A bacterial tyrosine aminomutase proceeds through retention or inversion of stereochemistry to catalyze its isomerization reaction

Article abstract of DOI:10.1021/ja403918w

β-Amino acids are biologically active compounds of interest in medicinal chemistry. A class I lyase-like family of aminomutases isomerizes (S)-α-arylalanines to the corresponding β-amino acids by exchange of the NH₂/H pair. This family uses a 3

Full text of DOI:10.1021/ja403918w

Article
pubs.acs.org/JACS
A Bacterial Tyrosine Aminomutase Proceeds through Retention or
Inversion of Stereochemistry To Catalyze Its Isomerization Reaction
Udayanga Wanninayake† and Kevin D. Walker†,‡,*
‡
†Department of Chemistry, and Department of Biochemistry and Molecular Biology, Michigan State University, East Lansing,
Michigan 48824, United States
S
* Supporting Information
ABSTRACT: β-Amino acids are biologically active com-
pounds of interest in medicinal chemistry. A class I lyase-like
family of aminomutases isomerizes (S)-α-arylalanines to the
corresponding β-amino acids by exchange of the NH2/H pair.
This family uses a 3,5-dihydro-5-methylidene-4H-imidazol-4-
one (MIO) group within the active site to initiate the reaction.
The absolute stereochemistry of the product is known for an
MIO-dependent tyrosine aminomutase from Chondromyces
crocatus (CcTAM) that isomerizes (S)-α-tyrosine to (R)-β-tyrosine. To evaluate the cryptic stereochemistry of the CcTAM
mechanism, (2S,3S)-[2,3-2H2]- and (2S,3R)-[3-2H]-α-tyrosine were stereoselectively synthesized from unlabeled (or [2H]-
labeled) (4′-hydroxyphenyl)acrylic acids by reduction with D2 (or H2) gas and a chiral Rh-Prophos catalyst. GC/EIMS analysis
of the [2H]-β-tyrosine biosynthesized by CcTAM revealed that the α-amino group was transferred to Cβ of the phenylpropanoid
skeleton with retention of configuration. These labeled substrates also showed that the pro-(3S) proton exchanges with protons
from the bulk media during its migration to Cα during catalysis. 1H- and 2H NMR analyses of the [2H]-β-tyrosine derived from
(2S)-[3,3-2H2]-α-tyrosine by CcTAM catalysis showed that the migratory proton attached to Cα of the product also with
retention of configuration. CcTAM is stereoselective for (R)-β-tyrosine (85%) yet also forms the (S)-β-tyrosine enantiomer
(15%) through inversion of configuration at both migration termini, as described herein. The proportion of the (S)-β-isomer
made by CcTAM during steady state interestingly increased with solvent pH, and this effect on the proposed reaction mechanism
is also discussed.
available natural and nonnatural α-amino acids through
aminomutase catalysis.
INTRODUCTION
■
β-Amino acids are emerging as an important class of
compounds that are present in bioactive natural products,
such as the antineoplastic pharmaceutical paclitaxel isolated
from Taxus plants,1,2 the aminopeptidase inhibitor bestatin
obtained from Streptomyces olivoreticuli,3 an antibacterial
blasticidin S from S. griseochromogenes,4 the antibiotic agent
enediyne C-1027 from S. globisporus,5 the anti-tuberculosis
agent viomycin from S. vinaceus,6 the antibiotic andrimid from
Pantoea agglomerans, and cytotoxic agents chondramides A−D
from Chondromyces crocatus.7 In addition, single β-aryl-β-
alanines show anti-epileptogenesis activity.8 Other β-aryl-β-
alanines have been used as building blocks toward the synthesis
of complex bioactive molecules, including β-lactams,9 β-
peptides as mimics of α-peptides,10,11 and antimicrobial
compounds.12
Several chiral synthetic strategies have been developed for
variously substituted β-arylalanines that include tandem one-
pot processes13 and conjugate addition of homochiral lithium
amides to α,β-unsaturated acceptors.14 The Knoevenagel
condensation of benzaldehyde and malonic acid in the presence
of NH4OAc produced a series of racemic β-arylalanines.15 By
contrast, there are only a few reports on the biocatalysis of
asymmetric β-arylalanines from the corresponding readily
So far, five 3,5-dihydro-5-methylidene-4H-imidazol-4-one
(MIO)-dependent aminomutases are known to isomerize
either (S)-α-phenylalanine (EncP from S. maritimus,16 AdmH
(PaPAM) from P. agglomerans,17,18 and TcPAM from Taxus
plants19,20), or (S)-α-tyrosine ((S)-1) (SgcC4 (SgTAM) from
S. globisporus21,22 and CmdF (CcTAM) from C. crocatus,23
further described herein) to their respective β-amino acids.
Efforts to optimize the turnover of this family of aminomutases
provides a potentially alternative means toward scaleable
biocatalytic production of novel enantiomerically pure β-
amino acids as synthetic building blocks in medicinal chemistry.
Earlier substrate specificity studies showed that TcPAM can
convert substituted aromatic and heteroaromatic α-alanines to
the corresponding β-alanines. Apparently, since TAMs require
the 4′-hydroxyl group on the substrate for catalysis, their use to
biosynthesize β-tyrosine analogues is limited.21,23
These aminomutases belong to a class I lyase-like family
(comprising ammonia lyases24−26 and aminomutases18,19,23,27
)
where, mechanistically, the amino group of the arylalanine
substrate nucleophilically attacks the MIO to form an amino
Received: April 19, 2013
Published: June 25, 2013
© 2013 American Chemical Society
11193
dx.doi.org/10.1021/ja403918w | J. Am. Chem. Soc. 2013, 135, 11193−11204

Journal of the American Chemical Society
Article
acid−MIO adduct (Scheme 1). Removal of a β-proton via a
tyrosine residue and concomitant elimination of the NH2−
end, we add further mechanistic detail to CcTAM, which
converts (S)-1 to (R)-2 on the biosynthetic pathway of the
cytotoxic cyclodepsipeptide chondramides A−D in C. crocatus
myxobacteria (Scheme 2).7,23,31 The (R)-product stereo-
Scheme 1. General MIO-Dependent Aminomutase
Mechanism
Scheme 2. CcTAM Reaction on the Chondramides A−D
Pathway
MIO adduct from the substrate produces an intermediary
arylacrylate product. Interchange and rebound of the transient
hydrogen and NH2−MIO to the intermediate, and release of
the β-amino acid completes the conversion. Stereochemical
evidence shows that these enzymes can be sorted by their
enantioselectivity for the β-amino acid product; EncP, PaPAM,
and SgTAM make (S)-β-arylalanines, and CcTAM and TcPAM
make (R)-β-arylalanines. Further, the properties of the MIO
aminomutases segregate according to whether the cryptic
stereochemistry at Cα and Cβ of the product is inverted or
retained. The different enantioselectivities of these isozymes is
determined by the fate of the intermediate formed on distinct
reaction pathways.
For example, both PaPAM and SgTAM presumably bind
their respective substrates and displace the NH2−MIO adduct
and pro-(3S) hydrogen by anti-elimination. For both enzymes,
the NH2−MIO and hydrogen interchange positions and
rebound to the same face of the cinnamate intermediate from
which they were removed to form the (S)-β-amino acids, with
inversion of configuration at Cα and Cβ. Therefore, the
intermediate on this reaction sequence must reside as a single
rotamer.17,27 By contrast, for the isozyme TcPAM, the
cinnamate intermediate is proposed to rotate 180° about the
C1−Cα and Cβ−Cipso bonds. Then the labile NH2 and pro-(3S)
hydrogen exchange positions and reattach to the opposite face
of the intermediate from which they were removed. This
mechanism results in retention of configuration at Cα and Cβ to
form (R)-β-phenylalanine.17,19,28,29
Interestingly, the tyrosine aminomutases (TAMs) are less
enantioselective than the PAMs. The former make both
enantiomers of the β-tyrosine (2) at steady state and show
significantly lower enantiomeric excess after reaching equili-
brium,23,30 while the PAM enzymes make product at >99% ee,
even after reaching equilibrium, and racemization is not
observed.17,20,28,29 It should be noted that earlier mutational
studies on CcTAM (Glu399 → Lys399) improved the
enantioselectivity for (R)-2 from 69% to 97% ee.23 Despite
this earlier study to correlate active site residues with the
stereoselectivity of CcTAM, the reason why the β-amino acid
enantioselectivity was enhanced remained unexplored. In
addition, the mode of attachment (i.e., retention or inversion
of configuration) of the reciprocally migratory NH2 group and
the hydrogen at Cα and Cβ during the reaction is unknown.
To understand and ultimately improve the substrate
specificity profile, turnover rate, and enantioselectivity of
CcTAM, its mechanisms must be fully understood. To this
chemistry catalyzed by CcTAM suggested that its stereo-
chemical course is related to that catalyzed by TcPAM. Herein,
we used deuterium-labeled isotopomers of α-tyrosine to
evaluate the cryptic stereochemistry of the CcTAM mechanism
and compare it to that of other aminomutases.
RESULTS
■
CcTAM Activity and Stereochemistry. CcTAM was
expressed from the pET-28a(+) vector in Escherichia coli
(BL21), and then CcTAM and (S)-1 were incubated. The
amino acids and hydroxycinnamate were derivatized and
analyzed by GC/EIMS to show that product 2 (90 mol %)
and byproduct 4′-hydroxycinnamate (10 mol %) were formed.
The biosynthetic β-tyrosine was also derivatized with a chiral
auxiliary that indicated a mixture containing (R)-2 and (S)-2 at
an 85:15 ratio (See Figures S1 and S2 in Supporting
Information [SI]). This product distribution was consistent
with that shown in an earlier study.23
Assignment of the Prochiral Hydrogens of (R)-2 by 1H
1
NMR. The H NMR of authentic unlabeled (R)-2 showed
signals at δ 4.41 (dd, J = 4.4, 10.1 Hz, Cβ−H), 2.73 (dd, J =
10.1, 16.6 Hz, Cα−H), 2.61 (dd, J = 4.4, 16.7 Hz, Cα−H) in
CD3OD solvent (Figure 1). According to the Karplus equation
32
1
for H NMR, 3J coupling constants of vicinal protons are
largest when the dihedral torsion angle (ϕ) is constrained at 0°
(eclipsed conformation) or 180° (anti conformation). Smaller
3J coupling constants are observed when ϕ approaches 90°.
Thus, the magnitudes of the 3J couplings (Jeq−eq and Jeq−ax (ϕ =
1
3
Figure 1. Partial H NMR profile of unlabeled 2 and the J coupling
constants for the ABX spin system of (R)-2.
11194
dx.doi.org/10.1021/ja403918w | J. Am. Chem. Soc. 2013, 135, 11193−11204

Journal of the American Chemical Society
Article
60°) between 3 and 4 Hz, and Jax−ax (ϕ = 180°) between 8 and
13 Hz) of the conformationally restricted vicinal axial (ax) and
equatorial (eq) protons of cyclohexane33,34 were used as a
guide to assign the 1H NMR signals of (R)-2. The geminal Cα−
H signals (designated as A and B spins) of the unlabeled (R)-2
spin-coupled with Cβ−H (designated as the X spin) were
identified by their distinct J-values [AX (3JAX = 10.1 Hz, ϕ ≈
180°) and BX (3JBX = 4.4 Hz, ϕ ≈ 60°)] (Figure 1).
Assessing the Mode of the Amino Group Attachment
at Cβ by CcTAM. Synthesis of (2S,3S)-[2,3-2H2]- and (2S,3R)-
[3-2H]-(S)-1. The mode of the amino group transfer to Cβ
during the isomerization reaction catalyzed by CcTAM was
assessed. (2S,3S)-[2,3-2H2]- and (2S,3R)-[3-2H]-1 were
synthesized by stereospecific reduction of unlabeled and
[3-2H]-labeled (Z)-2-benzamido-3-(4′-hydroxyphenyl)acrylic
acid using a chiral [Rh((R)-Prophos)(NBD)]ClO4 catalyst
with deuterium or hydrogen gas, according to an earlier
procedure.35 The [2H]-labeled isotopomers of 1 were treated
with a chiral auxiliary (S)-2-methylbutyric anhydride and
titrated with diazomethane to make the 4′-O,3-N-di((S)-2-
methylbutanoyl)-α-tyrosine methyl ester derivatives. The
retention times and mass spectrometry fragmentation of the
synthesized [2H]-labeled (S)-1 derivatives observed by GC/
EIMS analysis were identical to those of an identically
derivatized sample of authentic (S)-1.
Using NMR to Assess the Mechanism of the Hydro-
2
gen Transfer at Cα in the CcTAM Reaction. The H NMR
chemical shifts of the propanoid side chain deuteriums at Cα
and Cβ of [2H2]-2 biosynthesized from (2S)-[3,3-2H2]-1 by
CcTAM were at δ 2.61 (relative peak area = 0.7) and 4.41
(relative peak area = 1.0), respectively (Figure 2A). The
Analysis of 2 Made by CcTAM Catalysis from (2S,3S)-
[2,3-2H2]- and (2S,3R)-[3-2H]-1. CcTAM was incubated
separately with (2S,3S)-[2,3-2H2]- and (2S,3R)-[3-2H]-1.
Afterward, the reaction was basified, and the amino acids
were derivatized to their 4′-O,3-N-di(ethoxycarbonyl) deriva-
tives, acidified, extracted from the aqueous reaction buffer, and
reacted with diazomethane to make the methyl esters.
Interestingly, during the basification step of the derivatization
procedure, ethanol byproduct from the excess ethyl chlor-
oformate caused partial ethyl esterification of the amino acids
(10−30 mol % compared to the methyl esters). The mass
spectrum of the methyl ester derivatives of the β-isomers
yielded ions of diagnostic labeled fragments that were
perturbed by overlapping satellite fragment ions. Therefore,
the ethyl esters of the β-isomers were separated by GC, and the
fates of the deuteriums on the β-amino acids, originating from
the substrate, were evaluated by the identity of diagnostic
fragment ions made in the mass spectrometer (see SI Figures
S3−S5 for fragment ions). The molecular ion [M+] (m/z 353)
of the 4′-O,3-N-di(ethoxycarbonyl) ethyl ester derivative of
unlabeled (R)-2 fragmented into diagnostic ions F1A, F2A, and
F3A (m/z 280, 266, and 194, respectively) (Table 1A). The
corresponding fragment ions of the derivatized β-amino acid
biosynthesized by CcTAM from (2S,3S)-[2,3-2H2]-1 were F1B,
F2B, and F3B (m/z 282, 266, and 194, respectively, Table 1B).
Fragment ions F2B (m/z 266) and F3B (m/z 194) are
identical to those of the unlabeled derivatized product (Table
1A), indicating that deuterium was not at Cβ of the biosynthetic
2. Further, the fragment ion F1B (m/z 282, Table 1B)
corresponding to the intact phenylpropanoid was shifted by
two mass units above the similar fragment ion of the unlabeled
isotopomer. These data indicate that both deuteriums are at Cα
of the biosynthetic product derived from (2S,3S)-[2,3-2H2]-1.
Notably, the mass spectrum of the biosynthetic product 2 made
from (2S,3S)-[2,3-2H2]-1 showed a molecular ion ([2H2](M)+,
m/z 355) and ion m/z 354 ([2H1](M)+), one mass unit lower
(Table 1B), indicating 36% deuterium was lost from the
product compared to the substrate. This deuterium-to-hydro-
gen (D → H) exchange during the CcTAM isomerization of
(2S,3S)-[2,3-2H2]-1 was also supported by the ratio of
diagnostic fragment ion F1B (m/z 282) (Table 1B) and its
[F1B − 1] partner (m/z 281), showing 32% deuterium loss.
Conversely, when (2S,3R)-[3-2H]-1 was used as the substrate
with CcTAM, the resulting derivatized [2H]-labeled 2 yielded
fragment ion F1C (m/z 281) that was shifted one mass unit
over its unlabeled counterpart (Table 1C), indicating retention
Figure 2. 2H NMR (after solvent exchange into CH3OH); the relative
area of the peaks at δ 4.41 and 2.61 are shown (A) and 1H NMR (after
solvent exchange into CD3OD) (B) of a mixture containing the
remaining substrate (2S)-[3,3-2H2]-1 and the biosynthetic (2S,3R)-
1
[2,3-2H2]-2 after a CcTAM-catalyzed reaction. H NMR (in CD3OD)
of authentic (R)-2 (C). The signals for the prochiral protons of
authentic (R)-2 are aligned (boxes) with signals for the deuterium
labeled product in the biosynthetic sample.
1
deuterium signals were compared to the H NMR chemical
shifts of the prochiral hydrogens of authentic (R)-2 (Figure
2A,C) to assign the absolute stereochemistry of the [2H]-
1
labeled biosynthetic sample. The H NMR signal at δ 2.73
(singlet) produced by the biosynthetic [2H]-labeled (R)-2
(Figure 2B) coincided with the chemical shift of the pro-(2R)
proton of authentic (R)-2 (Figure 2C). Two nearby doublets
were also observed at δ 2.75 (d, J = 16.8 Hz, 1 H) and 2.63 (d, J
= 16.8 Hz, 1 H) with a peak area that was ∼30% of the singlet
at δ 2.73 (Figure 2B). These resonances corresponded to a
[2H]-labeled isotopomer of 2 containing two geminal hydro-
gens at Cα, creating an AB-coupled spin system.
11195
dx.doi.org/10.1021/ja403918w | J. Am. Chem. Soc. 2013, 135, 11193−11204

Journal of the American Chemical Society
Article
Table 1. EI-MS Fragmentation of 4′-O,3-N-Di(ethoxycarbonyl) Ethyl Ester Derivatives of Labeled and [2H]-Labeled
Biosynthesized Isotopomers of 2
g
D-transfer was negligible compared to H-transfer based on analysis of mass spectrometry fragments.
of one deuterium in the propanoid side chain. Fragment ions
F2C (m/z 267) and F3C (m/z 195) (Table 1C) were also one
mass unit higher than the corresponding fragments derived
from the unlabeled product. The latter clarified that the
deuterium at Cβ of the substrate remained at this position after
the CcTAM isomerization reaction to 2.
Thus, one of the geminal deuteriums of (2S)-[3,3-2H2]-1 was
100% retained at Cβ of the product, while the migratory
deuterium was partially exchanged with hydrogen. This
deuterium loss during the reaction was supported by the
parallel decrease in the ratio between ion F1D (m/z 282) (a
fragment containing two deuteriums, one at Cα (exchanged
during catalysis), the other at Cβ (100% retained)) and ion
[F1D − 1] (m/z 281) (a fragment containing one deuterium at
Cβ) in the same spectrum (Table 1D and see Figure S6 in SI as
reference).
Re-evaluation of the Stereoisomeric Product Distri-
bution Catalyzed by CcTAM. A previous study on CcTAM
reported that the enzyme stereoselectivity was 69% ee for (R)-2
at pH 8.8.23 The earlier study, however, also showed that (S)-2
was produced at <5 mol % for the duration of the steady state
(0−4 h) of the reaction containing CcTAM (10−50 μg) but
increased significantly as the reaction entered equilibrium.23
This interesting perturbation in the production of (R)- and (S)-
2 at equilibrium prompted us to re-evaluate this reaction
phenomenon. Here, GC/EIMS was used to separate the
diastereoisomeric 4′-O,3-N-di((S)-2-methylbutanoyl)-β-tyro-
sine methyl ester derivatives of the enantiomers of 2 produced
by CcTAM at steady state. Even after 1% conversion of the
substrate, after 5 min, CcTAM (50 μg) was already producing
Assessing the D → H Exchange Rate during CcTAM
Catalysis. (2S)-[3,3-2H2]-1 was incubated with CcTAM in a
time course experiment to assess the extent of the D → H
exchange during the isomerization (Table 1D). The β-amino
acids isolated at designated time intervals were derivatized, and
their 4′-O,3-N-di(ethoxycarbonyl) ethyl esters were analyzed by
GC/EIMS, as before. Tracking the molecular ion [2H2](M)+
(m/z 355) and its [2H1](M)+ partner (m/z 354) showed that
the deuterium enrichment decreased from 85% at 10 min to
60% after 12 h, under steady-state reaction conditions. After 45
h, the reaction reached equilibrium (70% conversion of
substrate to products 4′-hydroxycinnamate and 2, and the D
→ H exchange stabilized at 30% deuterium retention (see
Table 1D and SI Figure S6−S8). Moreover, the relative
abundances of fragment ions F2D (m/z 267) and F3D (m/z
195) (Table 1D) (both one mass unit above their unlabeled
counterparts) from the derivatized biosynthetic product 2, at
each time point, confirmed that a deuterium remained at Cβ.
11196
dx.doi.org/10.1021/ja403918w | J. Am. Chem. Soc. 2013, 135, 11193−11204

Journal of the American Chemical Society
Article
an enantiomeric mixture containing 80% ee of (R)-2 from 1
mM of substrate. The mixture approached 85:15 R/S (i.e., 70%
ee of (R)-2) at a steady-state rate just before reaching
equilibrium at 120 min (Figure 3 and Figure S2 of the
Figure 4. (S)-2 (as % of total 2) measured after substrate (S)-1 was
depleted by 13%, 35%, and 56% at pH 7, 8, and 9 while incubated with
CcTAM.
of the (2S,3S)-[2,3-2H2]-1 substrate was replaced by the amino
group en route to 2 (Table 1).
Figure 3. Analysis of the diastereomeric mixture of products catalyzed
■
●
by CcTAM. Plotted are mol % of (R)-2 ( ) and (S)-2 ( ) relative to
Fragment ion F1B revealed that the deuterium migrated
reciprocally to Cα (Table 1B), thus confirming that CcTAM
migrated the pro-(3S) hydrogen from Cβ to Cα. Further,
fragments ions F2C and F3C of the derivatized [2H]-labeled
product (Table 1C) showed complementary that the Cβ
deuterium of the (2S,3R)-[3-2H]-substrate was retained at its
original position. Coupled with the known (R)-2 stereo-
chemistry made in the aminomutase reaction, the deuterium
remaining at Cβ confirmed that CcTAM uses a retention-of-
configuration mechanism at the amino migration terminus.
Hydrogen Migration Stereochemistry. The transient Cβ-
deuterium of (2S,3S)-[2,3-2H2]-1 migrates from Cβ to Cα as the
amino group moves reciprocally from Cα to Cβ. The
stereochemical mode of attachment of the deuterium to Cα
was assessed by 1H- and 2H NMR analyses of the zwitterion of
the product dissolved in methanol. The magnitude of the
difference between the AX and BX coupling constants (Δ3J ≈ 6
Hz) of (R)-2 was measured and compared to the
experimentally calculated Δ3J (∼7 Hz) for the vicinal protons
of the cyclohexane structure.33,34 These similar Δ3J magnitudes
calculated for these two structures suggested that the
propionate hydrogens of (R)-2 are conformationally restricted
as they are in cyclohexane. It can be imagined that a
monodentate interaction between the ammonium cation and
carboxylate anion of the (R)-2 zwitterion forms a pseudo-
staggered six-membered ring to account for the magnitude of
the observed Δ3J (Figure 5).
▲
amount of (S)-1 added. The amount of (R)-2 (as %) ( ) relative to
the total amount of (R)- and (S)-2 made at steady state. (Average of
duplicate assays is plotted).
Supporting Information). The steady-state production of (S)-2
(from 5 to 15 mol %) over 2 h starkly contrasted the <5 mol %
production of (S)-2 prior to reaching equilibrium over 4 h, as
reported earlier.23
pH Effect on the Stereoselectivity of the Reaction
Catalyzed by CcTAM. The CcTAM reaction was incubated
with (S)-1 separately at different pH values. After the
conversion of (S)-1 to 2 reached 13, 35, and 56% at each
pH, the amount of the (S)-isomer (the minor antipode) relative
to the total amount of 2 was calculated. The rate of the reaction
was similar at pHs 8 and 9, yet half as fast at pH 7. After the
conversion of substrate to product reached ∼13% in each assay,
the samples incubated at pHs 7, 8, and 9 contained respectively
10.8%, 13.0%, and 16.7% (S)-2, relative to the (R)-antipode
(Figure 4). At higher percentages of conversion, the proportion
of (S)-2 continued to increase, and the trend of increasing
production of (S)-2 with higher pH was also maintained
(Figure 4).
DISCUSSION
■
Retention of Configuration at the Migration Termini
during the CcTAM Reaction. Amino Group Migration. The
stereochemistry of the major biosynthetic product 2 made by
CcTAM was confirmed herein as (3R); this was consistent with
the assignment made in an earlier report.23 The synthesis of
stereospecifically deuterium-labeled isotopomers (2S,3S)-
[2,3-2H2]- and (2S,3R)-[3-2H]-1 made it possible to assess
the mode of attachment of the amino group at Cβ during the
isomerization reaction catalyzed by CcTAM. GC/EIMS profiles
of derivatized isotopomers of 2 showed diagnostic fragment
ions that informed on the location of the deuteriums. Fragment
ions F2B and F3B (Table 1B) revealed that the Cβ deuterium
To support the proposed restricted rotamer conformation,
the carboxylate group of (R)-2 was methyl esterified to disrupt
the ionic ammonium/carboxylate ionic interaction. In acyclic
systems, the conformational preference of hydrogens in an ABX
spin system is commonly lower than in a restricted conformer.
3
3
As an example, J(AX) (7.8 Hz) and J(BX) (6.3 Hz) coupling
constants (Δ3J(AX)−(BX) = 1.5 Hz) of the (R)-2 methyl ester
were nearly identical in CD3OD. This observation suggested
that HA and HB in the (R)-2 methyl ester was conformationally
more equivalent with respect to HX due to greater rotational
freedom about the Cα−Cβ bond. This likely resulted from
11197
dx.doi.org/10.1021/ja403918w | J. Am. Chem. Soc. 2013, 135, 11193−11204

Journal of the American Chemical Society
Article
Scheme 3. Inversion and Retention of Configuration
a
Pathways Catalyzed by CcTAM
Figure 5. Intramolecular salt-bridge between the ammonium ion and
carboxylate group of (R)-2 in methanol. Dihedral angles (ϕ1 and ϕ2)
between HA and HX and between HB and HX, respectively, in the
pseudo-six-membered ring formed by 2 are shown in Newman
Projection.
a
(Path a) Retention and (Path b) inversion-of-configuration at Cα and
Cβ after exchange and reattachment of the pro-(3S) proton (H†) and
the NH2 of the (S)-1 substrate to make (R)-2 and (S)-2, respectively
by CcTAM catalysis.
removal of the intramolecular ionic interaction of underivatized
(R)-2 (see Figure 5).
The proposed six-membered ring conformation of (R)-2
places the side chain hydrogens of the amino acid in defined
axial and equatorial positions by restricted rotation about the
Cα−Cβ bond (Figure 5). In addition, the carboxylate and
aromatic ring of (R)-2 are considered to be positioned anti,
according to a described lowest energy rotamer of 1.36
Therefore, based on the 1H NMR chemical shifts and coupling
constants, HA was definitively assigned as pro-(2R) and HB as
pro-(2S) (see Figure 1). These chemical shift assignments were
used as reference in 1H- and 2H NMR analyses of the
deuterium-labeled biosynthetic samples. CcTAM shuttled one
deuterium of (2S)-[3,3-2H2]-1 from Cβ and attached it to Cα
with retention of configuration, placing the deuterium in the
formal pro-(2S) position of (R)-2.
Previous stereochemical studies showed that a related MIO-
dependent phenylalanine aminomutase from Taxus plants
(TcPAM) exchanges the position of the NH2 and hydrogen
migration partners of (S)-α-phenylalanine with retention of
configuration at the terminal carbons.7 Thus, the CcTAM and
TcPAM reactions are likely mechanistically similar,19 where
both remove NH2 and H from their substrates to form an
acrylate intermediate. Apparently, these enzymes can rotate
their intermediates 180° about the C1−Cα and Cβ-Cipso bonds
prior to the rebound of NH2 and H to retain the
stereochemistry in the corresponding β-products.28
TcPAM with (2S)-[3,3-2H2]-α-phenylalanine as the substrate.19
However, no deuterium exchange was observed in the reaction
catalyzed by an isozyme (PaPAM) from the bacteria P.
agglomerans with (2S)-[3,3-2H2]-α-phenylalanine substrate.17
Structural data showed that the PaPAM active site is slightly
smaller and more ordered than TcPAM.28,29 Thus, bulk water
likely can access the active site of TcPAM better than that of
PaPAM. While the CcTAM structure is not yet solved, it can be
imagined that its active site, like that of TcPAM, allows access
by bulk water to account for the observed proton exchange
during catalysis.
The D → H exchange rate progressively increases as the
reaction progresses under steady-state conditions. A monop-
rotic Tyr52 residue, also reported for other MIO aminomutases
(and located on a flexible loop structure),22,28,29 presumably
serves as the general base in the CcTAM reaction. In this case,
after deuterium abstraction from the (2S)-[3,3-2H2]-1, the D →
H exchange on the catalytic Tyr52 should theoretically remain
constant if all of the conditions stayed the same for the duration
of the reaction; however, this is not the case. One notion to
explain the observed increase in deuterium washout uses
ureases as a model. Ureases are known to change the pH of the
enzyme-microenvironment upon release of ammonia from
urea.37,38 Likewise, each of the known MIO-dependent
arylalanine aminomutases catalyzes an ammonia lyase reaction
that releases ammonia and produces an arylacrylate as a
byproduct (see Figure S8 in SI).17,23,28,30 Thus, the micro-
environment of CcTAM, containing increasing amounts of
ammonia, might increase the pH and influence the proton
exchange rates during catalysis. It is conceivable that as the
microenvironment ultimately reaches equilibrium with bulk
water, the D → H exchange becomes zero order as the reaction
reaches equilibrium (see Figure S8 of the SI). To assess this,
CcTAM was separated from the (2S)-[3,3-2H2]-1 substrate and
biosynthesized products after a 24-h time course study. The
isolated enzyme was reincubated with more (2S)-[3,3-2H2]-1,
and the change in the deuterium washout rate was the same as
observed before (see Figure S8 in SI). Thus, the factors
affecting the deuterium exchange were reversible, and the
microenvironment was reset to its initial conditions.
It is important to note that (R)-2 produced by CcTAM
decreased from 90% to 80% (while (S)-2 increased
accordingly) during the steady-state phase of the reaction
from 5 to 120 min (Figure 3). The isotope enrichment in
fragment ion clusters of the biosynthetic [2H]-2 (Table 1) in
the mass spectrometer did not indicate that the (S)-isomer was
2
labeled regioselectively different from (R)-2. In addition, H
NMR analysis of the labeled mixture of 2 indicated isochronous
signals for the deuteriums at Cα of both enantiomers of 2
present in the reaction mixture. Therefore, the biosynthetic
[2H]-(S)-2 was labeled as the enantiomer of (R)-β-isomer. In
contrast to the retention-of-configuration mechanism to access
(R)-2, CcTAM must use an inversion-of-configuration process
to obtain the (S)-isomer (Scheme 3).
Hydrogen Exchange during Migration. (2S)-[3,3-2H2]-1
was incubated with CcTAM in a time course study to track the
D → H exchange rate of the migratory pro-(3S) deuterium
during the isomerization reaction. The deuterium enrichment
in the [2H]-2 decreased from 85% at 5 min to 30% after 10 h as
the reaction reached equilibrium. This level of D → H
exchange was also observed for the (R)-β-phenylalanine
product made in a similar reaction catalyzed by the plant
Diastereomeric Product Ratio Catalyzed by CcTAM. Thus
far, the MIO-dependent aminomutases can be categorized
according to their stereoselectivity. The catalyses of SgTAM
and PaPAM stereoselectively produce (S)-β-arylalanines.17,18 It
can be imagined that these aminomutases follow a similar
stereochemical course during removal and rebounding of the
transient hydrogen and NH2 groups. The crystal structures of a
11198
dx.doi.org/10.1021/ja403918w | J. Am. Chem. Soc. 2013, 135, 11193−11204

Journal of the American Chemical Society
Article
phenylalanine aminomutase (PaPAM) (Figure 6A) and a
tyrosine aminomutase (SgTAM) (Figure 6B) show that the
composed of residues Leu108 and Ile431, which also helps
align the reaction intermediate in TcPAM ∼15° different from
that of phenylalanine about the Cβ-axis in the PaPAM29 (Figure
6A). The altered angle presumably allows the TcPAM
intermediate to rotate 180° and retain the configuration at
the chiral and prochiral centers of the (R)-β-phenylalanine
product. CcTAM probably also proceeds through an identical
process to obtain (R)-2.
The CcTAM structure has not yet been solved. However, its
reaction stereochemistry (retention of configuration at both Cα
and Cβ in the major reaction product) is identical to that of
TcPAM. This suggested that the two enzymes have common
active-site architecture and likely position their reaction
intermediates similarly. Thus, CcTAM was modeled on
TcPAM (PDB 3NZ4) (30% sequence similarity). Placing the
4′-hydroxycinnamate in an orientation identical to that of
cinnamate in TcPAM shows the hydroxyaryl portion of the
former interacts with His81 and Tyr403 residues of CcTAM
(Figure 6D). The plausible hydrogen bonding between CcTAM
and the 4′-hydroxyaryl of the substrate likely helps align the
substrate with Arg298 to form a bidentate salt bridge
interaction (Figure 6D), similar to the interaction in TcPAM.
As mentioned, a significant difference between the amino-
mutase isozymes is the TAMs use His and Tyr residues in the
aromatic pocket to help bind and orient the tyrosine substrate
through hydrogen bonding. The PAMs, instead, use hydro-
phobic residues in the aromatic pocket and likely other not yet
fully understood factors to direct the substrate binding
orientation and stereoselectivity. It is feasible that hydrogen
bonding with the phenol of the substrate, in part, governs the
stereoselectivity of the TAMs and enables the minor antipodal
product to form through a new reaction route. As mentioned
previously, each MIO-dependent arylalanine aminomutases
catalyzes an ammonia lyase reaction, releasing ammonia and an
arylacrylate as byproducts (see Figure S8 in SI).17,23,28,30 We
posited earlier that the observed D → H exchange for the
CcTAM reaction was likely influenced by the increased pH of
the microenvironment caused by ammonia release (Figure S8
in SI). Variation in bulk and local pH is known generally to
disturb H-bonding networks within an enzyme active site.40−42
In the TAM enzymes, the perturbation of the local pH could
also affect the H-bonding network within the active site.
Alteration of the H-bonding network in the 4′-hydroxyaryl
binding pocket of the TAM enzymes could potentially change
the enantioselectivity, since this region likely contributes
significantly to the substrate docking conformation. Therefore,
in this study, the buffered pH of a CcTAM reaction was
adjusted to 9 and below to assess whether this would change
the % ee of the product.
Figure 6. Comparison of the TcPAM, PaPAM, and SgTAM active-site
structures cocrystallized with phenylpropanoid adducts or complexes
and the CcTAM active site with 4′-hydroxycinnamate was modeled on
the TcPAM crystal structure (PDB 3NZ4). The orientation of
phenylpropanoid (center of each diagram) relative to the Arg residue
(at the right of each diagram) is shown. Also shown are key
noncatalytic residues involved in binding and positioning the
substrate; the catalytic tyrosine residue is above the plane in each
drawing and is not shown.
trajectories of the phenylalanine and tyrosine substrates,
respectively, are nearly identical. However, these aminomutases
use distinct enzyme/substrate interactions to orient their
substrates. SgTAM apparently uses residues His93 and
Tyr415 to form a hydrogen-bond network with the 4′-OH of
the tyrosine substrate and place the carboxylate in a
monodentate salt bridge interaction with Arg311.39 By contrast,
these H-bond interactions are absent in PaPAM, which has
hydrophobic residues Val108 and Phe428 positioned analo-
gously to His93 and Tyr415, respectively, of SgTAM. Instead,
PaPAM uses the steric bulk of Phe455 to force the
phenylalanine substrate into a monodentate salt bridge
interaction with Arg323. This steric interaction aligns phenyl-
alanine in PaPAM at an orientation similar to that of the
enzyme/substrate pairing in SgTAM. It was postulated that,
after elimination of the hydrogen and NH2 from the PaPAM
substrate, the distinct angle somehow prevents rotation of the
cinnamate intermediate. In this way, the hydrogen and NH2 can
rebound to the same face of the intermediate from which they
were removed. The similar substrate-docking conformations
and trajectories of PaPAM and SgTAM likely, in part, define
their identical (3S)-product stereoselectivities.
By analogy, CcTAM (described herein) and TcPAM from
Taxus plants both make β-amino acid products with (3R)-
stereochemistry;7,20,23 thus, these aminomutases likely progress
through similar stereochemical profiles. Earlier structural
analyses showed that TcPAM has Asn458 positioned in
contrast to the sterically larger Phe455 of isozyme PaPAM.
The smaller Asn of TcPAM enables the carboxylate of the
transient cinnamate intermediate to engage in a bidentate salt
bridge with Arg325 (Figure 6C). In addition, the aryl portion of
cinnamate sits in a hydrophobic pocket that is, in part,
A lower % conversion of (S)-1 to 2 correlated with a higher
% ee of (R)-2 for each pH tested. Also, at higher pH, the
relative proportion of (S)-2 made by CcTAM increased
compared to that of (R)-2. CcTAM consistently produced
more (S)-2 enantiomer at higher pH, regardless of the amount
of (S)-1 converted to 2 (Figure 4). As we proposed earlier, (S)-
2 is made via a conformer of a 4′-hydroxycinnamate
intermediate that receives the NH2 group and hydrogen
(dwelling temporarily on the enzyme) on the same face from
which they were removed (S)-1. (R)-2 is provided via a rotamer
of the initially formed conformer of 4′-hydroxycinnamate.
Therefore, since the production of (S)-2 increases with higher
pH, these conditions can be imagined to restrict the rotation of
the 4′-hydroxycinnamate intermediate within the CcTAM
11199
dx.doi.org/10.1021/ja403918w | J. Am. Chem. Soc. 2013, 135, 11193−11204

Journal of the American Chemical Society
Article
medium, and the cells were grown at 16 °C for 16 h. The cells were
then harvested by centrifugation, and the resulting pellet was
resuspended in buffer (50 mL of 50 mM sodium phosphate containing
5% (v/v) glycerol, 300 mM NaCl, and 10 mM imidazole, pH 8.0). The
cells were lysed by sonication, and the cellular debris and light
membranes were removed by centrifugation. The crude, functionally
soluble aminomutase was purified by nickel nitrilotriacetic acid (Ni-
NTA) affinity chromatography according to the protocol described by
the manufacturer (Qiagen, Valencia, CA). Fractions that eluted from
the column, containing active CcTAM (62 kDa) in 250 mM imidazole,
were combined. The buffer was exchanged with 50 mM sodium
phosphate (pH 8.0) containing 5% (v/v) glycerol through several
concentration/dilution cycles, using a Centriprep centrifugal filter
(30,000 MWCO, Millipore). The purity of the concentrated CcTAM
(15 mg/mL in 4 mL, estimated by the Bradford method) was >95% by
SDS−PAGE with Coomassie Blue staining.
Assessing the Activity and Stereochemistry of the CcTAM
Reaction. (S)-1 (1 mM) was incubated with CcTAM (0.1 mg) at 31
°C in 50 mM phosphate buffer (1 mL, pH 8.5) for 1 h. The assay
mixtures were treated in twice (5 min each time) with pyridine (0.6
mmol) and ethyl chloroformate (0.5 mmol). This step, containing
excess ethyl chloroformate, caused partial ethyl esterification. After-
ward, the reaction was acidified to pH 2 (6 M HCl), and the 4′-O,2-N-
and 4′-O,3-N-di(ethoxycarbonyl) derivative of α- and β-tyrosine,
respectively, were extracted into ether and treated with a slight excess
of diazomethane. The resulting sample contained a mixture of ethyl
(∼20 mol %) and methyl (80 mol %) esters. 4′-Hydroxycinnamic acid
byproduct was converted to its 4′-O-ethoxycarbonyl-(E)-coumaric acid
ethyl (10 mol %) and methyl (90 mol %) esters under these
conditions. The methyl esters of the α-tyrosine derivative and the 4′-
O-ethoxycarbonyl-(E)-coumaric acid, and the ethyl ester of the β-
tyrosine derivative were analyzed by GC/EIMS.
active site by possibly altering the substrate binding angle or by
strengthening the H-bond network with the 4′-OH of the
substrate. Increases in pH are known to partition the phenol
groups of active-site tyrosines to the phenolates and thereby
shorten H-bond distances interacting with these oxyanions.41
However, the mechanism through which the pH exactly affects
the product stereochemistry of CcTAM remains unclear and
will require further inquiry.
CONCLUSIONS
■
In summary, the synthesis of stereospecifically [2H]-labeled 1
enabled us to provide evidence to support that CcTAM
catalyzes its isomerization reaction along a path similar to that
of TcPAM from a plant. That is, both likely overcome a
torsional barrier to access two pivotal rotameric intermediates.
CcTAM converts one rotamer to (R)-2 as the major product
through retention of configuration. This pathway competes
with a route involving a second rotamer, producing (S)-2 as the
minor enantiomer through inversion of configuration. TcPAM
uses only one rotamer to advance to a single enantiomer, (R)-
β-phenylalanine.
Thus, this study potentially starts to shed light on how
CcTAM catalysis uses the 4′-hydroxyl group to orient the
substrate in the active site. In addition, this investigation
provides a basis by which to probe the driving force that rotates
the branch point intermediate in CcTAM to produce two
enantiomers.
To confirm the stereochemistry of the β-isomer product,23 another
sample of CcTAM (0.1 mg) was incubated at 31 °C with (S)-1 (1
mM) in phosphate buffer (1 mL, pH 8.5) for 1 h. To this solution
were added pyridine (50 μL, 0.64 mmol) and (S)-2-methylbutyric
anhydride (10 μL, 0.05 mmol), and the mixture was stirred vigorously
for 5 min. Another batch of pyridine (0.64 mmol) and (S)-2-
methylbutyric anhydride (0.05 mmol) was added, and the reaction was
stirred for 5 min. The solution was acidified to pH 2 (6 M HCl) and
extracted with diethyl ether (3 × 2 mL). The ether fractions were
combined and dried. The resulting residue was dissolved in diethyl
ether (100 μL) and the solution was titrated with a dilute solution of
diazomethane dissolved in diethyl ether until the yellow color
persisted. These samples were analyzed by GC/EIMS and compared
to the retention time and mass fragmentation of authentic standards
(See SI)
Synthesis of Authentic (S)-1, (R)-2, and (E)-4-Hydroxycin-
namic Acid Derivatives. To (S)-α-, (R)-β-tyrosine or (E)-4′-
hydroxycinnamic acid (0.1 mmol of each) dissolved in 50 mM
phosphate buffer (1 mL, pH 8.5) were added pyridine (200 μL, 2.4
mmol) and ethyl chloroformate (200 μL, 1.6 mmol). The reactions
were stirred for 5 min and treated with another batch of pyridine and
ethyl chloroformate (both at 200 μL) for 5 min with stirring. The
ethanolic mixtures caused partial ethyl esterification. Each mixture was
acidified to pH 2 (6 M HCl), and extracted into diethyl ether (3 × 2
mL). The ether fractions were combined, dried under vacuum, and the
residue was dissolved in methanol (100 μL). To this solution was
added a dilute diazomethane solution dissolved in diethyl ether, until
the yellow color persisted. This procedure resulted in a mixture of
ethyl and methyl esters for each sample, as described earlier.
MATERIALS AND METHODS
■
Chemicals and Reagents. (S)-α-Tyrosine, 4′-hydroxycinnamic
acid, unlabeled- and [α-2H]-4′-hydroxybenzaldehyde, hippuric acid,
ethyl chloroformate, (S)-2-methylbutyric anhydride, acetic anhydride,
pyridine, bicyclo[2.2.1]hepta-2,5-diene (norbornadiene), 2H2 gas,
(2S)-[3,3-2H2]-α-tyrosine ((2S)-[3,3-2H2]-1) and Dowex 50W
(100−200 mesh, H+ form) ion-exchange resin were purchased from
Sigma Aldrich (St. Louis, MO). (R)-3-Amino-3-(4′-hydroxyphenyl)-
propanoic acid and (S)-3-amino-3-(4′-hydroxyphenyl)propanoic acid
were purchased from Peptech Inc. (Bedford, MA). (R)-1,2-Bis-
(diphenylphosphino)propane ((R)-Prophos) was purchased from Alfa
Aesar (Ward Hill, MA). Bicyclo[2.2.1]hepta-2,5-diene-rhodium(I)
chloride dimer ([Rh(NBD)Cl]2) and silver perchlorate were
purchased from Strem Chemicals (Newburyport, MA). Hydrogen
gas (99.995% purity) was obtained from Airgas Great Lakes
(Independence, OH 44131).
1
2
Instrumentation. H NMR (500 MHz), H NMR (76.7 MHz)
and 13C NMR (126 MHz) spectra were obtained on a Varian NMR-
Spectrometer using standard acquisition parameters. X-ray crystallo-
graphic data were collected using a Bruker CCD (charge coupled
device)−based diffractometer equipped with an Oxford Cryostream
low-temperature apparatus operating at 173 K. The biosynthetic
products were quantified and analyzed by gas chromatography/
electron-impact mass spectrometry (GC/EIMS): GC (model 6800N,
Agilent, Santa Clara, CA) was coupled to a mass analyzer (model 5973
inert, Agilent, Santa Clara, CA) in ion scan mode from 50 to 400
atomic mass units. The GC conditions were as follows: column
temperature was held at 200 °C for 1 min and then increased linearly
at 20 °C/min to 250 °C with a 1-min hold. Splitless injection was
selected, and helium was used as the carrier gas.
Subcloning, Expression, and Purification of CcTAM. The
cctam cDNA was codon-optimized by GenScript (Piscataway, NJ
08854) in the pUC57 vector. The clone was ligated into the pET-
28a(+) vector between the NdeI/BamHI cloning sites. Recombinant
plasmids were used to transform Escherichia coli BL21(DE3) cells,
which were grown in 1 L Luria−Bertani medium supplemented with
kanamycin (50 μg/mL). Overexpression of CcTAM was induced by
the addition of isopropyl-β-D-thiogalactopyranoside (100 μM) to the
4′-O-Ethoxycarbonyl-(E)-coumaric Acid Methyl Ester (i.e., 4′-O-
Ethylcarboxy-(E)-cinnamic Acid Methyl Ester). The coumarate ester
was recrystallized from ethanol and isolated at 80% yield (20 mg).
HRMS: m/z 251.0890 ([M + H]+); calculated m/z 251.0919
1
([C13H15O5]+). H NMR (500 MHz, CDCl3) δ: 7.68 (d, J = 16.1
Hz, 1 H), 7.55 (d, J = 8.3 Hz, 2 H), 7.22 (d, J = 8.3 Hz, 1 H), 6.41 (d, J
= 16.1 Hz, 1 H), 4.34 (q, J = 7.3 Hz, 2 H), 3.82 (s, 3 H), 1.41 (t, J =
7.3 Hz, 3 H). 13C NMR (126 MHz, CDCl3) δ: 167.2 (C(O)OCH3),
153.2 (OC(O)O), 152.4 (C4′), 143.6 (Cβ), 132.2 (C1′), 129.2 (C2′),
11200
dx.doi.org/10.1021/ja403918w | J. Am. Chem. Soc. 2013, 135, 11193−11204

Journal of the American Chemical Society
Article
121.6 (C3′), 118.1 (Cα), 65.1 (C(O)OCH3), 51.7 (OCH2CH3), 14.2
(OCH2CH3). (See SI Figures S9−S11 for NMR spectra and GC/
EIMS fragmentation data).
O), 51.0 (Cβ), 40.2 (Cα), 14.1 (CH3CH2OC(O)NH), 13.6
(CH3CH2OC(O)O) (See SI Figures S20−S21 for NMR spectra).
Synthesis of [Rh(NBD)2]ClO4 Complex. The [Rh(NBD)2]ClO4
complex was synthesized according to a described process.35,43 Briefly,
a mixture of dimeric [Rh(NBD)Cl]2 (0.46 g, 1 mmol) and
norbornadiene (0.19 g, 2 mmol) dissolved in CH2Cl2 (15 mL) was
added to silver perchlorate (0.42 g, 2 mmol) under N2, and stirred for
1 h. The suspension was filtered to remove the white precipitate, and
the filtrate was diluted with dry THF (15 mL). The sample was
concentrated under vacuum until the orange needles of [Rh(NBD)2]-
ClO4 appeared. The crystals were collected, washed with ice-cold, dry
THF, and dried under vacuum to obtain 0.65 g (85% isolated yield) of
The 4′-O,2-N-Di(ethoxycarbonyl)-α-tyrosine Methyl Ester. A
mixture of ethyl (10 mol %) and methyl (90 mol %) esters of 4′-
O,2-N-di(ethoxycarbonyl)-α-tyrosine was obtained at 74% yield (25
mg). The methyl ester of the α-tyrosine derivative was used in the GC/
EIMS analyses of labeled and unlabeled α-tyrosine substrates.
Therefore, the mixed ester sample, dissolved in chloroform (500
μL), was loaded onto a preparative silica gel TLC plate and eluted with
90:10 hexane/ethyl acetate. Authentic 4-O,2-N-diethoxycarbonyl-α-
tyrosine methyl ester was isolated (Rf = 0.65) at 67% yield (22.5 mg).
HRMS: m/z 419.1781 ([M + H + pyridine]+); calculated m/z
1
rust brown crystals. H NMR (500 MHz, CDCl3) δ: 5.20 (q, J = 2.0
Hz, 4 H), 4.13 (br s, 2 H), 1.51 (t, J = 1.6 Hz, 2 H).
1
419.1818 ([C21H27N2O7]+). H NMR (500 MHz, CDCl3) δ: 7.21−
Synthesis of [Rh(NBD)((R)-Prophos)]ClO4 Complex. The Rh-
Prophos catalyst was prepared by an established procedure.35,43 To the
orange-red solution of [Rh(NBD)2]ClO4 (0.54 g, 1.4 mmol) and (R)-
Prophos (0.57 g, 1.4 mmol) dissolved in a mixture of dry CH2Cl2 and
THF (5 mL of each) was added hexane (5 mL) dropwise under N2.
The solution stood, undisturbed at room temperature for 5 h, and then
at 5 °C for 16 h. Orange-red crystals were collected by vacuum
filtration and washed with ice-cold, dry THF and then with hexane.
The crystals were dried under N2 to obtain 0.8 g (80% isolated yield)
6.98 (m, 4 H), 5.10 (d, J = 7.3 Hz, 1 H), 4.67−4.55 (m, 1 H), 4.28 (q,
J = 7.1 Hz, 2 H), 4.08 (q, J = 6.7 Hz, 2 H), 3.69 (s, 3 H), 3.10 (dd, J =
6.1, 14.0 Hz, 1 H), 3.05 (dd, J = 6.1, 14.0 Hz, 1 H), 1.36 (t, J = 7.0 Hz,
3 H), 1.20 (t, J = 6.7 Hz, 3 H). 13C NMR (126 MHz, CDCl3) δ: 171.9
(C(O)OCH3), 155.8 (OC(O)NH), 153.5 (OC(O)O), 150.2 (C4′),
133.6 (C1′), 130.3 (C2′), 121.0 (C3′), 64.8 (C(O)OCH3), 61.2
(CH3CH2OC(O)NH), 54.5 (CH3CH2OC(O)O), 52.3 (Cα), 37.6
(Cβ), 14.4 (CH3CH2OC(O)NH), 14.1 (CH3CH2OC(O)O) (See
̂
Figures S12−S16 of SI A for NMR spectra and GC/EIMS
fragmentation data).
1
of the Rh-catalyst. H NMR (500 MHz, CDCl3) δ: 7.79−7.73 (m, 2
H), 7.72−7.67 (m, 3 H), 7.65−7.61 (m, 2 H), 7.61−7.58 (m, 6 H),
7.58−7.55 (m, 6 H), 7.47−7.41 (m, 2 H), 7.35−7.29 (m, 2 H), 5.42
(br s, 2 H), 5.33 (br s, 1 H), 5.31 (s, 1 H), 4.87 (br s, 1 H), 4.28 (br s,
1 H), 4.16 (br s, 1 H), 3.84−3.67 (m, 1 H), 2.71−2.59 (m, 2 H), 2.57
(d, J = 2.2 Hz, 1 H), 2.04 (td, J = 7.4, 12.9 Hz, 1 H), 1.88−1.84 (m, 1
H), 1.84−1.76 (m, 2 H), 1.20 (dd, J = 6.5, 12.3 Hz, 4 H). 13C NMR
(126 MHz, DMSO-d6) δ: 143.1, 135.1, 135.1, 134.8, 134.5, 134.1,
133.9, 132.8, 132.7, 132.0, 131.9, 131.7, 131.1, 131.0, 130.9, 130.1,
129.3, 129.3, 128.9, 128.8, 128.8, 128.6, 128.5, 128.2, 127.8, 125.9,
125.5, 67.0, 63.4, 48.2, 25.1, 14.8, 14.6.
4′-O,2-N-Di(ethoxycarbonyl)-α-tyrosine Ethyl Ester. The corre-
sponding ethyl ester was isolated (Rf = 0.50) from the TLC plate at 7%
yield (2.5 mg). HRMS: m/z 433.1933 ([M + H + pyridine]+);
calculated m/z 433.1974 ([C22H29N2O7]+). 1H NMR (500 MHz,
CDCl3) δ: 7.17 (d, J = 8.6 Hz, 2 H), 7.14−7.11 (m, J = 8.6 Hz, 2 H),
5.14 (d, J = 7.8 Hz, 1 H), 4.63 (dd, J = 5.6, 13.4 Hz, 1 H), 4.33 (q, J =
7.1 Hz, 2 H), 4.17 (q, J = 7.2 Hz, 2 H), 4.12 (q, J = 7.1 Hz, 2 H), 3.11
(d, J = 5.1 Hz, 2 H), 1.40 (t, J = 7.1 Hz, 3 H), 1.25 (t, J = 7.3 Hz, 3 H),
1.24 (t, J = 7.2 Hz, 3 H). 13C NMR (126 MHz, CHCl3) δ: 171.7
(C(O)OCH3), 156.1 (OC(O)NH), 153.8 (OC(O)O), 150.4 (C4′),
133.9 (C1′), 130.6 (C2′), 121.3 (C3′), 65.1 (CH3CH2OC(O)), 61.8
(CH3CH2OC(O)NH), 61.4 (CH3CH2OC(O)O), 54.9 (Cα), 38.0
(Cβ), 14.7 (CH3CH2OC(O)), 14.4 (CH3CH2OC(O)NH), 14.3
(CH3CH2OC(O)O) (See SI Figures S15 and S16 for NMR spectra
and GC/EIMS fragmentation data).
4′-O,3-N-Di(ethoxycarbonyl)-β-tyrosine Ethyl Ester. A mixture of
ethyl (33 mol %) and methyl (67 mol %) esters of 4′-O,3-N-
diethoxycarbonyl-β-tyrosine was isolated (Rf = 0.38) at 68% yield (24
mg). The ethyl ester of the β-tyrosine derivative was used in the GC/
EIMS analyses of labeled and unlabeled biosynthetic β-tyrosines. Thus,
authentic ethyl ester was purified by silica gel TLC (90:10 hexane/
ethyl acetate). 4′-O,2-N-Di(ethoxycarbonyl)-β-tyrosine ethyl ester was
isolated at 23%yield (8 mg). HRMS: m/z 433.1922 ([M + H +
pyridine]+); calculated m/z 433.1974 ([C22H29N2O7]+). 1H NMR
(500 MHz, CDCl3) δ: 7.30 (d, J = 8.8 Hz, 2 H), 7.12 (d, J = 8.8 Hz, 2
H), 5.67 (d, J = 6.8 Hz, 1 H), 5.12 (br s, 1 H), 4.28 (q, J = 7.3 Hz, 2
H), 4.08 (q, J = 6.8 Hz, 2 H), 4.05 (q, J = 7.0 Hz, 2 H), 2.91−2.69 (m,
2 H), 1.36 (t, J = 7.1 Hz, 3 H), 1.20 (t, J = 6.8 Hz, 3 H), 1.14 (t, J = 7.3
Hz, 3 H). (See SI Figures S17−S19 for NMR spectra and GC/EIMS
fragmentation data). 13C NMR (126 MHz, CDCl3) δ: 170.7
(C(O)OCH2CH3), 155.8 (OC(O)NH), 153.5 (OC(O)O), 150.3
(C4′), 138.8 (C1′), 127.4 (C2′), 121.3 (C3′), 64.8 (CH3CH2OC-
(O)), 61.0 (CH3CH2OC(O)NH), 60.7 (CH3CH2OC(O)O), 51.0
(Cβ), 40.6 (Cα), 14.5 (CH3CH2OC(O)), 14.1 (CH3CH2OC(O)-
NH), 14.0 (CH3CH2OC(O)O).
Synthesis of [2H]-Labeled (2S)-1 Isotopomers. Synthesis of
(Z)-2-Benzamido-3-(4′-hydroxyphenyl)acrylic Acid. According to a
described procedure,44 a mixture of 4-hydroxybenzaldehyde (1.9 g,
12.5 mmol), K2HPO4 (2.2 g, 12.5 mmol), and acetic anhydride (3.8
mL, 40 mmol) was stirred and heated at 80 °C under N2 for 5 min. To
the mixture was added hippuric acid (2.3 g, 12.5 mmol) in one lot, and
the reaction was stirred at 80 °C for 2 h. Yellow crystals were collected
by vacuum filtration and washed with water to obtain an oxazolone
intermediate (3.1 g, 81% yield) that was used without further
purification. To the oxazolone (3.07 g, 10 mmol) was added 2%
NaOH in 70% aqueous ethanol (100 mL), and the suspension was
refluxed for 12 h. The reaction mixture was cooled to room
temperature, diluted with distilled water (∼50 mL) and titrated with
12 M HCl until precipitation of the product ceased. The mixture was
vacuum filtered, washed with distilled water, dried, and recrystallized
from ethanol/water (70:30, v/v) to obtain 3 g (85% yield) of the
desired product. 1H NMR (500 MHz, DMSO-d6) δ: 12.49 (br s, 1 H),
9.91 (s, 1 H), 9.77 (s, 1 H), 7.99 (d, J = 7.1 Hz, 2 H), 7.59 (t, J = 7.3
Hz, 1 H), 7.54 (dd, J = 3.4, 8.8 Hz, 3 H), 7.51 (d, J = 7.3 Hz, 1 H),
6.77 (d, J = 8.7 Hz, 2 H, 13C NMR (126 MHz, DMSO-d6) δ: 166.6,
165.9, 158.8, 134.0, 133.7, 131.9, 131.8, 128.5, 128.3, 127.7, 127.5,
124.6, 124.0, 115.5, 115.4. See SI for crystallographic data.
Synthesis of (Z)-2-Benzamido-[3-2H]-3-(4′-hydroxyphenyl)acrylic
Acid. The [3-2H]-acrylic acid isotopomer was synthesized analogously
to the unlabeled isomer (above), except [α-2H]-4′-Hydroxybenzalde-
hyde (0.62 g, 5 mmol) was used. Acetic anhydride (1.6 mL, 17 mmol),
K2HPO4 (0.9 g, 5 mmol), and hippuric acid (0.9 g, 5 mmol) were
varied to make the intermediate oxazolone (1.26 g, 82% yield). The
oxazolone (1.23 g, 4 mmol) was saponified under reflux with ethanolic
NaOH as before. The mixture was diluted with distilled water (25
mL), and the product was precipitated by the adding 12 M HCl at
room temperature. The suspension was worked up and the product
was recrystallized as described previously to obtain 1 g (88% yield) of
product. 1H NMR (500 MHz, DMSO-d6) δ: 12.49 (s, 1 H), 9.91 (s, 1
H), 9.77 (s, 1 H), 7.99 (d, J = 7.1 Hz, 1 H), 7.60 (t, J = 7.3 Hz, 1 H),
7.56−7.48 (m, 4 H), 6.77 (d, J = 8.8 Hz, 1 H). 13C NMR (126 MHz,
4′-O,3-N-Di(ethoxycarbonyl)-β-tyrosine Methyl Ester: isolated (Rf
= 0.54) at 46% yield (16 mg). HRMS: m/z 419.1770 ([M + H +
pyridine]+); calculated m/z 419.1818 ([C21H27N2O7]+). 1H NMR
(500 MHz, CDCl3) δ: 7.29 (d, J = 8.5 Hz, 2 H), 7.12 (d, J = 8.5 Hz, 1
H), 5.66 (br s., 1 H), 5.17−5.07 (m, 1 H), 4.28 (q, J = 6.9 Hz, 2 H),
4.08 (q, J = 7.3 Hz, 2 H), 3.60 (s, 3 H), 2.90−2.73 (m, 2 H), 1.36 (t, J
= 7.3 Hz, 3 H), 1.20 (t, J = 7.0 Hz, 3 H). 13C NMR (126 MHz,
CD3COCD3) δ: 170.7 (C(O)OCH3), 155.9 (OC(O)NH), 153.4
(OC(O)O), 150.5 (C4′), 140.3 (C1′), 128.4 (C2′), 121.7 (C3′), 65.0
(C(O)OCH3), 64.4 (CH3CH2OC(O)NH), 60.0 (CH3CH2OC(O)-
11201
dx.doi.org/10.1021/ja403918w | J. Am. Chem. Soc. 2013, 135, 11193−11204

Journal of the American Chemical Society
Article
DMSO-d6) δ: 166.6, 165.9, 158.8, 133.8, 131.9, 131.8, 131.6, 128.5,
128.4, 127.7, 127.5, 124.6, 123.9, 115.6, 115.4. See SI for crystallo-
graphic data.
= 8.5 Hz, 2 H), 6.93 (d, J = 8.5 Hz, 2 H), 3.07 (s, 1 H). 2H NMR (77
MHz, H2O, 512 scans, 25 °C) δ: 3.97 (bs, 1 H), 3.21 (bs, 1 H).
(2S,3R)-[3-2H]-1: 1H NMR (500 MHz, D2O, 32 scans, 25 °C) δ: 7.23
(d, J = 8.3 Hz, 2 H), 6.93 (d, J = 8.8 Hz, 2 H), 3.97 (d, J = 4.9 Hz, 1
H), 3.22 (d, J = 4.9 Hz, 1 H). 2H NMR (77 MHz, H2O, 512 scans, 25
°C) δ: 3.07 (bs, 1 H) (Figure S25 of SI).
Synthesis of Authentic 4′-O,3-N-di((S)-2-Methylbutanoyl)
Methyl Esters of (R)- and (S)-2. To a sample of (R)-and (S)-2
(0.5 μmol of each) dissolved in 50 mM phosphate buffer (1 mL) were
added pyridine (50 μL, 0.6 mmol) and (S)-2-methylbutyric anhydride
(10 μL, 5 μmol) at 0 °C. The reactions were stirred for 5 min and
treated with another batch of pyridine (50 μL) and (S)-2-
methylbutyric anhydride (10 μL) and stirred for 5 min. The mixtures
were each acidified (pH 2 with 6 M HCl) and extracted with diethyl
ether (3 × 2 mL). The ether fractions were combined, dried under
vacuum, and the residue was dissolved in methanol (100 μL). To this
solution was added a dilute diazomethane solution dissolved in diethyl
ether, until the yellow color persisted. The resultant 4′-O,3-N-di((S)-
2-methylbutanoyl)-β-tyrosine methyl ester diastereoisomers were
analyzed by GC/EIMS; their retention times and fragment ions are
noted in SI Figure S26.
Assessing the Stereospecificity of the Cβ-Hydrogen Ab-
straction Catalyzed by CcTAM. (2S,3S)-[2,3-2H2]- and (2S,3R)-
[3-2H]-1 (each at 1 mM) were separately incubated with CcTAM (0.1
mg) at 31 °C in 50 mM phosphate buffer (1 mL, pH 8.5) for 1 h. The
tyrosine isomers were derivatized in situ with ethyl chloroformate (50
μL, 0.5 mmol) and pyridine (50 μL, 0.6 mmol) to their 4′-O,N-
di(ethoxycarbonyl) derivatives. The samples were acidified to pH 2 (6
M HCl) and extracted into diethyl ether (1 mL). After treatment of
the ethanolic extract with excess diazomethane, the amino acid
derivatives were isolated as a mixture of ethyl and methyl esters and
analyzed by GC/EIMS, as before. Diagnostic fragment ions from the
ethyl ester derivative of the biosynthetic β-2 were analyzed to
determine the regiochemistry of the deuterium atoms.
Synthesis of (2S,3S)-[2,3-2H2]- and (2S,3R)-[3-2H]-1. The following
procedure is based on previously described methods.35,43,45 (Z)-2-
Benzamido-[3-2H]-3-(4′-hydroxyphenyl) acrylic acid (1 g, 3.5 mmol)
and 10 mg of [Rh(NBD)((R)-Prophos)]ClO4 were dissolved in dry
THF (25 mL) in a Parr reactor. The reactor was successively
evacuated and filled with H2 gas and kept under H2 gas (2.0 bar) for
16 h. The solvent was evaporated, followed by azeotropic removal of
residual THF with methanol. To the resultant yellow residue dissolved
in methanol (15 mL) was added dry Dowex 50 W (100−200 mesh)
cation exchange resin (1.5 g). The mixture was stirred until a clear
solution was obtained, and then filtered. The retained resin was
washed with warm methanol. All methanol filtrates were combined
and dried under vacuum to obtain the crude N-benzoyl-(2S,3R)-
[3-2H]-tyrosine (0.9 g). The product was then refluxed with 40% HBr
(aqueous) (10 mL) for 3 h. The mixture was cooled to room
temperature, and the benzoic acid crystals were removed by filtration.
The filtrate was washed with diethyl ether (3 × 20 mL) to remove
residual benzoic acid. The aqueous layer was lyophilized, and the crude
product was dissolved in 0.2 M NaOH (10 mL) and acidified with
acetic acid to yield crystals after 1h at 0 °C. The product was isolated
by vacuum filtration and the retentate was washed with ice cold water
to yield (2S,3R)-[3-2H]-1 as its zwitterionic amino acid (0.41 g, 70%
yield)
(2S,3S)-[2,3-2H2]-1 was synthesized by a procedure analogous to
that described for the synthesis of (2S,3R)-[3-2H]-1 with the following
exceptions. (Z)-2-Benzamido-3-(4′-hydroxyphenyl)acrylic acid was
used instead of the unlabeled isotopomer, and D2 gas was used in
place of H2 gas for the reduction step to obtain (2S,3S)-[2,3-2H2]-1 as
the zwitterionic amino acid (038 g, 65% yield).
Characterization of (2S,3S)-[2,3-2H2]- and (2S,3R)-[3-2H]-1.
(2S,3S)-[2,3-2H2]- and (2S,3R)-[3-2H]-1 (1 μmol of each) were
separately derivatized to their 4′-O,2-N-di(ethoxycarbonyl) derivatives
in water (1 mL) with ethyl chloroformate (50 μL, 0.5 mmol) and
pyridine (50 μL, 0.6 mmol), and the samples were acidified to pH 2 (6
M HCl) and extracted into diethyl ether (1 mL). After treatment of
the ethanolic extract with diazomethane, the derivatives were isolated
as a mixture of ethyl (10 mol %) and methyl (90 mol %) esters, as
described before. The methyl-esterified amino acid derivative was
analyzed by GC/EIMS. The [2H]-labeled fragment ions originating
from the identically derivatized (2S)-[3,3-2H2]-1 were used to
interpret the structures of various ions. The relative ratio of diagnostic
[2H]-labeled ions was compared to that of identical fragment ions of
the unlabeled (S)-1 derivative to calculate the deuterium enrichment
of labeled isotopomers: (2S,3S)-[2,3-2H2]-1 and (2S,3R)-[3-2H]-1
(Table S1 and Figures S21−S24 of SI).
Assessing the Stereospecificity of the Hydrogen Rebound
at Cα Catalyzed by CcTAM. (2S)-[3,3-2H2]-1 (5 mM) was incubated
with CcTAM (0.8 mg) at 31 °C in 50 mM phosphate buffer (40 mL,
pH 8.5) for 36 h. The incubation mixture was lyophilized, and the
2
remaining residue was dissolved in methanol (7 mL) for H NMR
analysis or CD3OD (7 mL) for 1H NMR analysis. As a control sample,
authentic unlabeled (R)-2 (2 mM) in 50 mM phosphate buffer (40
mL, pH 8.5) was lyophilized, and the remaining residue was dissolved
in CD3OD (7 mL) for 1H NMR analysis. The magnitude of the
1
coupling constants of the ABX spin system observed in the H NMR
for authentic unlabeled (R)-2 in CD3OD was used to assign the
chemical shifts of the protons at the prochiral center of 2. This data
was then used to assign the absolute stereochemistry at Cα of the
biosynthetic [2H]-2.
Assessing the Intramolecular Proton Transfer Step of the
CcTAM Reaction. (2S)-[3,3-2H2]-1 (1 mM) was incubated with
CcTAM (0.1 mg) at 31 °C in 50 mM phosphate buffer (10 mL, pH
8.5). Aliquots (1 mL) were withdrawn from the reaction mixture at 10
min, then at 2, 4, 7, 12, 22, 33, and 45 h. The samples were derivatized
in situ with ethyl chloroformate (50 μL, 0.5 mmol) and pyridine (50
μL, 0.6 mmol) to their 4′-O,N-di(ethoxycarbonyl) derivatives, acidified
to pH 2 (6 M HCl), and extracted into diethyl ether (1 mL). After
treatment of the ethanolic extract with diazomethane, the amino acid
derivatives were isolated as a mixture of ethyl and methyl esters.
During the derivatization, 4′-hydroxycinnamate was converted to its
4′-O-ethoxycarbonyl-(E)-coumaric acid as an ethyl and methyl ester
mixture. The methyl esters of the α-tyrosine and (E)-coumaric acid
derivatives and the ethyl ester of the β-tyrosine derivative were
analyzed by GC/EIMS (see SI Figures S6 and S7 for representative
diagnostic fragment ions observed for the derivatized biosynthetic 2
and 4′-hydroxycinnamate in this time course study). GC/EIMS was
used in selected-ion mode to calculate the ion abundance ratio of
[2H1](M)+ (m/z 354) and [2H2](M)+ (m/z 355) for the derivatives of
the biosynthetic deuterium labeled 2. This ratio informed on the D →
H exchange during the reaction. The ion abundance of [2H2](M)+ (m/
z 355) was corrected by subtracting the abundance of the naturally
The coupling constants (J) for the geminal protons (HA and HB)
against HX observed by 1H NMR for authentic (S)-1 in D2O were JAX
= 5.0 Hz at δ 3.23 (HA) and JBX = 7.8 Hz at δ 3.08 (HB). In a previous
study, the preferred conformation of (S)-1 in water was established by
NMR using stereospecifically deuterium isotopomers of 1 to correlate
chemical shift and the magnitude of coupling constants with the
conformation of the prochiral hydrogens.36 These earlier findings put
the dihedral angle between the carboxy group and the aromatic ring at
180°. In this conformation, HA and HX were separated by ∼60°, and
HB and HX by 180°. Using this earlier data along with the magnitude
of the J values36 observed for (S)-1, we assigned HA as pro-(3S) and
HB as pro-(3R). This data was used to assign the absolute
stereochemistry at Cβ of the synthetic (2S,3S)-[2,3-2H2]- and
(2S,3R)-[3-2H]-1.
Authentic (S)-1 (10 mg) was dissolved in D2O (1 mL), and
transferred to a 5 mm diameter NMR tube, and analyzed by 1H NMR
[(500 MHz, D2O, 32 scans, 25 °C) δ: 7.23 (d, J = 8.5 Hz, 2 H), 6.93
(d, J = 8.5 Hz, 2 H), 3.97 (dd, J = 5.1, 7.8 Hz, 1 H), 3.23 (dd, J = 5.0,
14.7 Hz, 1 H), 3.08 (dd, J = 7.8, 14.7 Hz, 1 H)]. (2S,3S)-[2,3-2H2]-
and (2S,3R)-[3-2H]-1 (10 mg of each) were separately dissolved in
1
2
D2O (and H2O) and analyzed by H NMR (and H NMR). (2S,3S)-
[2,3-2H2]-1: 1H NMR (500 MHz, D2O, 32 scans, 25 °C) δ: 7.23 (d, J
11202
dx.doi.org/10.1021/ja403918w | J. Am. Chem. Soc. 2013, 135, 11193−11204

Journal of the American Chemical Society
Article
occurring 13C-isotopomer (i.e., the [2H1](M + 1)+ (m/z 355) of the
molecular ion [2H1](M)+ (m/z 354)) of the singly deuterium-labeled
derivative.
(10) Gademann, K.; Kimmerlin, T.; Hoyer, D.; Seebach, D. J. Med.
Chem. 2001, 44, 2460−2468.
(11) Sonti, R.; Gopi, H. N.; Muddegowda, U.; Ragothama, S.;
Balaram, P. Chem.Eur. J. 2013, 19, 5955−5965.
Assessing the Effect of pH on CcTAM Stereoselectivity. (S)-1
(1 mM) was incubated with CcTAM (0.1 mg) separately at pHs 7, 8,
and 9 (6 mL of each) in 50 mM phosphate buffer at 31 °C. Aliquots (1
mL) were withdrawn at 1, 2.5, 5, 11, and 25 h. The amino acids were
derivatized in situ with (S)-2-methylbutyric anhydride to form the 4′-
O,3-N-di((S)-2-methylbutanoyl)-2 and methyl esterified with diazo-
methane and analyzed by GC/EIMS. The sum of ion abundances for
fragment ions m/z 278 [M − methyl butyl]+ (90% of base peak) and
m/z 194 [M − (2 × methyl butyl) + H]+ (base peak) for each
enantiomer was compared to calculate the ratio of (R)- and (S)-2 in
the sample.
Synthesis of (R)-2 Methyl Ester. In brief, to (R)-2 (18.1 mg, 0.1
mmol) dissolved in methanol (1 mL, 25 mmol) was added
trimethylsilyl chloride (25 μL, 0.2 mmol).46 The suspension was
stirred for 16 h and the methanol was evaporated in vacuo to obtain
the product (19.5 mg, quantitative yield). 1H NMR (500 MHz,
CD3OD) δ: 7.29 (d, 8.5 Hz, 2 H), 6.85 (d, 8.5 Hz, 2 H), 4.63 (dd, J =
6.3, 7.8 Hz, 1 H), 3.69 (s, 3 H), 3.09 (dd, J = 7.8, 17.0 Hz, 1 H), 2.98
(dd, J = 6.3, 17.0 Hz, 1 H).
(12) Murray, J. K.; Farooqi, B.; Sadowsky, J. D.; Scalf, M.; Freund, W.
A.; Smith, L. M.; Chen, J.; Gellman, S. H. J. Am. Chem. Soc. 2005, 127,
13271−13280.
(13) Ibrahem, I.; Rios, R.; Vesely, J.; Zhao, G.-L.; Cordova, A.
Synthesis 2008, 1153−1157.
(14) Davies, S. G.; Mulvaney, A. W.; Russell, A. J.; Smith, A. D.
Tetrahedron: Asymmetry 2007, 18, 1554−1566.
(15) Tan, C. Y. K.; Weaver, D. F. Tetrahedron 2002, 58, 7449−7461.
(16) Chesters, C.; Wilding, M.; Goodall, M.; Micklefield, J. Angew.
Chem., Int. Ed. 2012, 51, 4344−4348.
(17) Ratnayake, N. D.; Wanninayake, U.; Geiger, J. H.; Walker, K. D.
J. Am. Chem. Soc. 2011, 133, 8531−8533.
(18) Magarvey, N. A.; Fortin, P. D.; Thomas, P. M.; Kelleher, N. L.;
Walsh, C. T. ACS Chem. Biol. 2008, 3, 542−554.
(19) Mutatu, W.; Klettke, K. L.; Foster, C.; Walker, K. D.
Biochemistry 2007, 46, 9785−9794.
(20) Walker, K. D.; Klettke, K.; Akiyama, T.; Croteau, R. J. Biol.
Chem. 2004, 279, 53947−53954.
(21) Christianson, C. V.; Montavon, T. J.; Festin, G. M.; Cooke, H.
A.; Shen, B.; Bruner, S. D. J. Am. Chem. Soc. 2007, 129, 15744−15745.
(22) Christianson, C. V.; Montavon, T. J.; Van Lanen, S. G.; Shen, B.;
Bruner, S. D. Biochemistry 2007, 46, 7205−7214.
ASSOCIATED CONTENT
■
S
* Supporting Information
Gas chromatography and mass spectrometry data, NMR data,
D → H exchange plot, and CIF data. This material is available
free of charge via the Internet at http://pubs.acs.org.
(23) Krug, D.; Muller, R. ChemBioChem 2009, 10, 741−750.
̈
(24) Rother, D.; Poppe, L.; Viergutz, S.; Langer, B.; Retey, J. Eur. J.
Biochem. 2001, 268, 6011−6019.
(25) Asano, Y.; Kato, Y.; Levy, C.; Baker, P.; Rice, D. Biocatal.
Biotransform. 2004, 22, 131−138.
AUTHOR INFORMATION
■
(26) Rettig, M.; Sigrist, A.; Ret
2246−2265.
́
ey, J. Helv. Chim. Acta 2000, 83,
Corresponding Author
walker@chemistry.msu.edu
(27) Montavon, T. J.; Christianson, C. V.; Festin, G. M.; Shen, B.;
Bruner, S. D. Bioorg. Med. Chem. Lett. 2008, 18, 3099−3102.
(28) Feng, L.; Wanninayake, U.; Strom, S.; Geiger, J.; Walker, K. D.
Biochemistry 2011, 50, 2919−2930.
Notes
The authors declare no competing financial interest.
(29) Strom, S.; Wanninayake, U.; Ratnayake, N. D.; Walker, K. D.;
Geiger, J. H. Angew. Chem., Int. Ed. 2012, 51, 11−17.
(30) Christenson, S. D.; Wu, W.; Spies, M. A.; Shen, B.; Toney, M.
D. Biochemistry 2003, 42, 12708−12718.
(31) Sasse, F.; Kunze, B.; Gronewold, T. M. A.; Reichenbach, H. J.
Natl. Cancer Inst. 1998, 90, 1559−1563.
(32) Lambert, J. B.; Shurvell, H. F.; Lightner, D.; Cooks, R. G.
Organic Structural Spectroscopy; 1st ed.; Prentice Hall: NJ, 1998.
(33) Jensen, F. R.; Bushweller, C. H. J. Am. Chem. Soc. 1969, 91,
3223−3225.
ACKNOWLEDGMENTS
■
This material is based on work supported by NSF-CAREER
Award 0746432. We thank Dr. Richard Staples (Center for
Crystallographic Research, Michigan State University) for
providing the X-ray crystallographic data. The high-resolution
mass spectrometry was done by Dr. Gavin Reid at Michigan
State University, Department of Chemistry.
REFERENCES
■
(34) Garbisch, E. W.; Griffith, M. G. J. Am. Chem. Soc. 1968, 90,
6543−6544.
(35) Fryzuk, M. D.; Bosnich, B. J. Am. Chem. Soc. 1978, 100, 5491−
5494.
(1) Ahn, H. J.; Kim, Y. S.; Kim, J.-U.; Han, S. M.; Shin, J. W.; Yang,
H. O. J. Cell. Biochem. 2004, 91, 1043−1052.
(2) Garcion, E.; Lamprecht, A.; Heurtault, B.; Paillard, A.; Aubert-
Pouessel, A.; Denizot, B.; Menei, P.; Benoit, J. P. Mol. Cancer Ther.
2006, 5, 1710−1722.
(36) Oba, M.; Ueno, R.; Fukuoka, M.; Kainosho, M.; Nishiyama, K. J.
Chem. Soc., Perkin Trans. 1 1995, 1603−1609.
(3) Belyaev, N. A.; Kolesanova, E. F.; Kelesheva, L. F.; Rotanova, T.
V.; Panchenko, L. F. Bull. Exp. Biol. Med. 1990, 110, 1483−1485.
(4) Prabhakaran, P. C.; Woo, N. T.; Yorgey, P. S.; Gould, S. J. J. Am.
Chem. Soc. 1988, 110, 5785−5791.
(37) Kumar, S.; Dwevedi, A.; Kayastha, A. M. J. Mol. Catal. B: Enzym.
2009, 58, 138−145.
(38) Stingl, K.; Uhlemann, E. M.; Schmid, R.; Altendorf, K.; Bakker,
E. P. J. Bacteriol. 2002, 184, 3053−3060.
(5) Liu, W.; Christenson, S. D.; Standage, S.; Shen, B. Science 2002,
297, 1170−1173.
(39) Cooke, H. A.; Bruner, S. D. Biopolymers 2010, 93, 802−810.
(40) Katz, B. A.; Elrod, K.; Verner, E.; Mackman, R. L.; Luong, C.;
Shrader, W. D.; Sendzik, M.; Spencer, J. R.; Sprengeler, P. A.;
Kolesnikov, A.; Tai, V. W. F.; Hui, H. C.; Breitenbucher, J. G.; Allen,
D.; Janc, J. W. J. Mol. Biol. 2003, 329, 93−120.
(6) Yin, X. H.; O’Hare, T.; Gould, S. J.; Zabriskie, T. M. Gene 2003,
312, 215−224.
(7) Rachid, S.; Krug, D.; Weissman, K. J.; Muller, R. J. Biol. Chem.
̈
2007, 282, 21810−21817.
(41) Hanoian, P.; Sigala, P. A.; Herschlag, D.; Hammes-Schiffer, S.
Biochemistry 2010, 49, 10339−10348.
(8) Weaver, D. F.; Tan, C. Y. K.; Kim, S. T.; Kong, X.; Wei, L.;
Carran, J. R. (Queen’s University at Kingston, Canada; Neurochem
Inc.). Antiepileptogenic agents. Patent WO/2002/073208, March 13,
2002.
(42) Lyubimov, A. Y.; Lario, P. I.; Moustafa, I.; Vrielink, A. Nat.
Chem. Biol. 2006, 2, 259−264.
(43) Fryzuk, M. D.; Bosnich, B. J. Am. Chem. Soc. 1977, 99, 6262−
6267.
(9) Georg, G. I. The Organic Chemistry of β-Lactams; Wiley-VCH:
New York, 1993.
11203
dx.doi.org/10.1021/ja403918w | J. Am. Chem. Soc. 2013, 135, 11193−11204

Journal of the American Chemical Society
Article
(44) Cleary, T.; Brice, J.; Kennedy, N.; Chavez, F. Tetrahedron Lett.
2010, 51, 625−628.
(45) Kirby, G. W.; Michael, J. J. Chem. Soc., Perkin Trans. 1 1973,
115−120.
(46) Li, J.; Sha, Y. Molecules 2008, 13, 1111−1119.
11204
dx.doi.org/10.1021/ja403918w | J. Am. Chem. Soc. 2013, 135, 11193−11204