and that
employ standard readily available substrates are
required. Using the serine protease subtilisin Bacillus
lentus (SBL), we have successfully exploited a combined
site directed mutagenesis and chemical modification
approach for enhancing enzyme activity and tailoring
enzyme specificity (Scheme 1).12–23 This strategy
involves the introduction of a cysteine residue through
site directed mutagenesis and the reaction of its thiol
side chain with chemoselective methanethiosulfonate
reagents24,25 to produce chemically modified mutant
* Corresponding authors. Tel.: 0191 374 3136; fax: 0191 384 4737;
e-mail: bendavis@durham.ac.uk
0957-4166/01/$ - see front matter © 2001 Published by Elsevier Science Ltd.
PII: S0957-4166(01)00024-6

250
K. Khumta6eeporn et al. / Tetrahedron: Asymmetry 12 (2001) 249–261
Scheme 1. The creation of CMM biocatalysts for use in amide bond formation.
(CMM) enzymes (Scheme 1). Since wild-type (WT)
SBL contains no natural cysteines the site of modifica-
tion corresponds only to the site of mutagenesis. The
ease of this method is such that large arrays of different
CMM biocatalysts can be created and screened in a
short space of time.26 In this way, groups intended to
influence the specificity of SBL were introduced to
position 166 at the base of the primary specificity
catalysed reactions by assaying the pH change during
the reaction course has been used by several research
groups30 and was chosen as a suitable method for our
system. The assay developed is based on the mechanism
of serine protease-catalysed ester hydrolysis as shown in
Scheme 2.
Productive binding of the acyl donor 1 leads to the
formation of the acyl-enzyme intermediate 2. Deacyla-
tion of 2 in aqueous media leads to hydrolysis and
generates 3 and a hydronium ion, which may be
detected using a pH indicator. It is the aminolysis of 2
that, in the presence of an amine acyl acceptor, leads to
amide bond formation. Therefore, an assay that mea-
sures formation of 3 gives a valuable indication of
which CMMs successfully form the acyl-enzyme inter-
mediate 2 with a given acyl donor 1. The reaction
conditions for the assay were chosen in such a way that
the degree of colour change of the indicator system was
a qualitative indication of the concentration of 3 that
was formed. Therefore, the pH of the buffering system
was chosen to match the pKa of the indicator. From a
variety of indicators tested, bromothymol blue [pKa=
7.1, pH range 6.0 (yellow)–7.6 (blue)] in phosphate
buffer (pH 7.1) mixed with DMF was selected.
27
determining S1 pocket through the creation of a wide
range of S166C-CMMs.
Such modification of SBL provided CMMs that display
broadened structural as well as stereospecificities when
used as synthetic catalysts. Recently, we reported the
improved synthetic utility of both CMM-catalysed
transesterifications28 and peptide ligations.29 These
reactions also allowed successful incorporation of
D-
amino acid esters as acyl donors and a-branched amino
acid amides as acyl acceptors. This expanded specificity
prompted us to investigate further the application of
CMMs in other challenging coupling reactions with
non-coded acids. We now report the high synthetic
utility of CMMs of SBL in the coupling of representa-
tive esters derived from b-amino acids and chiral car-
boxylic acids as unnatural acyl donors with both
-amino acid acyl acceptors.
L- and
D
Using this screening system with representative acyl
donors 1a–m (Scheme 3), CMMs S166C-S-a–d (Scheme
1) were identified. Their selection corresponded well
with the results of previous detailed analyses of amidase
and esterase activities, using standard substrates suc-
AAPF-pNA and suc-AAPF-SBn, respectively, that by
virtue of their high esterase activity to amidase activity
ratios (E/A) had identified them as excellent candidates
for efficient amide bond formation.20,26
2. Results and discussion
2.1. pH Indicator-based screening of CMMs
A fast screening method was developed to identify
enzymes suitable for amide formation reactions from
our CMM library. Monitoring the progress of enzyme-
Scheme 2. The mechanism of serine protease-catalysed ester hydrolysis.

K. Khumta6eeporn et al. / Tetrahedron: Asymmetry 12 (2001) 249–261
251
2.2. Coupling of 1a–m with glycinamide
employed as substrates for proteases and typically acy-
lases have proved more successful catalysts for mande-
lyl transfer.34,35 The active site spatial requirements of
SBL are such that with this shorter side chain the
CHOH of mandelate would also be required to bind in
the S1 pocket.36 Presumably, the hydrophilicity of the
hydroxyl group prevents such a binding mode in the
hydrophobic S1 pocket. Therefore, we speculated that
protection of this hydroxyl group with hydrophobic
moieties, such as methoxymethyl (MOM) or benzyl
groups, would solve this problem. In all cases it was
found that reaction of the corresponding hydroxyl-pro-
tected acyl donors 1e–h with 4, although slow,37 led to
the formation of the mandelamides 5e–h in yields of up
to 63% (S166C-S-d catalysed coupling of 1e with 4). In
spite of some low absolute yields, all of the CMM-
catalysed reactions of 1e–h showed improved synthetic
utilities compared with WT-catalysed reactions. Inter-
estingly, although SBL-WT and S166C-S-a,b showed
little stereochemical preference for either (S)-1e or (R)-
1f, S166C-S-c,d both showed an approximately two-
fold (S)- over (R)- preference. Consistently lower yields
for the more bulky Bn-protected substrates 1g,h (trace
to 18%) compared with MOM-protected substrates 1e,f
(12–63%) provided further support for a binding mode
in which the CHOR substituent of the mandelate acyl
donors is bound in the S1 pocket, whereas the greater
bulk of the OBn moiety is only partially accommo-
dated. Moreover, no significant (R)- or (S)-stereospe-
cificity was observed for 1g,h with any of the catalysts
tested.
The substrate acceptance of the S1 pocket of our
CMMs S166C-S-a–d was investigated using glyci-
namide 4 as the acyl acceptor and benzyl esters 1a–m of
a wide range of chiral carboxylic acids as the acyl
donors (Scheme 3). In accord with our goal of using
simple solvent systems a 1:1 water:DMF mixture was
used.
The results of these coupling reactions with glycinamide
are shown in Table 1. The natural P1 preference of
SBL-WT is for phenylalanine and typically high yields
result from the use of phenylalanine esters as acyl
donors.29 Benzyl esters of (S)- and (R)-phenyllactic
acid 1a,b were chosen as close analogues of
L- and
D
-phenylalanine, respectively.31–33 As Table 1 illus-
trates, although SBL-WT catalyses the coupling of
esters 1a,b with glycinamide 4 yielding 5a,b in moderate
yields, the use of CMMs engenders a greater than
two-fold increase in yield and S166C-S-a–d are all
superior catalysts. In particular, S166C-S-CH2C6F5
(-b), which is modified with a perfluorobenzyl sub-
stituent in its S1 pocket, allows coupling in a good 69%
yield. Interestingly, and in contrast to SBL-WT and all
the other CMMs, S166C-S-p-CH2C6H4COO− (-c),
which is modified with a benzoate in its S1 pocket,
shows a slight but reversed preference for the (R)-ester
1b (64%) over the (S)-ester 1a (50%).
Shortening the side chain of the acyl donor by one
methylene unit from CH2C6H5 in 1a,b to C6H5 in 1c,d
had a dramatic effect and no coupling products were
observed for either (R)- or (S)-mandelic acid benzyl
esters as acyl donors. This is consistent with the limited
examples of mandelates that have been successfully
b-Amino acids are important sources of secondary
structure in the synthesis of peptide isosteres.38 They
are typically stable to proteolytic degradation and as
Scheme 3. CMM-catalysed coupling of glycinamide 4 with a variety of non-coded carboxylic acid esters 1.

252
K. Khumta6eeporn et al. / Tetrahedron: Asymmetry 12 (2001) 249–261
such are not only potentially useful in vivo pepti-
domimetics, but are also testing substrates for protease-
catalysed peptide ligation.39,40 SBL-catalysed coupling
of acyl donor Z-b-Ala-OBn 1i successfully formed
amide 5i, containing naturally occurring but non-coded
amino acid b-alanine, in moderate yields. CMMs
Table 1. Coupling of non-coded chiral carboxylic acid or b-amino acid esters 1 with glycinamide 4 catalysed by SBL-WT
and SBL-CMMs S166C-S-a–d

K. Khumta6eeporn et al. / Tetrahedron: Asymmetry 12 (2001) 249–261
253
Scheme 4. CMM-catalysed coupling of
D
- and L-alaninamide with a variety of non-coded carboxylic acid esters 1.
+
S166C-SBn (-a) and S166C-SEtNH3
(-d) gave
phenyllactates as acyl donors in coupling reactions
with -Ala-NH2 6 with an approximate two-fold
improved yields of 49 and 45%, respectively, when
used as catalysts compared with SBL-WT (39%).
Despite these positive indications for the coupling of
b-amino acids, the use of acyl donors with an addi-
tional b-benzyl substituent as phenylalanine homo-
logues proved dramatically detrimental and no
reaction of (R)- or (S)-Z-b-benzyl-b-Ala-OEt 1j and
1k41 was observed even after 5 days. The substituent
at the b-position of compounds 1j and 1k seems to be
responsible for the failure of this reaction. Corre-
sponding a-benzyl substituted esters 1l, and 1m, which
were synthesised by slight modification of a literature
procedure,42 proved to be good substrates. SBL-WT
shows a strong stereospecificity for (S)-Boc-a-benzyl-
b-Ala-OBn 1l (51% yield of 5l) over (R)-Boc-a-benzyl-
b-Ala-OBn 1m (9% yield of 5m). The use of CMMs
greatly enhances the synthetic utilities of these reac-
tions and S166C-S-a–d catalyse the formation of 5l in
good yields of 60–79%. Similarly, the syntheses of 5m
were significantly improved with CMMs as catalysts
and up to 40% yield of product was obtained using
L
stereospecificity for (S)- (27% yield of 7a) over (R)-
(12% yield of 7b) that reflects the natural P1 stereospe-
cificity for
L
-phenylalanine over
D-phenylalanine.
However, the stringent P1% stereospecificity of SBL-WT
was demonstrated by the complete absence of coupling
product in these reactions using
D-Ala-NH2 as the acyl
acceptor with either donor 1a,b. In contrast, corre-
sponding CMM-catalysed amide formation was dra-
matically enhanced. In all cases the yields of amides
7a,b and 9a,b were greater than those for SBL-WT.
Indeed, the specificities of S166C-S-b,c were suffi-
ciently broadened that they allowed the synthesis of 7a
from 1a in excellent yields of 88% that were, in fact,
superior to those obtained in glycinamide couplings of
1a. Excitingly, S166C-S-a–d were also capable of
catalysing coupling reactions (which SBL-WT will not)
of both (S)-1a and (R)-1b acyl donors with
D-Ala-
NH2 to give 9a,b, respectively, in low yields of 14–
25%. Reaction of b-amino ester Z-b-Ala-OBn 1i as
the acyl donor allowed coupling to
L
-Ala-NH2 6 in
+
S166C-SEtNH3 (-d). This four-fold broadening of
stereospecificity to accept both (S)- and (R)-isomers as
acyl donors in fair yield closely mirrors that found for
good yields that were similar or even, in the case of
S166C-S-b–d, superior to those found for coupling to
glycinamide 4. Addition of a benzyl group at the a-
carbon of this b-amino acid scaffold gives the more
challenging acyl donors (S)-1l and (R)-1m Boc-a-ben-
zyl-b-Ala-OBn. Little or no coupling was observed in
the reaction of these b-alaninyl esters 1l, and 1m with
L
- and D
-phenylalanine esters.29
2.3. Coupling of 1 with
the specificity of the S%1 pocket
L- and D-alaninamide: probing
D
-Ala-NH2 8 in the presence of SBL-WT, which is in
Having determined that the CMM strategy success-
fully creates catalysts S166C-S-a–d with broadened P1
specificities in the coupling of non-coded acids 1a–m,
we probed the altered specificities of the S%1 pocket in
these same catalysts using a-branched amino acid
alaninamide as an acyl acceptor. In addition to prob-
ing the ability of these catalysts to accept a-branched
line with the strict P%1 stereospecificity observed for the
SBL-WT catalysed coupling of phenyllactates,
although low yields of 7l and 7m (20 and 7%, respec-
tively) were obtained with
in all cases S166C-S-a–d were superior catalysts for
these reactions. Reaction of 1l,m with -Ala-NH2
L
-Ala-NH2 6. Gratifyingly,
L
6
gave yields only slightly lower than those obtained for
glycinamide 4 as an acyl acceptor (64% yield of 7l,
using S166C-S-b as a catalyst and 20% of 7m, using
S166C-S-d). Moreover, the broadening of the P1%
stereospecificity of these S166C CMMs was further
P1% substrates, parallel reactions using both
L
-6 and
D
-8
alaninamide also tested the P%1 stereochemical prefer-
ence (Scheme 4 and Table 2). As representative acyl
donors 1a, 1b, 1i, 1l and 1m were chosen on the basis
of superior yields in couplings with glycinamide.
demonstrated by the successful coupling of
D-Ala-NH2
8 as an acyl acceptor with 1l,m to give 9l in 41% yield
(using S166C-S-b as a catalyst) and 9m in 13% yield
(using S166C-S-c as a catalyst). This demonstrates
that these catalysts tolerate not only a-branched acyl
acceptors in their P%1 subsites but also those with either
(R)- or (S)-configurations at a-stereocentres.
The results of these coupling reactions are shown in
Table 2. In agreement with our previous studies,29
these reactions using Ala-NH2 6,8 as the acyl accep-
tors typically required longer reaction times than for
those using Gly-NH2 4. As for the glycinamide cou-
plings, the WT enzyme accepts both (S)-1a and (R)-1b

254
K. Khumta6eeporn et al. / Tetrahedron: Asymmetry 12 (2001) 249–261
Table 2. Coupling of chiral carboxylic acids or b-amino acids 1 with alaninamide catalysed by SBL-WT and SBL-CMMs
S166C-S-a–d
2.4. Probing the selectivity of CMMs in the aminolysis of
a prochiral non-coded diester
(Scheme 5). Such reactions offer the potential for very
efficient generation of optically active materials and
mono amino acid-glutaramides such as 11 have shown
potent activity as important endopeptidase inhibitors.43
As Table 3 shows, in all cases SBL-catalysed coupling
of the dibenzyl ester of glutaric acid 10 as the acyl
donor and glycinamide 4 as the acyl acceptor proceeded
with good stereoselectivity to yield monoamide 11 in
Following the successful coupling of amino acid acyl
acceptors 4,6,8 with a variety of non-coded acid esters
described in Sections 2.2 and 2.3, we investigated the
potential of CMMs of SBL as catalysts for the selective
amidation of a prochiral meso-diester substrate 10
Scheme 5. CMM-catalysed enantiotopic aminolysis of prochiral meso-diester 10.

K. Khumta6eeporn et al. / Tetrahedron: Asymmetry 12 (2001) 249–261
255
Table 3. CMM-catalysed coupling reactions of 10 with
glycinamide 4
sis and purification were performed on pre-coated
Merck silica gel plates (60 F-254, 250 mm) visualised
1
with UV light or iodine. H and 13C NMR spectra were
Enzymes
% Yield of 11 % E.e. of 11a
% Yield of 12
recorded on a Varian Gemini 200 (200 MHz for 1H and
1
50.3 MHz for 13C) or Unity 400 (400 MHz for H and
WT
S166C-S-b
S166C-S-c
6
11
10
90
91
87
Not determined
22
16
100 MHz for 13C) spectrometer and chemical shifts are
given in ppm (l) using CDCl3, acetone-d6 or DMSO-d6
as an internal standard. High resolution mass spectra
(HRMS) were recorded using Micromass ZAB-SE
(FAB+). Optical rotations were measured with a
Perkin–Elmer 243B polarimeter.
a Column: Chiralcel OJ, hexane/i-PrOH=75/25, flow rate: 0.5 mL/
min.
high enantiomeric excess (87–91%) as determined by
chiral HPLC. Although rather slow and low in yield,
these first examples of selective amidation of prochiral
meso-diesters by proteases clearly demonstrate the
potential of such enzyme catalysed reactions. Further
studies to improve the yield of 11 by enhancing conver-
sion levels44 and reducing competing hydrolysis to
mono-ester 12 are in progress.
4.2. CMM screen using pH indicator
The screen was carried out on a 96 well plate with an
overall volume of 200 mL per well. The assay solutions
were prepared by mixing phosphate buffer (45 mL of a
0.02 M solution, pH 7.1), DMF (125–140 mL), substrate
(5 mL of a 1 M solution in DMF), and bromothymol
blue [45 mL of a 1.9 M solution in phosphate buffer (0.1
M, pH 7.1)] in the well. Enzyme [5–20 mL of a solution
in MES-buffer (1–3 mg/mL enzyme, 10 mmol MES, 1
mM CaCl2, pH 5.8)] was added and all components
were carefully mixed. The enzymes were used in the
same concentrations and were evaluated qualitatively as
good, poor and not working hydrolysis catalysts by
comparing the time needed for colour change. The
enzyme concentrations necessary for a reasonably fast
assay were substrate dependent and had to be tested
individually. The concentrations given can be used as a
guide. Controls without enzyme were employed for
better detection.
3. Conclusions
CMMs S166C-S-a–d of the serine protease subtilisin
Bacillus lentus (SBL) are powerful catalysts in the ami-
dation of important non-coded chiral carboxylic acids
and amino acids. The modification of position 166 at
the base of the primary specificity determining S1
pocket in these enzymes broadens not only P1 structural
and stereospecificities, but also P%1 structural and
stereospecificities. They allowed the coupling of phenyl-
lactates, hydroxyl-protected mandelates and b-amino
acids. Highest yields were obtained for a-benzylcar-
boxylate acyl donors 1a,b,l,m, consistent with the natu-
ral P1 preference of SBL for phenylalanine. Although
much broadened over SBL-WT, the stereochemical
preference of these CMMs typically favours the
configuration of the a-benzyl stereocentre in these
4.3. General procedure for the coupling reaction of acyl
donors and amino acids as the acyl acceptors catalysed
by CMMs
To a solution of acyl donor (0.1 mmol) in DMF (0.4
mL) and water (0.4 mL), glycinamide hydrochloride 4
donors that is homochiral with that of L-phenylalanine.
(0.3 mmol) or
L
-(6) or -(8) alaninamide hydrochloride
D
The dramatically broadened P%1 specificity has allowed
the use of single CMMs to construct all four possible
stereoisomers of both phenyllactylalaninamide (7a,b
and 9a,b) and a-benzyl-b-alaninyl-alaninamide (7l,m
and 9l,m) as examples of potentially important non-
coded peptide isosteres. Furthermore, although low in
yield, SBL-WT and SBL-CMMs catalysed the first
examples of a stereoselective amidation of a prochiral
meso-diester with an amino acid.
(0.2 mmol) and Et3N (0.083–0.125 mL, 0.3–0.4 mmol)
were added, followed by addition of a solution of 1 mg
of active enzyme (0.0037 mmol, 0.037 equiv.), as deter-
mined by titration with PMSF.45 The resulting volume
was 1.0–1.2 mL. The reaction was left stirring at rt for
the period indicated in Tables 1 and 2. In the case of
reaction times longer than 48 h, a solution containing 1
mg more of active enzyme as well as an equal volume
of DMF were added at 48 h. Although activity
decreases over time due to autolysis, variations in rates
of autolysis between CMMs are minimal.29 After the
reaction had finished, the mixture was extracted with
EtOAc and brine and then the organic layer was dried
(MgSO4) and concentrated in vacuo. The residue was
purified by preparative TLC (2–5% MeOH in CH2Cl2).
4. Experimental
4.1. General methods
WT-subtilisin Bacillus lentus and mutant enzyme
S166C were purified and prepared as previously
reported.12,13,20 Protected amino acids were purchased
from Sigma or Bachem and were used as received.
4.4. (S)-Phenyllactic-Gly-NH2 5a
(R)-1j
and
(S)-3-(benzyloxycarbonylamino)-4-
A mixture of (S)-phenyllactic acid (0.4 g, 2.5 mmol),
benzyl alcohol (0.27 g, 2.5 mmol) and p-TsOH (cat.) in
benzene (50 mL) was refluxed for 3 h. The water was
removed from the reaction using a Dean–Stark trap.
phenylbutanoate, 1k, were prepared according to litera-
ture methods.41 All solvents were reagent grade and
distilled prior to use. Thin-layer chromatography analy-

256
K. Khumta6eeporn et al. / Tetrahedron: Asymmetry 12 (2001) 249–261
The reaction was worked up by removing solvent,
diluting with CH2Cl2 and washing with water and
brine. The organic layer was dried (MgSO4) and the
solvent was removed in vacuo. The crude product was
purified by column chromatography (gradient elution
with EtOAc in hexane) to yield 0.55 g (86%) of (S)-ben-
zyl phenyllactate 1a. 1H NMR (CDCl3) l 2.40 (brs, 1H,
OH), 2.99 (dd, J=4.5, 14 Hz, 1H, CHH%CHOH), 3.15
(dd, J=7, 14 Hz, 1H, CHH%CHOH), 4.50 (dd, J=4.5,
7 Hz, 1H, CHOH), 5.21 (s, 2H, OCH2Ph), 7.11–7.52
(m, 10H, 2×Ph); 13C (CDCl3) l 41.0, 67.9, 71.8, 127.3,
128.9, 129.1, 129.2, 130.5, 135.6, 136.7, 174.5. [h]2D7=−
15.8 (c 0.89, MeOH) {lit. enantiomer,46 [h]D25=+13.6 (c
1, MeOH)}.
The mixture was diluted with CH2Cl2 and washed with
water and brine. Purification by column chromatogra-
phy (gradient elution with EtOAc in hexane) provides
(S)-O-MOM-benzyl mandelate 1e (0.32 g, 75%). 1H
NMR (CDCl3) l 3.34 (s, 3H, OCH3), 4.67 (d, J=12
Hz, 1H, OCHH%O), 4.75 (d, J=12 Hz, 1H, OCHH%O),
5.10 (d, J=12 Hz, 1H, OCHH%Ph), 5.17 (d, J=12 Hz,
1H, OCHH%Ph), 5.20 (s, 1H, CHOMOM), 7.18–7.45
(m, 10H, 2×Ph); 13C (CDCl3) l 56.2, 67.0, 76.9, 95.2,
127.5 128.0, 128.4, 128.6, 128.8, 128.9, 135.6, 136.1,
170.8. [h]2D6=+50.0 (c 1.01, MeOH).
Using the general procedure, 5e was obtained in the
1
yield indicated in Table 1. H NMR (CDCl3) l 3.35 (s,
3H, OCH3), 3.92 (dd, J=6, 12 Hz, 1H, CON-
HCHH%CONH2), 4.00 (dd, J=4, 12 Hz, 1H, CON-
HCHH%CONH2), 4.62 (d, J=12 Hz, 1H, OCHH%O),
4.70 (d, J=12 Hz, 1H, OCHH%O), 5.07 (s, 1H,
OCHPh), 5.63, 6.41, 7.57 (3×brs, NH, NH2), 7.26–7.40
(m, 5H, Ph); 13C (CDCl3) l 42.5, 56.3, 78.0, 94.7, 127.5,
128.8, 128.9, 136.7, 170.9, 171.5. HRMS (MH+) calcd
for C12H16N2O4, 253.1188; found 253.1184. [h]2D4=+
56.7 (c 1.78, MeOH).
The coupling reaction of (S)-benzyl phenyllactate 1a
and glycinamide hydrochloride was carried out using
the general procedure and 5a was obtained in the yield
1
as indicated in Table 1. H NMR (DMSO-d6) l 2.68
(dd, J=8, 13 Hz, 1H, CHH%CHOH), 2.99 (dd, J=4, 13
Hz, 1H, CHH%CHOH), 3.62 (dd, J=5, 14 Hz, 1H,
CONHCHH%CONH2), 3.70 (dd, J=7, 14 Hz, 1H,
CONHCHH%CONH2), 4.10 (m, 1H, CHOH), 5.74,
7.11 (2×brs, 3H, NH2, OH), 7.16–7.28 (m, 5H, Ph),
7.95 (m, 1H, NH); 13C (acetone-d6) l 41.6, 42.7, 73.8,
127.0, 128.9, 130.5, 139.5, 171.6, 174.3. HRMS (MH+)
calcd for C11H14N2O3, 223.1083; found 223.1100.
[h]2D6=+72.1 (c 0.19, MeOH).
4.7. (R)-O-MOM-Mandelic-Gly-NH2 5f
(R)-Benzyl mandelate 1d was prepared in the same
manner as (S)-benzyl mandelate 1c from (R)-mandelic
1
4.5. (R)-Phenyllactic-Gly-NH2 5b
acid (41% yield). H and 13C data were identical to 1c.
[h]2D4=−49.2 (c 1.51, MeOH).
(R)-Benzyl phenyllactate 1b was prepared in the same
manner as (S)-benzyl phenyllactate from (R)-phenyllac-
(R)-O-MOM-benzyl mandelate 1f (75% yield) was pre-
pared in the same manner as (S)-O-MOM-benzyl man-
delate 1e. 1H and 13C data were identical to 1e.
[h]2D4=−49.2 (c 1.4, MeOH).
1
tic acid (96% yield). H and 13C data were identical to
(S)-benzyl phenyllactate. [h]2D6=+16.2 (c 1.0, MeOH)
{lit.,46 [h]2D5=+13.6 (c 1, MeOH)}.
Using the general procedure, 5b was obtained in the
Using the general procedure, 5f was obtained in the
1
1
yield as indicated in Table 1. H NMR and 13C data
yield as indicated in Table 1. H NMR and 13C data
were identical to 5a. HRMS (MH+) calcd for
C11H14N2O3, 223.1083; found 223.1097. [h]2D6=−72.9 (c
0.23, MeOH).
were identical to 5e. HRMS (MH+) calcd for
C12H16N2O4, 253.1188; found 253.1191. [h]2D6=−57.6 (c
1.54, MeOH).
4.6. (S)-O-MOM-Mandelic-Gly-NH2 5e
4.8. (S)-O-Bn-Mandelic-Gly-NH2 5g
(S)-Benzyl mandelate 1c was prepared in the same
manner as 1a from (S)-mandelic acid (1.52 g, 10 mmol),
benzyl alcohol (1.08 g, 10 mmol) and p-TsOH (cat.) in
benzene (50 mL). The crude product was purified by
column chromatography (10% EtOAc in hexane) to
To a solution of (S)-benzyl mandelate 1c (0.367 g, 1.5
mmol) in THF (20 mL) at 0°C, NaH (80% in oil, 0.023
g, 0.75 mmol) was added. The reaction mixture was
stirred for 15 min, then a solution of benzyl bromide
(0.18 mL, 1.5 mmol) in THF (5 mL) was added. After
stirring at rt for 5 h the reaction was worked up by
adding water and extracting with CH2Cl2. The organic
layer was washed with brine, dried over MgSO4 and the
solvent was removed in vacuo. The crude product
was purified by column chromatography (0.4% EtOAc
in hexane) to yield (S)-O-Bn-benzyl mandelate 1g49
(0.12 g, 49%). 1H NMR (CDCl3) l 4.59 (br s, 2H,
OCH2Ph), 4.97 (s, 1H, OCHPh), 5.12 (d, J=12 Hz,
1H, OCHH%Ph), 5.19 (d, J=12 Hz, 1H, OCHH%Ph),
7.20–7.47 (m, 15H, 3×Ph); 13C (CDCl3) l 66.9, 71.3,
79.7, 127.6, 128.1, 128.2, 128.4, 128.6, 128.7, 128.8,
128.9, 135.6, 136.3, 137.2, 170.8. [h]2D6=+19.0 (c 0.83,
MeOH).
1
obtain pure (S)-benzyl mandelate 1c47 (1 g, 41%). H
NMR (CDCl3) l 3.58 (d, J=7 Hz, 1H, OH), 5.14 (dd,
J=12 Hz, 1H, OCHH%Ph), 5.24 (d, J=7 Hz, 1H,
CHOH), 5.25 (d, J=12 Hz, 1H, OCHH%Ph), 7.20–7.45
(m, 10H, 2×Ph); 13C (CDCl3) l 67.8, 73.1, 126.7, 128.1,
128.5, 128.6, 128.7, 128.8, 135.1, 138.3, 173.6. [h]2D6=+
49.2 (c 0.455, MeOH) {lit.,48 [h]D23=+55.7 (c 1,
CHCl3)}.
To the mixture of 1c (0.367 g, 1.5 mmol) and i-Pr2NEt
(0.52 mL, 3 mmol) in CH2Cl2 (10 mL) at 0°C,
ClCH2OMe (0.23 mL, 3 mmol) was added. The reac-
tion was then warmed up to rt and was stirred for 24 h.

K. Khumta6eeporn et al. / Tetrahedron: Asymmetry 12 (2001) 249–261
257
Using the general procedure, (S)-O-Bn-mandelic-Gly-
4.11. (2S)-3-[(t-Butyloxycarbonyl)amino]-2-benzyl
propanoic-Gly-NH2 5l42
NH2 5g was obtained in the yield as indicated in Table
1. 1H NMR (CDCl3) l 3.92 (br d, 2H, J=7 Hz,
CH2CONH2), 4.43 (d, J=11 Hz, 1H, OCHH%Ph), 4.58
(d, J=11 Hz, 1H, OCHH%Ph), 4.85 (s, 1H, OCHPh),
5.55, 5.99 (brs×2, 1H×2, NH2), 7.23–7.47 (m, 11H,
2×Ph, NH); 13C NMR (CDCl3) l 42.7, 71.4, 81.0, 127.3,
128.3, 128.7, 128.8, 128.9, 129.0, 136.7, 136.8, 171.1,
171.6. HRMS (MH+) calcd for C17H18N2O3, 299.1391;
found 299.1391. [h]2D4=+6.8 (c 0.78, MeOH).
Dimethyl malonate (11.9 g, 90 mmol) in THF (120 mL)
was added to a stirred suspension of NaH (80% in oil,
2.7 g, 90 mmol) in THF (60 mL) at 0°C. After warming
to rt and stirring for 0.5 h, a solution of benzyl bromide
(9.5 mL, 80 mmol) in THF (120 mL) was added. The
reaction was stirred for 2 h and then quenched with satd
NH4Cl. The organic layer was separated and the
aqueous phase was extracted with EtOAc (2×50 mL).
The combined organic layers were dried (MgSO4) and
the solvent was removed in vacuo. Purification by
column chromatography (gradient elution with EtOAc
in hexane) furnished 15.4 g (82%) of 2-benzyl dimethyl-
malonate. 2-Benzyl dimethylmalonate (5.4 g, 24.3
mmol) in THF (20 mL) was added to a stirred suspen-
sion of LiAlH4 (0.95 g, 25 mmol) in THF (100 mL) at
0°C. The mixture was stirred at rt for 3 h and quenched
by adding satd NaSO4 until all the suspension solid
became white. The solid was filtered out and the organic
layer was dried (MgSO4). The solvent was removed in
vacuo to yield 3.81 g (95%) of 2-benzyl-1,3-propanediol,
which was pure by NMR and was used in the next step
4.9. (R)-O-Bn-Mandelic-Gly-NH2 5h
(R)-O-Bn-Benzyl mandelate 1h (60% yield) was pre-
pared according to the procedure described for (S)-O-
1
Bn-benzyl mandelate 1g. The H and 13C NMR were
identical to 1g. [h]2D4=−18.3 (c 1.53, MeOH).
Using the general procedure, 5h was obtained in the
1
yield as indicated in Table 1. The H and 13C NMR
were identical to the 5g. HRMS (MH+) calcd for
C17H18N2O3, 299.1391; found 299.1386. [h]2D4=−6.1 (c
1.29, MeOH).
1
without further purification. H NMR (CDCl3) l 2.05
(m, 1H, CHCH2Ph), 2.61 (d, J=7 Hz, 2H, CH2Ph),
3.67 (dd, J=7, 11 Hz, 2H, CH2OH), 3.80 (dd, J=4, 11
Hz, 2H, CH2OH), 7.17–7.29 (m, 5H, Ph); 13C (CDCl3)
l 34.4, 44.0, 65.9, 126.3, 128.6, 129.1, 140.0.
4.10. N-Z-b-Alaninyl-Gly-NH2 5i50
To a solution of b-alanine (1.8 g, 20 mmol) in 2N
NaOH (10 mL, 20 mmol) at 0°C, benzyl chloroformate
(3.16 mL, 22 mmol) and 2N NaOH (11 mL) were added
alternatively. After the addition was finished, the reac-
tion was stirred at rt for a further 0.5 h. The mixture
was then worked up by washing with ether (3×50 mL).
The aqueous layer was acidified with 2.5N HCl and
extracted with CH2Cl2 (3×50 mL). The combined
organic layer was dried (MgSO4) and the solvent was
removed in vacuo. The crude product was used without
further purification.
2-Benzyl-1,3-propanediol (1.84 g, 11 mmol) was treated
with Ac2O (4 mL) and Et3N (4 mL) and DMAP (0.244
g, 2 mmol) at rt for 2 h. The mixture was diluted with
1N HCl and extracted with CH2Cl2. The solution was
washed with satd NaHCO3, dried (MgSO4) and the
solvent removed in vacuo. The crude product was
purified by column chromatography (gradient elution
with EtOAc in hexane) to give 2.4 g (100%) of the
1
corresponding diacetate. H NMR (CDCl3) l 2.04 (s,
6H, 2×CH3), 2.32 (m, 1H, CHCH2Ph), 2.68 (d, J=7 Hz,
2H, CH2Ph), 4.00 (dd, J=6, 11 Hz, 2H, CH2OAc), 4.04
(dd, J=4, 11 Hz, 2H, CH2OAc), 7.13–7.30 (m, 5H, Ph);
13C (CDCl3) l 21.0, 34.8, 39.2, 63.9, 126.6, 128.7, 129.1,
138.8, 171.1.
The esterification of N-Z-b-Ala was carried out accord-
ing to the procedure used for the synthesis of (S)-benzyl
phenyllactate 1a, using N-Z-b-Ala (2.2 g, 10 mmol),
benzyl alcohol (1.18 g, 10 mmol) and p-TsOH (cat.) in
benzene (100 mL). Purification by column chromatogra-
phy furnished 2.1 g (67%) of N-Z-b-Ala-OBn 1i; mp
The diacetate (2.4 g, 10 mmol) was suspended in a 70:30
mixture (100 mL) of 0.1 M phosphate buffer (pH 7.0)
and i-Pr2O. Lipase PS-30 (Amano) (600 mg) was added
at rt and the mixture was stirred overnight. The mixture
was filtered and extracted with CH2Cl2. The extract was
washed with brine, dried (MgSO4) and the solvent was
removed in vacuo. The crude product was separated
from unreacted starting material by column chromatog-
raphy (gradient elution with EtOAc in hexane) to yield
1
22°C [lit.51 mp 22°C]. H NMR (CDCl3) l 2.59 (t, J=6
Hz, 2H, CH2CH2CO), 3.49 (q, J=6 Hz, 2H,
NHCH2CH2), 5.08, 5.12 (2×s, 2H×2, 2×OCH2Ph), 5.30
(br m, 1H, NH) 7.24–7.37 (m, 10H, 2×Ph); 13C (CDCl3)
l 34.6, 36.7, 66.7, 66.9, 127.7, 128.2, 128.3, 128.4, 128.5,
128.6, 135.8, 136.3, 156.4, 172.2.
Using the general procedure, 5i was obtained in the
1
1
1.1 g (58%) of (2S)-benzyl-1-acetoxypropane-3-ol. H
yield as indicated in Table 1. H NMR (DMSO-d6+
NMR (CDCl3) l 2.08 (s, 3H, CH3), 2.11 (m, 1H,
CHCH2Ph), 2.62 (m, 2H, CH2Ph), 3.49 (dd, J=7, 12
Hz, 1H, CHHOH), 3.59 (dd, J=4, 11 Hz, 1H,
CHHOH), 4.06 (dd, J=7, 12 Hz, 1H, CHHOAc), 4.17
(dd, J=7, 11 Hz, 1H, CHHOAc), 7.16–7.30 (m, 5H,
Ph); 13C (CDCl3) l 21.1, 34.4, 42.6, 62.1, 64.1, 126.4,
128.7, 129.2, 139.5, 171.9. [h]2D4=−26.2 (c 0.98, MeOH).
CDCl3) l 2.35 (t, J=7 Hz, 2H, CH2CH2CO), 3.33 (q,
J=7 Hz, 2H, NHCH2CH2), 3.72 (d, J=6 Hz, 2H,
CH2CONH2), 5.01 (s, 2H, CH2OPh), 7.05, 7.23, 8.08
(3×m, 3×1H, NH, NH2), 7.30–7.40 (m, 5H, Ph); 13C
(DMSO-d6+CDCl3) l 35.6, 37.1, 41.9, 65.3, 127.7,
128.4, 137.2, 156.1, 170.7, 171.1. HRMS (MH+) calcd
for C13H17N3O4, 280.1297; found 280.1294.

258
K. Khumta6eeporn et al. / Tetrahedron: Asymmetry 12 (2001) 249–261
HN3 [generated by NaN3 (1.5 g, 23 mmol) and 50%
H2SO4 (10 mL) in benzene (15 mL)] and diethyl azodi-
carboxylate (DEAD) (1.2 mL, 6 mmol) was added to a
solution of (2S)-benzyl-1-acetoxypropane-3-ol (1 g,
5.15 mmol) and PPh3 (1.6 g, 6 mmol) in THF (50 mL)
at −35°C. The mixture was stirred at this temperature
for 0.5 h and then at rt for 2 h. After evaporation the
residue was dissolved in EtOAc (50 mL) and di-t-butyl
dicarbonate (1.13 g, 6 mmol) and Pd-C (250 mg) were
added. The reaction mixture was hydrogenated at rt
and at atmospheric pressure overnight. The catalyst
was removed by filtering through Celite. The solvent
was removed and the crude product was purified using
column chromatography (gradient elution with EtOAc
in hexane) to yield 1.0 g (70%) of (2S)-2-benzyl-3-[(t-
butyloxycarbonyl)amino]propyl acetate. 1H NMR
(CDCl3) l 1.41 (s, 9H, t-Bu), 2.04 (s, 3H, CH3), 2.08
(m, 1H, CHCH2Ph), 2.65 (m, 2H, CH2Ph), 3.08 (m,
1H, CHHN), 3.22 (m, 1H, CHHN), 3.90 (dd, J=4, 11
Hz, 1H, CHHOAc), 4.10 (dd, J=6, 11 Hz, 1H,
CHHOAc), 7.16–7.32 (m, 5H, Ph); 13C (CDCl3) l 21.1,
28.5, 35.7, 40.5, 41.7, 64.2, 79.5, 126.5, 128.7, 129.2,
139.3, 156.2, 171.4. [h]2D6=−6.2 (c 0.79, MeOH).
overnight. The reaction was worked up by extraction
with water and CH2Cl2. The organic layer was dried
(MgSO4) and the solvent was removed in vacuo. The
crude product was purified by column chromatography
(gradient elution with EtOAc in hexane) to obtain
benzyl
(2S)-3-[(t-butyloxycarbonyl)amino]-2-benzyl-
1
propanoate 1l (0.29 g, 70%). H NMR (CDCl3) l 1.42
(s, 9H, t-Bu), 2.82–3.01 (m, 3H, CH2Ph, CHCO), 3.28
(m, 1H, CHHN), 3.38 (m, 1H, CHHN), 4.83 (m, 1H,
NH), 5.06 (s, 2H, OCH2Ph), 7.11–7.35 (m, 10H, 2×Ph);
13C (CDCl3) l 28.5, 36.1, 41.9, 47.6, 66.6, 79.6, 126.7,
128.3, 128.4, 128.7, 129.0, 135.8, 138.3, 155.9, 174.3.
[h]2D4=−7.2 (c 0.43, MeOH).
Using the general procedure for CMM catalysed cou-
pling reactions, reaction of benzyl (2S)-3-[(t-butyloxy-
carbonyl)amino]-2-benzyl-propanoate 1l with glycin-
amide hydrochloride gave 5l in the yield as indicated in
1
Table 1. H NMR (DMSO-d6) l 1.35 (s, 9H, t-Bu),
2.60–2.80 (m, 3H, CH2Ph, CHCO), 3.06 (m, 2H,
CH2N), 3.62 (m, 2H, CH2CONH2), 6.68, 7.00, 8.01
(m×3, 1H×3, NH, NH2), 7.05–7.27 (m, 5H, Ph); 13C
(DMSO-d6) l 28.3, 35.5, 40.1, 41.9, 47.5, 77.8, 126.1,
128.2, 128.9, 139.6, 155.6, 171.1, 173.3. HRMS (MH+)
calcd for C17H25N3O4, 336.1917; found 336.1915.
[h]2D5=−3.8 (c 0.16, MeOH).
K2CO3 (0.94 g, 7.2 mmol) was added to a solution of
(2S)-2-benzyl-3-[(t-butyloxycarbonyl)amino]propyl ace-
tate (1 g, 3.6 mmol) in 5% aq. MeOH (50 mL). The
reaction mixture was stirred at rt for 2 h and then
worked up by extraction with water and CH2Cl2 to
obtain the crude 0.56 g (62%) of (2S)-2-benzyl-3-[(t-
butyloxycarbonyl)amino]propanol. The product was
4.12. (2R)-3-[(t-Butyloxycarbonyl)amino]-2-benzyl
propanoic-Gly-NH2 5m
A mixture of 2-benzyl-1,3-propanediol (0.92 g, 5.5
mmol), vinyl acetate (0.52 mL, 11 mmol) and lipase
PS-30 (Amano) (0.92 g, 1 g per 1 g of the substrate) in
THF (25 mL) was stirred at rt for 2 h. The reaction was
then filtered and the solvent was removed to yield 1 g
(94%) of 2(R)-benzyl-1-acetoxypropane-3-ol. [h]2D6=
+25.4 (c 0.54, MeOH); {lit.,52 [h]D20=+27.7 (c 1.3,
CHCl3); lit.,53 [h]D23=+20.6 (c 1.0, CHCl3)}.
1
used in the next step without further purification. H
NMR (CDCl3) l 1.46 (s, 9H, t-Bu), 1.91 (m, 1H,
CHCH2Ph), 2.49 (dd, J=8, 14 Hz, 1H, CHHPh), 2.62
(dd, J=7, 13 Hz, 1H, CHHPh), 3.08 (dd, J=7, 14 Hz,
1H, CHHN), 3.22 (dd, J=4, 14 Hz, 1H, CHHN), 3.41
(dd, J=4, 11 Hz, 1H, CHHOH), 3.58 (dd, J=7, 11 Hz,
1H, CHHOH), 7.19–7.41 (m, 5H, Ph); 13C (CDCl3) l
28.5, 35.5, 40.4, 43.7, 62.0, 80.1, 126.3, 128.6, 129.1,
140.1, 157.8. [h]D26=−29.8 (c 0.43, MeOH).
Benzyl-2(R)-3-[(t-butyloxycarbonyl)amino]-2-benzyl
propanoate 1m was prepared from (2R)-benzyl-1-acet-
oxypropane-3-ol in the same manner as for 1l from
Jones reagent (5 mL) was added to a solution of
(2S)-2-benzyl-3-[(t-butyloxycarbonyl)amino]propanol
(0.56 g, 2.25 mmol) in acetone (45 mL) at 0°C. The
reaction mixture was stirred at rt for 1 h, then
quenched with i-PrOH until the colour of the reaction
mixture turned to green. After dilution with water and
extraction with CH2Cl2 (3×50 mL), the combined
organic layers were dried (MgSO4) and the solvent was
removed in vacuo to obtain crude (2S)-2-benzyl-3-[(t-
butyloxycarbonyl)amino]propanoic acid (0.54 g, 90%).
1H NMR (CDCl3) l 1.42 (s, 9H, t-Bu), 2.41–3.60 (m,
5H, CH2Ph, CH2N, CHCH2Ph), 7.18–7.38 (m, 5H, Ph),
9.21 (brs, 1H, COOH).
1
2(S)-benzyl-1-acetoxypropane-3-ol. H NMR and 13C
data were identical to 1l. [h]2D4=+7.6 (c 2.11, MeOH).
Using the general procedure for CMM catalysed cou-
pling reactions, reaction of 1m with glycinamide hydro-
chloride gave 5m in the yield as indicated in Table 1. 1H
NMR and 13C data were identical to 5l. HRMS (MH+)
calcd for C17H25N3O4, 336.1917; found 336.1921.
[h]2D6=+2.9 (c 0.70, MeOH).
4.13. (S)-Phenyllactic-L-Ala-NH2 7a
Using the general procedure for CMM-catalysed cou-
(2S)-3-[(t-Butyloxycarbonyl)amino]propanoic
acid
pling reactions, reaction of (S)-benzyl phenylacetate 1a
(0.29 g, 1.1 mmol) was dissolved in DMF (20 mL) and
then CsCO3 (0.39 g, 1.21 mmol) was added. The mix-
ture was evaporated in vacuo to dryness. DMF (20 mL)
was added to the remaining solid and the solvent was
removed again in vacuo. The residue was dissolved in
DMF (20 mL) and treated with benzyl bromide (0.15
mL, 1.21 mmol). The mixture was stirred at rt
with L-alaninamide hydrochloride 6 gave 7a in the yield
as indicated in Table 2. 1H NMR (acetone-d6) l 1.38 (d,
J=7 Hz, 3H, CH3), 2.81 (dd, J=8, 12 Hz, 1H, CHH-
CHOH), 3.20 (dd, J=4, 12 Hz, 1H, CHH%CHOH),
4.25 (br dd, J=4, 8 Hz, 1H, CHOH), 4.42 (quin., J=7
Hz, 1H, CHCONH2), 4.83, 6.48, 6.87, 7.61 (4×brs, NH,
NH2, OH), 7.19–7.35 (m, 5H, Ph); 13C (acetone-d6) l

K. Khumta6eeporn et al. / Tetrahedron: Asymmetry 12 (2001) 249–261
259
19.1, 41.6, 48.8, 73.6, 127.1, 128.9, 130.5, 139.5, 173.4,
4.19. (2S)-3-[(t-Butyloxycarbonyl)amino]-2-benzyl
propanoic- -Ala-NH2 9l
174.7. HRMS (MH+) calcd for C12H16N2O3, 237.1239;
D
found 237.1231. [h]2D7=−55.1 (c 1.65, MeOH).
Using the general procedure, 9l was obtained in the
1
yield as indicated in Table 1. H NMR (DMSO-d6) l
4.14. (S)-Phenyllactic-D-Ala-NH2 9a
1.10 (d, J=7 Hz, 3H, CH3CH), 1.38 (s, 9H, t-Bu),
2.55–3.13 (m, 5H, CH2Ph, CHCO, CH2N), 4.01 (quin,
J=7 Hz, 1H, CHCONH2), 6.95, 6.98, 7.04, 7.80 (4×
brs, 4H, NH, NHBoc, NH2), 7.15–7.27 (m, 5H, Ph);
13C (DMSO-d6) l 18.3, 28.0, 35.0, 41.0, 47.0, 48.0, 77.9,
126.0, 128.2, 128.8, 139.5, 155.4, 172.4, 174.1. HRMS
(MH+) calcd for C18H27N3O4, 350.2073; found
350.2075. [h]2D4=−9.4 (c 0.47, MeOH).
Using the general procedure, 9a was obtained in the
1
yield as indicated in Table 2. H NMR (DMSO-d6) l
1.18 (d, J=7 Hz, 3H, CH3), 2.73 (dd, J=8, 14 Hz, 1H,
CHHCHOH), 2.98 (dd, J=4, 14 Hz, 1H, CHH-
CHOH), 4.14 (ddd, J=4, 6, 8 Hz, 1H, CHOH), 4.22
(quin, J=7 Hz, 1H, CHCONH2), 5.79 (d, OH, J=6
Hz), 7.15, 7.42, 7.61 (3×brs, NH, NH2, OH), 7.20–7.38
(m, 5H, Ph); 13C (acetone-d6) l 19.6, 40.9, 48.9, 73.8,
127.4, 129.3, 130.2, 139.9, 173.0, 175.8. HRMS (MH+)
calcd for C12H16N2O3, 237.1239; found 237.1241.
[h]2D7=−52.5 (c 1.19, MeOH).
4.20. (2R)-3-[(t-Butyloxycarbonyl)amino]-2-benzyl
propanoic-L-Ala-NH2 7m
Using the general procedure, 9m was obtained in the
1
yield as indicated in Table 1. H NMR and 13C data
4.15. (R)-Phenyllactic-
L
-Ala-NH2 7b
were identical to 9l. HRMS (MH+) calcd for
C18H27N3O4, 350.2073; found 350.2066. [h]2D4=+9.7 (c
0.64, MeOH).
Using the general procedure, 7b was obtained in the
1
yield as indicated in Table 2. H NMR (DMSO-d6) and
13C (acetone-d6) data were identical to 9a. HRMS
(MH+) calcd for C12H16N2O3, 237.1239; found
237.1229. [h]2D7=+52.6 (c 1.25, MeOH).
4.21. (2R)-3-[(t-Butyloxycarbonyl)amino]-2-benzyl
propanoic-D-Ala-NH2 9m
Using the general procedure, 9m was obtained in the
1
yield as indicated in Table 1. H NMR and 13C data
4.16. (R)-Phenyllactic-D-Ala-NH2 9b
were identical to 7l. HRMS (MH+) calcd for
C18H27N3O4, 350.2073; found 350.2069. [h]2D4=+11.7 (c
1.33, MeOH).
Using the general procedure, 9b was obtained in the
1
yield as indicated in Table 2. H NMR (acetone-d6) and
13C (acetone-d6) data were identical to 7a. HRMS
(MH+) calcd for C12H16N2O3, 237.1239; found
237.1227. [h]2D7=+54.0 (c 0.71, MeOH).
4.22. Monobenzyl 3-methylglutaryl-Gly-NH2 11
To a suspension of NaH (80% in oil, 843 mg, 28.1
mmol) in THF (10 mL), BnOH (2.91 mL, 28.1 mmol)
was added at rt under nitrogen atmosphere and the
resulting mixture was stirred for 15 min. After addition
of a solution of 3-methylglutaric anhydride (3.00 g, 23.4
mmol) in THF (40 mL) at rt the mixture was stirred
overnight. The reaction was quenched with water, and
the pH was adjusted to ca. 8 with 0.1 M NaOH. After
extraction with EtOAc, the pH of the water layer was
adjusted with 1 M HCl, the solution was saturated with
NaCl and extracted with EtOAc. After evaporation,
crude monobenzyl 3-methylglutarate was obtained as a
colourless oil. This compound was used in the next step
4.17. N-Z-b-Alaninyl-L-Ala-NH2 7i
Using the general procedure, 7i was obtained in the
1
yield as indicated in Table 2. H NMR (DMSO-d6) l
1.16 (d, J=7 Hz, 3H, CH3), 2.29 (m, 2H, CH2CH2CO),
3.19 (q, J=7 Hz, 2H, NHCH2CH2), 4.18 (quin, J=7
Hz, 1H, CH3CHCONH2), 5.00 (s, 2H, CH2OPh), 6.96,
(brs, 2H, NH2), 7.21 (t, J=7 Hz, 1H, NHZ), 7.29–7.36
(m, 5H, Ph), 8.01 (d, J=7 Hz, 1H, NH); 13C (DMSO-
d6) l 18.3, 35.5, 37.1, 48.0, 65.2, 127.8, 127.9, 128.4,
137.3, 156.0, 170.0, 174.4. HRMS (MH+) calcd for
C14H19N3O4, 294.1454; found 294.1445.
1
without further purification. H NMR (CDCl3) l 1.05
(d, J=5.5 Hz, 3H, CH3), 2.22–2.40 (m, 2H, CH2CO2),
2.40–2.58 (m, 2H, CH2CO), 2.46–2.52 (m, 1H,
CHCH3), 5.12 (s, 2H, CH2Ph), 7.30–7.48 (m, 5H, Ph),
10.5 (brs, 1H, CO2H).
4.18. (2S)-3-[(t-Butyloxycarbonyl)amino]-2-benzyl
propanoic-L-Ala-NH2 7l
Using the general procedure, 7l was obtained in the
1
yield as indicated in Table 2. H NMR (DMSO-d6) l
To a solution of monobenzyl 3-methylglutarate and
BnOH (2.23 g, 20.6 mmol) in benzene (50 mL) a
catalytic amount of p-TsOH was added, and the mix-
ture was refluxed using a Dean–Stark trap and stirred
overnight. The reaction was quenched with satd
NaHCO3, extracted with EtOAc, washed with brine
and dried over MgSO4. After evaporation the residue
was purified by column chromatography on silica gel
(hexane:EtOAc, 9:1) to afford dibenzyl 3-methylglu-
1.16 (d, J=7 Hz, 3H, CH3CH), 1.37 (s, 9H, t-Bu),
2.57–3.16 (m, 5H, CH2Ph, CHCO, CH2N), 4.18 (quin,
J=7 Hz, 1H, CHCONH2), 6.70 (t, J=7 Hz, 1H,
NHBoc), 6.91, 6.95 (2×brs, 2H, NH2), 7.18–7.26 (m,
5H, Ph), 7.82 (d, J=7 Hz, 1H, NH); 13C (DMSO-d6) l
18.4, 28.3, 35.4, 42.1, 47.2, 48.0, 77.7, 126.1, 128.2,
128.9, 139.6, 155.5, 172.4, 174.2. HRMS (MH+) calcd
for C18H27N3O4, 350.2073; found 350.2082. [h]2D4=
−11.3 (c 0.46, MeOH).
1
tarate 10 as a colourless oil (5.34 g, 70%). H NMR

260
K. Khumta6eeporn et al. / Tetrahedron: Asymmetry 12 (2001) 249–261
(CDCl3) l 1.02 (d, J=6.5 Hz, 3H, CH3), 2.31 (dd,
J=7.5, 15.0 Hz, 2H, CH2CO2), 2.44–2.60 (m, 1H,
CHCH3), 2.45 (dd, J=6.0, 15.0 Hz, 2H, CH2CO2), 5.12
(s, 4H, CH2Ph), 7.32–7.41 (m, 10H, Ph); 13C NMR
(CDCl3) l 20.4, 28.0, 41.3, 66.7, 128.7, 128.9, 129.1,
136.5, 172.6.
Mitchinson, C.; Jones, J. B. J. Am. Chem. Soc. 1997, 119,
5265–5266.
15. DeSantis, G.; Berglund, P.; Stabile, M. R.; Gold, M.;
Jones, J. B. Biochemistry 1998, 37, 5968–5973.
16. DeSantis, G.; Jones, J. B. Bioorg. Med. Chem. 1999, 7,
1381–1387.
17. DeSantis, G.; Shang, X.; Jones, J. B. Biochemistry 1999,
38, 13391–13397.
Using the general procedure, 11 was obtained in the
1
yields and % e.e. as indicated in Table 3. H NMR
18. Davis, B. G.; Shang, X.; DeSantis, G.; Bott, R. R.; Jones,
J. B. Bioorg. Med. Chem. 1999, 7, 2293–2301.
19. Davis, B. G.; Khumtaveeporn, K.; Bott, R. R.; Jones, J.
B. Bioorg. Med. Chem. 1999, 7, 2303–2311.
20. Plettner, E.; DeSantis, G.; Stabile, M. R.; Jones, J. B. J.
Am. Chem. Soc. 1999, 121, 4977–4981.
(DMSO-d6) l 0.87 (d, J=6.5 Hz, 3H, CH3), 2.00–2.50
(m, 4H, CH2CO), 2.20–2.40 (m, 1H, CHCH3), 3.60 (d,
J=5.5 Hz, 2H, NHCH2CO), 5.08 (s, 2H, CH2Ph), 6.99
(brs, 1H, NH2), 7.28 (brs, 1H, NH2), 7.30–7.45 (m, 5H,
Ph), 8.01 (t, J=5.5 Hz, 1H, NH); 13C NMR (acetone-
d6) l 20.4, 29.1, 30.5, 41.7, 43.4, 66.8, 129.2, 129.3,
129.7, 137.9, 172.7, 173.0, 173.3.
21. DeSantis, G.; Paech, C.; Jones, J. B. Bioorg. Med. Chem.
2000, 8, 563–570.
22. Davis, B. G.; Lloyd, R. C.; Jones, J. B. Bioorg. Med.
Chem. 2000, 8, 1527–1535.
23. Lloyd, R. C.; Davis, B. G.; Jones, J. B. Bioorg. Med.
Chem. 2000, 8, 1537–1544.
Acknowledgements
24. Kenyon, G. L.; Bruce, T. W. Methods Enzymol. 1977, 47,
407–430.
25. Wynn, R.; Richards, F. M. Methods Enzymol. 1995, 251,
351–356.
26. For an example of combinatorial type screening, see:
Plettner, E.; Khumtaveeporn, K.; Shang, X.; Jones, J. B.
Bioorg. Med. Chem. Lett. 1998, 8, 2291–2296.
27. Nomenclature according to: Schechter, I.; Berger, A.
Biochem. Biophys. Res. Commun. 1967, 27, 152–162.
28. Dickman, M.; Lloyd, R. C.; Jones, J. B. Tetrahedron:
Asymmetry 1998, 9, 4099–4102.
We are grateful to the National Science and Engineer-
ing Research Council of Canada and Genencor Inter-
national Inc. for generous funding; to Genencor
International Inc. for providing WT- and S166C-SBL
enzymes and to Dr. Rick Bott for helpful discussions.
References
1. Wong, C.-H.; Whitesides, G. M. Enzymes in Synthetic
Organic Chemistry; Pergamon Press: Oxford, 1994; pp.
41–130 and references cited therein.
2. Faber, K. Biotransformations in Organic Chemistry, 2nd
ed.; Springer: Heidelberg, 1995; pp. 298–305 and refer-
ences cited therein.
3. Gotor, V. Bioorg Med. Chem. 1999, 7, 2189–2197.
4. Schellenberger, V.; Schellenberger, U.; Jakubke, H. D.;
Zapevalova, N. P.; Mitin, Y. V. Biotechnol. Bioeng. 1991,
38, 319–321.
5. Bordusa, F.; Ullmann, D.; Elsner, C.; Jakubke, H. D.
Angew. Chem., Intl. Ed. Engl. 1997, 36, 2473–2475.
6. Gunther, R.; Bordusa, F. Chem. Eur. J. 2000, 6, 463–467.
7. Bonneau, P. R.; Graycar, T. P.; Estell, D. A.; Jones, J. B.
J. Am. Chem. Soc. 1991, 113, 1026–1030.
29. Khumtaveeporn, K.; DeSantis, G.; Jones, J. B. Tetra-
hedron: Asymmetry 1999, 10, 2563–2572.
30. Janes, L. E.; Lo¨wendahl, C.; Kazlauskas, R. J. Chem.
Eur. J. 1998, 4, 2324–2331 and references cited therein.
31. Phenyllactate esters acyl donors have been used previ-
ously in protease-catalysed ligations with amino acid acyl
acceptors. However, in these examples the stereochem-
istry of the phenyllactates was not given. See Refs. 32 and
33.
32. Gerosch, S.; Jakubke, H.-D.; Kreuzfeld, H.-J. Tetra-
hedron: Asymmetry 1995, 6, 3039–3045.
33. Haensler, M.; Thust, S.; Klossek, P.; Ullmann, G. J. Mol.
Catal. B 1999, 6, 95–98.
34. Rosell, C. M.; Fernandez-Lafuente, R.; Guisan, J. M. J.
Mol. Catal. 1993, 84, 365–371.
35. Fite, M.; Capellas, M.; Benaiges, M. D.; Caminal, G.;
Clapes, P.; Alvaro, G. Biocatal. Biotransform. 1997, 14,
317–332.
8. Jackson, D. Y.; Burnier, J.; Quan, C.; Stanley, M.; Tom,
J.; Wells, J. A. Science 1994, 266, 243–247.
9. Chen, K. Q.; Arnold, F. H. Proc. Natl. Acad. Sci. USA
1993, 90, 5618–5622.
10. Hopfner, K. P.; Kopetzki, E.; Kresse, G. B.; Bode, W.;
Huber, R.; Engh, R. A. Proc. Natl. Acad. Sci. USA 1998,
95, 9813–9818.
36. Knapp, M.; Daubermann, J.; Bott, R. R. RCMB-PDB
Entry 1jea.
37. Due to the slowness of these reactions they were typically
stopped when product was observed by TLC at the time
indicated in Table 1. It should be noted that in all cases
starting material was recovered and the low yields were
not due to competing hydrolysis.
38. Seebach, D.; Matthews, J. L. J. Chem. Soc., Chem.
Commun. 1997, 2015–2022.
39. The use of specifically designed b-alaninate ester acyl
donors for substrate engineered approaches to protease-
catalysed b-amino acid peptide ligation has been reported
previously. See Refs. 6 and 40.
11. Ness, J. E.; Welch, M.; Giver, L.; Bueno, M.; Cherry, J.
R.; Borchert, T. V.; Stemmer, W. P. C.; Minshull, J.
Nature Biotechnol. 1999, 17, 893–896.
12. Stabile, M. R.; Lai, G. W.; DeSantis, G.; Gold, M.;
Jones, J. B.; Mitchinson, C.; Bott, R. R.; Graycar, T. P.;
Liu, C.-C. Bioorg. Med. Chem. Lett. 1996, 6, 2501–2506.
13. Berglund, P.; Stabile, M. R.; Gold, M.; Jones, J. B.;
Mitchinson, C.; Bott, R. R.; Graycar, T. P. Bioorg. Med.
Chem. Lett. 1996, 6, 2507–2512.
14. Berglund, P.; DeSantis, G.; Stabile, M. R.; Shang, X.;
Gold, M.; Bott, R. R.; Graycar, T. P.; Lau, T. H.;
.

K. Khumta6eeporn et al. / Tetrahedron: Asymmetry 12 (2001) 249–261
40. Sekizaki, H.; Itoh, K.; Toyota, E.; Tanizawa, K. Chem. 1992, 48, 8775–8780.
261
Pharm. Bull. 1999, 47, 104–110.
48. Mravik, A.; Bocskei, Z.; Katona, Z.; Markovits, I.; Pokol,
G.; Menyhard, D. K.; Fogassy, E. Chem. Commun. 1996,
1983–1984.
49. Bajwa, J. S. Tetrahedron Lett. 1992, 33, 2299–2302.
50. Hanson, H. T.; Smith, E. J. Biol. Chem. 1948, 175,
833–848.
51. Ben-Ishai, D.; Berger, A. J. Org. Chem. 1952, 17, 1564–
1570.
52. Tsuji, K.; Terao, Y.; Achiwa, K. Tetrahedron Lett. 1989,
30, 6189–6192.
41. Podlech, J.; Seebach, D. Liebigs Ann. 1995, 1217–1228.
42. Yokomatsu, T.; Minowa, T.; Murano, T.; Shibuya, S.
Tetrahedron 1998, 54, 9341–9356.
43. Hernandez, J. F.; Soleilhac, J. M.; Roques, B. P.;
Fourniezaluski, M. C. J. Med. Chem. 1988, 31, 1825–1831.
44. The balance of mass was recovered starting material.
45. Hsia, C. Y.; Gunshaw, G.; Paech, C.; Murray, C. J. Anal.
Biochem. 1996, 242, 221–227.
46. Muller, D.; Zeltser, I.; Bitan, G.; Gilon, C. J. Org. Chem.
1997, 62, 411–416.
47. Gutman, A. L.; Shkolnik, E.; Shapira, M. Tetrahedron
53. Itoh, T.; Chika, J.; Takagi, Y.; Nishiyama, S. J. Org. Chem.
1993, 58, 5717–5723.
.