Intermolecular Amine Transfer to Enantioenriched trans-3Phenylglycidates by an α/β-Aminomutase to Access Both anti-Phenylserine Isomers

Prakash K. Shee, Honggao Yan, and Kevin D. Walker*

Research Article

Intermolecular Amine Transfer to Enantioenriched trans-3-

Phenylglycidates by an α/β-Aminomutase to Access Both anti-

Phenylserine Isomers

Cite This: ACS Catal. 2020, 10, 15071 −15082

Read Online

ACCESS

Metrics & More

Article Recommendations

sı Supporting Information

ABSTRACT: β-Hydroxy-α-amino acids are noncanonical amino

acids with two stereocenters and with useful applications in the

pharmaceutical and agrochemical sectors. Here, a 5-methylidene-

,5-dihydro-4H-imidazol-4-one-dependent aminomutase from

Taxus canadensis (TcPAM) was repurposed to transfer the amino

group irreversibly from (2S)-styryl-α-alanine to exogenously

supplied trans-3-phenylglycidate enantiomers, producing anti-

phenylserines stereoselectively. TcPAM catalysis inverted the

intrinsic regioselective chemistry from amination at C to C of

enantioenriched trans-3-phenylglycidates to make phenylserine

predominantly (97%) over phenylisoserine (∼3% relative abun-

dance). Gas chromatography−mass spectrometry analysis of the

chiral auxiliary derivatives of the biocatalyzed products conﬁrmed

that the amine transfer was stereoselective for each glycidate enantiomer. TcPAM converted (2S,3R)-3-phenylglycidate to (2S)-anti-

phenylserine predominantly (89%) and (2R,3S)-3-phenylglycidate to (2R)-anti-phenylserine (88%) over their antipodes, with

inversion of the conﬁguration at C in each case. Both glycidate enantiomers formed a small amount (∼10%) of syn-phenylserine by

retaining the conﬁguration at C . The minor syn-isomer likely came from a β-hydroxy oxiranone intermediate formed by

intramolecular ring opening of the oxirane ring by the carboxylate before amine transfer. TcPAM had a slight preference toward

−

−1

(

2S,3R)-3-phenylglycidate, which was turned over (k = 0.3 min ) 1.5 times faster than the (2R,3S)-glycidate (k = 0.2 min ).

app

−1 −1

The catalytic eﬃciencies (k_c_a_t/K ≈ 20 M s ) of TcPAM for the antipodes were similar. The kinetic data supported a two-

substrate ping-pong mechanism for the amination of the phenylglycidates, with competitive inhibition at higher glycidate substrate

concentrations.

KEYWORDS: biocatalysis, enzyme catalysis, MIO-aminomutase, 3-phenylglycidate, cinnamate epoxide, phenylserine, phenylisoserine,

β-hydroxy-α-amino acids

INTRODUCTION

extended to covert β-heteroaryl-α-amino acids to the

corresponding β-amino acids.

■

The 5-methylidene-3,5-dihydro-4H-imidazol-4-one (MIO)

family (EC 4.3.1.0 and 5.4.3.0) contains aminomutases

AMs) and ammonia lyases (ALs). MIO-AMs catalyze the

reversible cross-exchange of an amine group and a hydrogen

atom on the vicinal carbon of an amino acid, thus allowing the

interconversion between α-amino acids and their correspond-

ing β-amino acid isomers.

reversible conversion of an α- or β-amino acid to its

dehydroaminated acrylate (Figure 1A).

AMs have been engineered to convert aromatic α-amino

acids to the corresponding β-amino acids with higher

enantioselectivity over the wild-type enzyme.

applications of AMs include engineering by random muta-

Similarly, ALs have been repurposed to achieve a broader

substrate scope to convert acrylates to aromatic α- or β-amino

acids through regioselective amination directed by the

electron-donating or -withdrawing substituents on the aryl

(

ring. Enantioselective α-amination of arylacrylates was

achieved through active site mutagenesis of ALs from

Petroselinum crispum (PcPAL) and Anabaena variabilis

−7

The related ALs catalyze the

−14

Several MIO-

(

AvPAL). In contrast, a mutated AL from Streptomyces

15−17

Received: September 10, 2020

Revised: November 24, 2020

Additional

18−20

genesis of the active site (Figure 1B)

or repurposing using

a non-natural amine donor substrate (Figure 1C) to catalyze

the amination of various ring-substituted arylacrylates to

produce aromatic α- and β-amino acids. AMs have also been

XXXX American Chemical Society

5071

ACS Catal. 2020, 10, 15071−15082

Research Article

Figure 1. Reactions catalyzed by native or engineered MIO-dependent ALs and AMs that (A) make either the acrylate or β-amino acid

stereoisomers from the corresponding α-amino acid, where R = methyl, imidazyl, tolyl, or hydroxphenyl [the free- (MIO) and aminated-MIO

(

NH −MIO) moieties are shown]; (B) produce either enantiomer of α- and β-amino acid (a.a.) from an acrylate. ee: enantiomeric excess; and (C)

transfer an amine from a surrogate substrate (2S)-styryl-α-alanine (1) to an arylacrylate. Enzyme names are listed in the main text.

maritimus (EncP) and an AM from Taxus chinensis (TchPAM)

showed enantioselectivity toward β-amination of arylacrylates

mates. In this study, we looked to verify that the isomeric ratio

of the (2R)-anti and (2S)-anti-phenylserine products catalyzed

by TcPAM resulted from stereoselectivity for each isomer in

the racemic mixture and not from bond rotation within one of

the glycidate substrate isomers during catalysis. We also

established the correlation between the glycidate enantiomer

and the small amount of syn-phenylserine formed, which

provided valuable insight into the amine group transfer

1,14,24,25

(Figure 1B).

Besides employing MIO-enzymes to

make amino acids, other eﬀorts looked to convert α-amino

acids to their corresponding acrylates. For instance, active site

mutagenesis on TchPAM resulted in a 44-fold increase of the

PAL activity in the enantioselective deamination of (R)-β-

phenylalanine to cinnamic acid. Another example demon-

strated how the acyclic α-amino acid propargylglycine was

converted to its acrylate (E)-pent-2-en-4-ynoate by an MIO-

mechanism from the NH

during catalysis.

−MIO to the glycidate backbone

AL from Petroselinum crispum.

To further validate these hypotheses, here, we synthesized

enantioenriched 3-phenylglycidates from cinnamyl alcohol

Much of the mechanistic analysis involving MIO-enzymes

has centered on deaminating amino acids to their acrylates or

aminating acrylates to α- or β-amino acids. However, our

earlier work explored a new application of MIO-enzyme

chemistry, where we expanded the transaminase activity of

MIO-AMs to a new class of substrates, epoxides, producing

amino alcohols. We repurposed an MIO-dependent phenyl-

alanine AM from Taxus canadensis (TcPAM) to convert trans-

using the asymmetric Sharpless epoxidation reaction,

followed by RuCl -catalyzed oxidation of the primary alcohol

to the carboxylate. TcPAM was incubated with each 3-

phenylglycidate enantiomer, and the absolute conﬁgurations of

the biocatalyzed anti-phenylserine (major) and phenylisoserine

(minor) were evaluated to gain further insights into the

substrate enantioselectivity of TcPAM for aminating 3-

phenylglycidates. The apparent Michaelis−Menten kinetic

parameters of TcPAM in a competitive inhibition ping-pong

mechanism for each 3-phenylglycidate enantiomer were

measured.

-arylglycidates (i.e., trans-arylacrylate epoxides) to anti-

arylserines. TcPAM normally isomerizes (2S)-α-phenyl-

alanine to (3R)-β-phenylalanine

29,30

and was shown to have

broad substrate speciﬁcity for non-natural α-arylalanines

including the chain extended (2S)-styryl-α-alanine. In a

MATERIALS AND METHODS

kinetic study on TcPAM, the residence time of the NH −MIO

■

adduct was measured after (2S)-styryl-α-alanine donated its

amino group to the enzyme, and the (2E,4E)-styrylacrylate

Chemicals and Supplies. trans-Cinnamyl alcohol, (+)-di-

ethyl tartrate (DET), (−)-DET, sodium metaperiodate, tert-

butyl hydroperoxide solution (5.0−6.0 M) in nonane,

chlorotrimethylsilane, and (2S)-(+)-methylbutyric anhydride

were purchased from MilliporeSigma (Burlington, MA).

Titanium(IV) isopropoxide and ruthenium(III) chloride

hydrate were purchased from Oakwood Chemical (Estill,

SC). (2S)-Styryl-α-alanine and (2R,3R)-anti-phenylisoserine

were purchased from Chem Impex (Wood Dale, IL). (2S)-syn-

Phenylserine and (2R,3S)-phenylisoserine hydrochloride were

purchased from Bachem (Torrance, CA). All chemicals were

used without further puriﬁcation unless noted.

product was released. The burst-phase study revealed that

the aminated TcPAM (NH −MIO) lifetime was long enough

to transfer the amino group intermolecularly from (2S)-styryl-

α-alanine to various arylacrylates making enantiopure α- and β-

arylamino acids.

Encouraged by the intrinsic intermolecular transaminase

29,32

activities exhibited by MIO-PAMs,

we used TcPAM and

(

2S)-styryl-α-alanine to convert trans-3-arylglycidate racemates

to arylserines and arylisoserines. In our previous substrate

speciﬁcity study, we showed that TcPAM has broad substrate

speciﬁcity and stereoselectivity toward forming (2R)-anti-

arylserines over their (2S)-anti-isomers, ranging from 66:34 for

Synthesis of (2R,3R)-3-Phenylglycidol (3a). The follow-

ing synthesis is based on a reported procedure. A mixture of

′-F-phenylserine to 88:12 for 3′−NO −phenylserine. We

Ti(Oi-Pr) (288 mg, 1.01 mmol), (−)-DET (241 mg, 1.2

also used crystal structures of TcPAM in our previous study to

visualize the active site and infer the stereocontrol observed for

converting the trans-3-arylglycidates to their corresponding

mmol), and activated, powdered 4 Å MS (500 mg) was stirred

in anhydrous dichloromethane (CH Cl ) (20 mL) at −30 °C

for 30 min. The reaction mixture was incubated over 30 min

intervals before adding the next reagent, so the chiral Ti−DET

complex could form and coordinate with each subsequent

reactant at low temperature. Cinnamyl alcohol (2) (1.05 g, 7.8

8,33

anti-arylserines.

From these data, we hypothesized that two

active site tyrosine residues acted separately as general acids to

facilitate stereoselective ring opening of the glycidate race-

5072

ACS Catal. 2020, 10, 15071−15082

ACS Catalysis

Research Article

mmol) dissolved in anhydrous CH Cl (2 mL) was then

adjusted to pH 4.0 (6 M HCl). The aqueous layer was then

added, and the resulting mixture was stirred at −30 °C for 30

min. The mixture was treated with 5.5 M t-BuOOH in nonane

extracted with Et O (2 × 15 mL). The combined organic

layers were dried (Na SO ), and the solvent was removed

(

3.6 mL, 19.5 mmol), stirred for 16 h at −30 °C, warmed to 0

under vacuum to yield the carboxylic acid as a yellow oil. The

carboxylic acid was dissolved in acetone (5 mL) and water (0.5

C over 1 h, and poured into a freshly prepared solution of

FeSO (1.2 g) and tartaric acid (350 mg) in deionized H O (2

mL) and treated with KHCO (100 mg, 1 mmol) to convert

mL) precooled to 0 °C. The two-phase solution was stirred for

0 min, and the aqueous phase was separated and extracted

with diethyl ether (Et O) (2 × 10 mL). The combined organic

the acid to its potassium salt (4a) and avoid potential α-

decarboxylation.

The ﬁnal epoxide product mixture contained salt impurities;

thus, dioxane (10 μmol) was the internal standard used to

quantify the 3-phenylglycidate stock solutions in assay buﬀer

[50 mM NaH PO /Na HPO buﬀer (pH 8.0)]. Isolated yield:

layers were dried over anhydrous Na SO , ﬁltered, and

concentrated under vacuum to give a pale-yellow oil. The

product was dissolved in Et O (50 mL), cooled to 0 °C, and

treated with a 30% (w/v) solution of NaOH in saturated brine

71 mg (63%). H NMR (500 MHz, D O): δ 7.42−7.35 (m,

H), 3.97 (d, J = 2.5 Hz, 1H), 3.56 (d, J = 2.0 Hz, 1H).

(

1 mL) precooled to 0 °C. The two-phase mixture was stirred

for 1 h at rt, and the aqueous layer was separated and treated

NMR (126 MHz, D O): δ 175.4, 135.5, 128.9, 128.7, 126.0,

with Et O (2 × 10 mL). The combined organic extracts were

−

58.8, 57.4. HRMS (ESI-TOF) m/z: [M − K] calcd for

C H O , 163.0395; found, 163.0403.

dried (Na SO ) and concentrated under vacuum. The

resulting crude product was puriﬁed by ﬂash column

chromatography (silica gel, EtOAc/hexanes, 1:4) to yield

Synthesis of Potassium (2R,3S)-3-Phenylglycidate

4b) from 3b. The potassium salt of (2R,3S)-3-phenyl-

glycidate (4b) was prepared following the same procedure to

(

2R,3R)-3-phenylglycidol (3a) as a clear liquid. Isolated yield:

40 mg (38%); 92% ee [based on high-performance liquid

convert 3a to 4a. Isolated yield: 185 mg (68%). H NMR

chromatography (HPLC) separation]. [α] +49° (c 1.00,

(

500 MHz, D O): δ 7.42−7.35 (m, 5H), 3.97 (d, J = 2.5 Hz,

CHCl ), reported [α] +45.9° (c 1.00, CHCl₃). R_f= 0.5 in

H), 3.56 (d, J = 2.0 Hz, 1H). C NMR (126 MHz, D O): δ

silica gel thin-layer chromatography (TLC) (EtOAc/hexanes,

75.4, 135.5, 128.9, 128.7, 126.0, 58.8, 57.4. HRMS (ESI-

0:60). H NMR (500 MHz, CDCl ): δ 7.39−7.27 (m, 5H),

−

TOF) m/z: [M − K] calcd for C H O , 163.0395; found,

.06 (ddd, J = 12.7, 5.2, 2.4 Hz, 1H), 3.94 (d, J = 2.2 Hz, 1H),

.81 (ddd, J = 12.7, 7.9, 3.8 Hz, 1H), 3.23 (dt, J = 3.7, 2.3 Hz,

H), 1.79 (dd, J = 7.8, 5.2 Hz, 1H). ¹C NMR (126 MHz,

63.0402.

Enantiopurity of the Synthesized 3-Phenylglycidates.

The glycidates were suspended in H O (1 mL) at 0 °C and

CDCl ): δ 136.8, 128.7, 128.5, 125.9, 62.5, 61.3, 55.7.

titrated with 6 N HCl to pH 3. The resulting 3-phenylglycidic

acids were extracted into EtOAc (1 × 2 mL), and the organic

fractions were combined and treated with diazomethane (0.9

Synthesis of (2S,3S)-3-Phenylglycidol (3b). The follow-

ing synthesis is based on a reported procedure and is

analogous to that used to convert cinnamyl alcohol 2 to 3a,

except that (+)-DET was used instead of (−)-DET. The ﬁnal

glycidol product (3b) was puriﬁed by ﬂash column

chromatography (silica gel, EtOAc/hexanes, 1:4) to yield

equiv) dissolved in Et O. The resulting methyl esters were

used to measure the ee and infer the ee of each synthetic

potassium 3-phenylglycidate (0.05 mmol) enantiomer from

which the methyl esters were derived.

(

−

2S,3S)-3-phenylglycidol (3b) as a clear liquid. Isolated yield:

A stock of racemic potassium 3-phenylglycidate, synthesized

79 mg (33%); 90% ee (based on HPLC separation). [α]

earlier, was also converted to the corresponding methyl ester

48° (c 1.00, CHCl ), reported [α] −50.1° (c 1.00,

by reacting with diazomethane. An aliquot (1 μL) of the

racemic and enantioenriched methyl 3-phenylglycidates was

analyzed by gas chromatography/electron-impact mass spec-

trometry (GC/EI-MS) using an Agilent 6890N gas chromato-

graph equipped with a capillary chiral GC column (25 m ×

CHCl ). R_f= 0.5 on silica gel TLC (EtOAc/hexanes,

(

0:60). H NMR (500 MHz, CDCl ): δ 7.43−7.23 (m, 5H),

.04 (dd, J = 12.8, 2.4 Hz, 1H), 3.94 (d, J = 2.2 Hz, 1H), 3.78

dd, J = 12.8, 4.2 Hz, 1H), 3.31−3.16 (m, 2H). C NMR (126

MHz, CDCl ): δ 136.6, 128.5, 128.3, 125.7, 62.7, 61.4, 55.7.

0.25 mm × 0.39 mm; CP Chirasil-Dex CB, thickness 0.25 μm;

General Procedure to Characterize the Enantiomeric

Excess of Glycidols (3a and 3b). The enantiomeric excess

Agilent Technologies, Santa Clara, CA) with He as the carrier

gas (ﬂow rate, 1 mL/min). The injector port (at 250 °C) was

set to splitless injection mode. Each sample was injected using

an Agilent 7683 auto-sampler (Agilent, Atlanta, GA). The

initial column temperature started at 70 °C, then increased at

(

ee) was measured by dissolving 2 mg of the synthesized and

puriﬁed 3-phenylglycidols (3a or 3b) in Et O (2 mL), and an

aliquot (10 μL) of this solution was analyzed in HPLC

(

Agilent 1260) equipped with a chiral column (Chiralcel OD-

40 °C/min to 95 °C with a 7 min hold, ramped at 10 °C/min

H, 5μm, 4.6 mm × 150 mm) with i-PrOH/Hex (90:10) as the

solvent at a ﬂow rate of 1 mL/min.

to 150 °C, then increased by 30 °C/min to 175 °C, and

returned to 70 °C over 3 min. The gas chromatograph was

coupled to a mass-selective detector (Agilent, 5973 inert)

operated in electron-impact mode (70 eV ionization voltage).

All spectra were recorded in the mass range of 50−375 m/z to

analyze the methyl phenylglycidates.

Synthesis of Potassium (2S,3R)-3-Phenylglycidate

(

4a) from 3a. The following synthesis is based on a reported

method. RuCl ·H O (6 mg, 29 μM) was added to a biphasic

mixture of solvents [CCl (2 mL), acetonitrile (2 mL), and

water (3 mL)] containing 3a (0.2 g, 1.32 mmol), sodium

metaperiodate 0.85 g, 3.96 mmol), and sodium bicarbonate

Production of Phenylserine by TcPAM Biocatalysis. A

(2S)-styryl-α-alanine (1 mM) solution in assay buﬀer was

preincubated with TcPAM (100 μg/mL) for 2 min, and then a

potassium 3-phenylglycidate enantiomer (4a or 4b) (1 mM)

was added. The assay was incubated at 31 °C and gently mixed

on a rocking shaker (100 rpm) for 2.5 h. The reaction was

quenched with 10% formic acid (pH 3.0).

(

0.55 g, 6.6 mmol). The mixture was stirred for 72 h, then

additional RuCl ·H O (6 mg, 29 μM) and sodium

metaperiodate (0.32 g, 1.5 mmol) were added, and the

reaction was stirred for 24 h. CH Cl (8 mL) and then water

(

2 mL) were added to the sodium carboxylate solution. The

product mixture was cooled to 0 °C, and the water layer was

5073

ACS Catal. 2020, 10, 15071−15082

ACS Catalysis

Research Article

General Method for Chiral Auxiliary Derivatization of

Phenylserines and Phenylisoserines. A mixture of all four

stereoisomers of phenylserine (0.2 mmol), synthesized

phenylglycidate concentration (eq S4 of the Supporting

Information ) was used to calculate the true kinetic parameters

(K and k ) of TcPAM for each enantiomer.

cat

previously, was dissolved in assay buﬀer. To this solution

were added pyridine (50 μL, 0.62 mmol) and (2S)-2-

methylbutyric anhydride (50 μL, 0.25 mmol), and the reaction

mixture was stirred for 15 min at ∼25 °C. The pH was

adjusted to 3.0 (6 M HCl), and the N-protected phenylserine

was extracted with EtOAc (1 × 2 mL). The organic layer was

evaporated under a stream of nitrogen gas, and the resultant

residue was dissolved in 3:1 EtOAc/MeOH (v/v) (1 mL).

RESULTS AND DISCUSSION

Synthesis and Characterization of the Enantioen-

■

riched 3-Phenylglycidate Substrates (4a and 4b). The

enantiomeric potassium 3-phenylglycidates were made by

oxidizing cinnamyl alcohol with the Sharpless asymmetric

epoxidation catalyst using a (−)-DET or (+)-DET ligand to

make (2R,3R)-3-phenylglycidol (3a) or the (2S,3S)-isomer

(3b), respectively (Scheme 1). The absolute conﬁguration of

Diazomethane dissolved in Et O was added dropwise until the

yellow color persisted, and the solvent was removed under a

stream of nitrogen gas to obtain the methyl esters. The

resulting methyl ester was dissolved in CH Cl (1 mL), to

Scheme 1. Synthesis of (2S,3R)-3-Phenylglycidate (4a) and

(2R,3S)-3-Phenylglycidate (4b); (a) (−)-DET [Step a′ Used

which pyridine (70 μL, 0.86 mmol) and chlorotrimethylsilane

100 μL, 0.79 mmol) were added. The solution was stirred for

5 min at ∼25 °C, and the reaction was quenched with water

1 mL), and the organic layer was separated and analyzed by

(

+)-DET], Ti(Oi-Pr) , t-BuOOH, 4 Å Molecular Sieves,

(

CH Cl , 16 h, −30 °C; (b) (i) RuCl , NaIO , NaHCO ,

CH CN/CCl /H O, rt, 72 h.; (ii) KHCO , Acetone/H O, 0

C, 1 h

GC/EI-MS.

To access authentic standards of anti-phenylisoserine

enantiomers, enantioenriched potassium 3-phenylglycidates,

a and 4b, and a stock of racemic potassium 3-phenyl-

glycidates, characterized and synthesized previously, were

treated separately with NH OH (2 M) in assay buﬀer (1 mL)

for 1 h. The hydroxy amino acids derived from the synthetic

amination of the glycidates and authentic (2R,3R)-anti- and

(

2R,3S)-syn-phenylisoserine were then derivatized to their

2S)-2-methylbutyramide chiral auxiliaries, followed by methyl

each 3-phenylglycidol enantiomer was conﬁrmed by measuring

the corresponding speciﬁc rotation in chloroform and

comparing the measured values against those reported in the

esteriﬁcation and O-silyl etheriﬁcation, as done with the

phenylserines. Each derivatized phenylisoserine sample was

analyzed by GC/EI-MS on an Agilent 6890N gas chromato-

graph equipped with a capillary GC column (30 m × 0.25 mm

literature.

An asymmetric Sharpless epoxidation method employed

(−)-DET and resulted in making (2R,3R)-3-phenylglycidol

(3a), whereas the epoxidation with (+)-DET produced

(2S,3S)-3-phenylglycidol (3b). The enantiopurity of each

glycidol enantiomer was calculated by measuring the relative

peak areas of the glycidols separated on an HPLC ﬁtted with a

chiral column. Glycidol 3b eluted at 7.42 min (90% ee),

followed by its enantiomer 3a (8.42 min, 92% ee) (Figure S1

of the Supporting Information ).

0.25 μm + 5 m EZ-Guard column; VF-5MS; Agilent

Technologies, Santa Clara, CA) with He as the carrier gas

ﬂow rate, 1 mL/min). The injector port (at 250 °C) was set

(

to splitless injection mode. A 1 μL aliquot of each sample was

injected using an Agilent 7683 auto-sampler (Agilent, Atlanta,

GA). The initial column temperature started at 50 °C, then

increased at 50 °C/min to 150 °C, followed by 20 °C/min to

00 °C, then ramped at 10 °C/min to 225 °C with a 5 min

hold, and ﬁnally increased by 25 °C/min to 250 °C. The gas

chromatograph was coupled to a mass-selective detector

Glycidols 3a and 3b were oxidized to their carboxylic acids

using a RuCl /NaIO -based oxidation protocol (Scheme 1)

3 4

(

Agilent, 5973 inert) operated in electron-impact mode (70

under basic conditions (NaHCO ) to facilitate the deproto-

eV ionization voltage). All spectra were recorded in the mass

range of 50−375 m/z to analyze the hydroxy amino acids

derivatized as their O-trimethylsilyl N-[(2S)-2-methylbutyryl]

methyl ester.

nation of the carboxylic acid and prevent any potential

decomposition because of spontaneous decarboxylation.

Oxidation of 3a gave potassium (2S,3R)-3-phenylglycidate

(4a), while 3b was oxidized to the potassium (2R,3S)-3-

phenylglycidate (4b). In contrast to the 43 h reaction time

reported in the literature for the alcohol to carboxylate

oxidation, here, the oxidation reactions were run for 72 h to

convert the alcohol precursors fully to their enantioenriched 3-

phenylglycidates.

Kinetic Analysis. (2S)-Styryl-α-alanine, the amine donor,

was varied from 100 μM to 1 mM, at ﬁxed concentrations of

each epoxide enantiomer (100 μM, 500 μM, 1 mM, and 2.5

mM) in triplicate assays containing TcPAM (100 μg/mL) to

calculate the steady-state enzyme kinetic constants. The

reactions were terminated with 10% formic acid (pH 3.0),

and 3′-bromo-α-phenylalanine (50 nM) was added as an

internal standard. This reaction mixture was analyzed using a

liquid chromatography-electrospray ionization−multiple reac-

Enantiopurity of the Synthesized 3-Phenylglycidate

Isomers (4a and 4b). The ee of both (2S,3R)- (4a) and

(2R,3S)-3-phenylglycidate salts (4b) were measured by

converting them to their corresponding methyl esters using

CH N and characterized by GC/EI-MS ﬁtted with a chiral

tion monitoring (LC/ESI−MRM) method without further

app

chemical derivatization. The apparent kinetic parameters (K

column. The retention times of the eluted peaks for the

enantioenriched methyl 3-phenylglycidate isomers were

veriﬁed by comparing against the retention times of authentic

racemic methyl 3-phenylglycidate standards characterized in a

app

and k_c_a_t) were calculated by nonlinear regression with Origin

Pro 9.0 software (Northampton, MA) using a modi ﬁ ed

Michaelis −Menten equation (eq S1 of the Supporting

app

max

Information). A secondary plot of V

against the 3-

previous account (Figure 2). Comparing the relative peak

5074

ACS Catal. 2020, 10, 15071−15082

Research Article

Figure 3. Stereoisomerism convention for the (A) phenylserine and

B) phenylisoserine isomers.

(

separately with each enantioenriched 3-phenylglycidate (4a or

b) along with the amino group donor substrate (2S)-styryl-α-

alanine. Noticeably, TcPAM turned over 4a ∼1.5-fold faster to

its corresponding anti-phenylserine (5c) than 4b was turned

over to its anti-phenylserine (5d). The anti-phenylserines were

made by TcPAM preferentially over the syn-isomers (5a and

b) from 4a and 4b, respectively. TcPAM catalyzed the ring

opening of each 3-phenylglycidate primarily by directing the

amino group attack toward C to produce phenylserine as the

primary product. A small amount of anti-phenylisoserine (3%),

Figure 2. GC-EI/MS extracted-ion (m/z 121) chromatogram of (A)

racemic methyl 3-phenylglycidate (51:49), (B) methyl ester of

resulting from amination at C , was also made.

(

2S,3R)-3-phenylglycidate of (4a) (92% ee, 12.56 min, 4b at 4% at

2.37 min), and (C) methyl ester of (2R,3S)-3-phenylglycidate (4b)

90% ee, 12.37 min, 4a at 5% at 12.56 min).

The phenyl ring of 3-phenylglycidates intrinsically stabilizes

the partial positive charge (δ ) on the benzylic carbon (C )

created by the C −O bond dipole more than it stabilizes the

C −O bond dipole. The adjacent carboxylate functional group

areas of the glycidate methyl esters analyzed by GC/EI-MS

veriﬁed that the (2S,3R)-isomer was made at 92% ee, and its

reduces the magnitude of the C −O bond dipole. Thus, the

localized stabilization of the C −O bond dipole of phenyl-

(

2R,3S)-antipode was prepared at 90% ee. The ee values of the

glycidates typically promotes a nucleophilic attack at C . For

glycidates tracked those of the corresponding glycidols from

which they were synthesized.

example, nucleophilic epoxide ring opening by InCl -catalyzed

thiolysis, bromolysis, iodolysis, copper-catalyzed azidoly-

Glycidate Ring Opening by TcPAM Catalysis and

sis, and aminolysis occurred at C of 3-phenylglycidates. In

Analysis of Its Stereoselectivity. In our previous work,

this study, we observed a reversal of regioselectivity with

TcPAM as the biocatalyst and 1 as the amine nucleophile

donor. The 3-phenylglycidates were ring-opened by nucleo-

we reported that TcPAM could be repurposed to irreversibly

biocatalyze a reaction that transfers an amine intermolecularly

from (2S)-styryl-α-alanine to racemic 3-arylglycidates to access

arylserines and arylisoserines. Racemic glycidates were used to

explore the enantioselectivity of TcPAM when converting the

glycidates to their corresponding hydroxy amino acid products.

Surprisingly, TcPAM turned over each enantiomer of

arylglycidate to a mixture of anti-arylserine enantiomers as

the major biocatalytic products, with a modest stereo-

preference for the (2R)-anti isomer over its (2S)-anti antipode

philic cleavage of the C −O bond to produce phenylserine as

the major product made biocatalytically.

The relative and absolute stereoconﬁgurations of O-

trimethylsilyl N-[(2S)-2-methylbutyryl] methyl esters deriva-

tives of phenylserine and phenylisoserine, biocatalyzed by

TcPAM, were assessed. Authentic standards of phenylserine

chiral derivatives (Scheme 2A), made in a previous study,

were analyzed by a GC/EI-MS method (Figure 4A).

(

see Figure 3 for the naming convention). These results

An earlier study used (2S)-threonine aldolase-catalyzed

stereoisomeric resolution to establish the elution proﬁle of the

methyl ester, N-acyl chiral derivatives as (2S)-anti- (5c, 8.88

min), (2R)-anti-phenylserine (5d, 8.93 min), (2S)-syn- (5a,

9.02 min), and then (2R )-syn -phenylserine (5b, 9.05 min)

inspired us to examine whether the isomeric ratios of the (2R)-

anti- and (2S)-anti-arylserine products resulted from the

stereospeciﬁcity of TcPAM for a particular isomer of

arylglycidate in the racemic mixture and not from bond

rotation within one of the glycidate substrate isomers during

catalysis.

(Figures 4 and S2 of the Supporting Information).

41,42

Guided by accounts reported previously,

racemic 3-

The turnover (k^a_c_a^p_t^p ) of the trans-3-arylglycidate racemates

phenylglycidate was ﬁrst treated with 2 M NH OH to make a

−

ranged from 0.02 min for 3-(4′-chlorophenyl)glycidate to 1.3

mixture of anti-phenylisoserine enantiomers (6c and 6d) in situ

to establish the absolute stereoconﬁguration of phenylisoserine

diastereomers. Also, the enantioenriched 4a was treated with

NH OH to invert the conﬁguration at the C to make (2S,3S)-

−

min for 3-(3′-chlorophenyl)glycidate, and the turnover of

−

the 3-phenylglycidate was in-between at 0.4 min . Thus, we

examined the turnover of each trans-3-phenylglycidate

enantiomer by TcPAM to gain further insight into its

stereospeciﬁcity and stereoselectivity. TcPAM was incubated

41,42

anti-phenylisoserine (6c),

and enantioenriched 4b was

treated similarly to make (2R,3R)-anti-phenylisoserine (6d)

5075

ACS Catal. 2020, 10, 15071−15082

ACS Catalysis

analyzed by GC/EI-MS (Figure 5 ).

Research Article

Scheme 2. (A) Chiral Auxiliary Derivatization of Authentic

Phenylserine Diastereomers Using (a) (i) (2S)-2-

isoserine (6d) were derivatized with chiral auxiliaries and

Methylbutyric Anhydride, Pyridine, rt, 15 min; (ii) 6 M

HCl, pH 2; (iii) CH N , EtOAc/MeOH (3:1 v/v), rt, 10

min; and (iv) Chlorotrimethylsilane, Pyridine, CH Cl , rt,

15 min. (B) Synthetic Epoxy Ring Opening of 4a and 4b via

Step (b) 2 M aq. NH4OH, rt, 1 h to Make Phenylisoserine

(

PheIS) Diastereomers (2S,3S)-anti-(6c) and (2R,3R)-anti-

6d); (2R,3S)-syn-PheIS (6b), 6c, and 6d Were Derivatized

to the Methyl Ester Chiral Analogues as Described in Step

a) of (A)

(

Figure 5. GC/EI-MS extracted-ion (m/z 106) chromatograms of a

methyl ester, N-acyl chiral derivative of (A) authentic (2R,3S)-syn-

phenylisoserine (Bachem); (B) authentic (2R,3R)-anti-phenylisoser-

ine (Chem Impex); (C) anti-phenylisoserine enantiomers synthesized

41,42

from authentic racemic 3-phenylglycidate;

phenylisoserine synthesized from (2S,3R)-3-phenylglycidate (4a);

E) (2R,3R)-anti-phenylisoserine made from (2R,3S)-3-phenylglyci-

(D) (2S,3S)-anti-

(

date (4b); (F) TcPAM-catalyzed (2S,3S)-anti-phenylisoserine made

from (2S,3R)-3-phenylglycidate (4a); and (G) TcPAM-catalyzed

(

2R,3R)-anti-phenylisoserine made from (2R,3S)-3-phenylglycidate

4b). Isoserines in (C−E) were made from the corresponding

glycidates by adding 2M NH OH (aqueous) (see Scheme 2B).

The GC/EI-MS fragmentation patterns for the biocatalyzed

phenylisoserines were identical for the chiral auxiliary

derivatives of the commercial (2R,3S)-syn-, (2R,3R)-anti-,

and synthesized (2S,3S)-anti- and (2R,3R)-anti-phenylisoser-

ine isomers (Figures S3 and S4 of the Supporting

Information). The derivatives of the commercial (2R,3S)-syn-

phenylisoserine (6b) and (2R,3R)-anti-phenylisoserine (6d)

eluted at 9.05 and 9.33 min, respectively (Figure 5A,B),

whereas those of the anti-phenylisoserine racemate, prepared

from 3-phenylglycidate racemate, eluted at 9.33 and 9.41 min

(

Figure 5C).

These data show that the peak eluting at 9.41 min

corresponds to the derivatized (2S,3S)-anti-phenylisoserine

6c). The retention times of the anti-phenylisoserines in the

(

racemic mixture were identiﬁed by comparing them against the

retention times of enantioenriched (2S,3S)-anti-phenylisoser-

ine (6c, 9.41 min) (Figure 5D) and (2R,3R)-anti-phenyl-

isoserine (6d, 9.33 min) (Figure 5E) derivatized identically.

GC/EI-MS analysis of the chiral derivatives of the products

made by TcPAM from (2S,3R)-3-phenylglycidate (4a, 92% ee)

in the amine transfer reaction showed that (2R)-anti-

phenylserine was the major product (90%) with (2S)-syn-

isomer as the minor product (8%) (Figure 4C). The (2S)-syn-

phenylserine is believed to form when the carboxylate group of

the glycidate substrate nucleophilically and intramolecularly

Figure 4. GC/EI-MS extracted-ion (m/z 179) chromatograms of (A)

phenylserine stereoisomers (see Scheme 2A). Chiral derivatives of

(B) TcPAM-catalyzed phenylserine made from (2R,3S)-3-phenyl-

glycidate (4b) [the (2S)-anti:(2R)-anti:(2R)-syn isomers are at a

relative ratio of 76:15:9], (C) TcPAM-catalyzed phenylserine made

from (2S,3R)-3-phenylglycidate (4a) [the (2S)-anti:(2R)-anti:(2S)-

syn isomers are at a relative ratio of 2:90:8]; and (D) authentic (2S)-

syn-phenylserine (Bachem).

(

Scheme 2B). These synthesized anti-phenylisoserine isomers

and commercial (2R,3S)-syn- (6b) and (2R,3R)-anti-phenyl-

5076

ACS Catal. 2020, 10, 15071−15082

Research Article

Figure 6. Lowest-energy binding conformations calculated for (A) (R)-3-((R)-hydroxy(phenyl)methyl)oxiran-2-one (7a) (light gray sticks) and

B) (S)-3-((S)-hydroxy(phenyl)methyl)oxiran-2-one (7b) (sky blue sticks) made by intramolecular carboxylate-assisted oxiranone formation.

(

Proposed mechanisms for making (C) (2S)-syn-phenylserine (5a) from 4a via 7a and (D) (2R)-syn-phenylserine (5b) from 4b through 7b,

followed by amination at C by the NH −MIO. The active site residues include a catalytic Tyr80 (yellow sticks), a putative catalytic Tyr322

(

yellow sticks), binding contact Arg325 (orange sticks), and the methylidene imidazolone (MIO) moiety (green sticks). Heteroatoms are colored

red for oxygen and blue for nitrogen. The images were produced with UCSF Chimera, and the docking conformations were generated with

AutoDock Vina from TcPAM crystal structure (PDB code 3NZ4).

opens the epoxy ring and forms a three-membered oxiranone

intermediate, (R)-3-((R)-hydroxy(phenyl)methyl)oxiran-2-

Incubating the enantioenriched 3-phenylglycidates with 2 M

NH OH showed that nonenzymatic, regioselective ring

one (7a) (Figure 6A,B), before being aminated at the C by

opening took place predominantly (82%) at the β-carbon

(see Figure S5 of the Supporting Information), producing

phenylisoserine as the major product. This result showed that

nucleophilic attack on the phenylglycidate occurs inherently at

the NH −MIO adduct in the TcPAM active site. Also, a trace

amount of the (2R,3S)-3-phenylglycidate present at 4% in the

substrate (4a) was converted to (2S)-anti-phenylserine at a

relative abundance of 2% (Figure 4C).

the more electropositive benzylic carbon (C ). In contrast,

By comparison, the chiral derivatives of the products made

by TcPAM from (2R,3S)-3-phenylglycidate (4b, 90% ee)

showed that (2S)-anti-phenylserine was the major product at

only 76% relative abundance with the (2R)-syn-phenylserine

present at 9% abundance (Figure 4B). The (2R)-syn-

phenylserine byproduct is likely formed from the (2R,3S)-

substrate (4b) via a mechanism similar to that forming the

TcPAM used (2S)-styryl-α-alanine (1 mM) as an amine donor

and catalyzed the amination of enantioenriched 3-phenyl-

glycidates predominantly at the C to produce phenylserine. A

previous study showed that 1 mM NH OH did not stimulate

enzyme activity. It is worth noting, in its natural reaction,

TcPAM transfers the amino group from the NH −MIO adduct

equally to both C and C of cinnamate to make α- and β-

(

2S)-syn-isomer. We propose an analogous oxiranone inter-

amino acids. However, with 3-phenylglycidates, C amination

mediate [(S)-3-((S)-hydroxy(phenyl)methyl)oxiran-2-one

is substantially reduced, resulting in TcPAM converting

(2S,3R)-3-phenylglycidate (4a) and (2R,3S)-3-phenylglycidate

(4b) to small amounts of (2S,3S)-anti-phenylisoserine (3%)

(Figure 5F) and (2R,3R)-anti-phenylisoserine (1%) (Figure

5G), respectively.

(

7b)] before amination at C (Figure 6C,D).

It was interesting to ﬁnd that TcPAM converted the 5%

2S,3R)-3-phenylglycidate isomer present in the substrate to

2R)-anti-phenylserine at 15% relative abundance (Figure 4B).

(

The latter is compared to the 2% relative abundance of the

Computational Docking to Probe How the Anti- and

Syn-phenylserines Are Putatively Made. Protein−ligand

computational docking studies were performed to explain how

anti-phenylserine (major product) and syn-phenylserine

isomers (minor product) are biocatalyzed by TcPAM. The

crystal structure of TcPAM in complex with cinnamate [PDB:

3NZ4] was used to model structures of (2S,3R)- (4a) and

(

2S)-anti-phenylserine made from the 4% (2R,3S)-3-phenyl-

glycidate impurity in the previous assay (Figure 4C). These

results suggest that TcPAM has signiﬁcant stereospeciﬁc and

turnover preference for (2S,3R)-3-phenylglycidate over the

(

2R,3S)-enantiomer as competitive, enantiomeric substrates at

relatively low concentration.

5077

ACS Catal. 2020, 10, 15071−15082

ACS Catalysis

Research Article

Scheme 3. Proposed Mechanism of Phenylserine Biocatalysis

The pathway from E → P depicts the two-substrate ping-pong mechanism, where (2S)-styryl-α-alanine acts as the amino group donor. The

pathway involving E → ES shows the competitive substrate inhibition at a higher concentration (∼2500 μM) of 3-phenylglycidate. trans-3-PG:

trans-3-phenylglycidate; (2S)-SA: (2S)-styryl-α-alanine; (2E,4E)-SAcryl: (2E,4E)-styrylacrylate; E: enzyme (TcPAM); E*: modiﬁed enzyme; S :

substrate (n = 1, 2); (ES ): an enzyme−substrate complex (n = 1, 2); P : product (n = 1, 2). Inset 1: typical enzymatic ping-pong mechanism; Inset

2: competitive inhibition pathway.

(

2R,3S)-3-phenylglycidate (4b) and their corresponding

reaction. The docked conformation of 7b also placed C near

oxiranones 7a and 7b into the TcPAM active site using

the nitrogen atom of the NH −MIO priming the mechanism

AutoDock Vina by the method described in our previous

for favorable S 2 attack at the C , thus forming the (2R)-syn-

8,43

work.

phenylserine (Figure 6C,D).

Identifying low-energy conformations of 3-phenylglycidate

Proposed Mechanism and Derivation of the Kinetic

Equation. A two-substrate “ping-pong” mechanism is

proposed for the biocatalysis of phenylserine from a glycidate

and (2S)-styryl-α-alanine by TcPAM. The mechanism supports

a process where the enzyme reciprocates between a substrate

enantiomers modeled in the TcPAM active site has been

described in our previous work (Figure S6A,C of the

Supporting Information) that highlighted how the anti -

phenylserines might be made (Figure S6B,D of the Supporting

Information).

(

)

complex (E * S ) (where n = 1, 2, 3,...) and a standard state

49,50

We hypothesized that the (2S)-syn-phenylserine was

produced during TcPAM catalysis through an oxiranone (7a)

intermediate made by intramolecular carboxylate-assisted ring

opening of the glycidate. Insight into this proposal was made

by evaluating a low-energy docking conformation of 7a inside

the TcPAM active site calculated using the AutoDock Vina

(E* or E) (Scheme 3, inset 1).

Biocatalysis of phenylserine from 3-phenylglycidate by

TcPAM (E) involves two substrates, (2S)-styryl-α-alanine

(S ) and trans-3-phenylglycidate (S ), that bind the enzyme

forming separate enzyme−substrate complexes. We hypothe-

size that the amino donor substrate (2S)-styryl-α-alanine (S )

program. The oxiranone (OC−O) functional group is

placed into a salt−bridge complex with Arg325 (Figure 6B).

The calculated low-energy pose places the oxiranone oxygen

transfers its amino group to the MIO moiety of TcPAM,

forming a covalent N-alkylated adduct (ES ). Crystal

structures of analogous N-alkylated complexes of both α-

and β-phenylalanine have been reported previously for the

and the hydroxyl group close to Tyr80 and C of the oxiranone

close to the NH −MIO nitrogen. This docking pose suggests

mechanistically similar TchPAM (PDB:4C5R) and bacterial

that the Tyr80 residue, a highly conserved, general acid/base

PaPAM (PDB:3UNV) from Pantoea agglomerans. These

other MIO-enzyme structures suggest that a common N-

used during the natural α- to β-amino acid isomerase reaction

44−47

catalyzed by MIO-catalysts,

might help transition the

alkylated covalent mechanism is followed, where the NH of

hydroxy oxiranone intermediate to the (2S)-syn-phenylserine

isomer directly from intermediate 7a through proton-transfer

steps and S 2 attack of the NH −MIO at the C .

the substrate nucleophilically attacks the MIO moiety to form

an N-alkylated adduct.

The enzyme−substrate complex (ES ) is proposed to

Likewise, (S)-3-((S)-hydroxy(phenyl)methyl) oxiran-2-one

7b) is proposed as an intermediate to the minor product

2R)-syn-phenylserine made analogous to 7a. AutoDock

displace the H and the NH −MIO-enzyme adduct (E*) to

(

release (2E,4E)-styrylacrylate (P ) as the ﬁrst product. In an

earlier burst-phase kinetic study, P release was monitored,

Vina calculated a low-energy binding pose for 7b in complex

with TcPAM. While the antipode 7b was modeled in the active

site as the mirror image, it made binding interactions similar to

those made by 7a. The oxiranone formed a salt bridge with

Arg325, an oxiranone oxygen is near Tyr80, and the hydroxyl

group is positioned close to Tyr322. The latter two

interactions are possible catalytic hydrogen-bonding inter-

actions. It should be noted that Tyr322 is an essential residue

involved in forming the essential, catalytic MIO moiety and

where the lifetime of the aminated TcPAM (NH −MIO, E*)

was suﬃciently long enough to transfer the amine group to

exogenously supplied acceptor substrates.

After P is released, S (trans-3-phenylglycidate) likely binds

and covalently links to TcPAM as an NH −MIO adduct

(E*S ) by ring-opening the glycidate. This proposed covalent

adduct matches a previously reported crystal structure of a

homologous MIO-AM (SgTAM) from Streptomyces globispo-

rus, PDB:2RJR). SgTAM formed a similar covalent adduct

between the MIO moiety and an intermediate derived from 3-

(4′-ﬂuorophenyl)glycidate to form what was likely misinter-

keeping the NH −MIO adduct deprotonated so it functions as

a nucleophile during the normal TcPAM-catalyzed isomerase

5078

ACS Catal. 2020, 10, 15071−15082

Research Article

Figure 7. Modiﬁed Michaelis−Menten plots for the turnover of (A) (2S,3R)-3-phenylglycidate (4a) and (B) (2R,3S)-3-phenylglycidate (4b) to

their corresponding phenylserine at a ﬁxed concentration of the 3-phenylglycidate enantiomer. Double reciprocal Lineweaver−Burk plot of (C)

(

2S,3R)-3-phenylglycidate (4a) and (D) (2R,3S)-3-phenylglycidate (4b) to their corresponding phenylserine at a ﬁxed concentration of 3-

phenylglycidate enantiomer [100 μM (■), 500 μM (●), 1000 μM (▲), and 2500 μM (▼)]. n = 3 for plots (A−D), error bars represent one

app

standard deviation of uncertainty. (E) Secondary plot of V (calculated from each plot of Figure 7A) vs [4a], R = 0.87; and (F) secondary plot of

max

app

max

(calculated from each plot of Figure 7B) vs [4b] to calculate the intrinsic K_Mof the glycidate substrates, R = 0.95. The error bars for plots in

app

(

E,F) are propagated from the variance in v values calculated from the Michaelis−Menten plots in (A,B), respectively.

max

preted as an ether linkage (R′−O−R″) instead of an N-alkyl

rate of TcPAM to make phenylserine. The concentration of

(2S)-styryl-α-alanine was varied from 100 μM to 1 mM for

each of the four ﬁxed concentrations of the 3-phenylglycidate

linkage with the MIO.

Our proposed mechanism proceeds through proton-transfer

steps, and the major biocatalyzed product anti-phenylserine

(4a or 4b) (100 μM, 500 μM, 1 mM, and 2 mM). For each

app

(P₂) is released after the MIO is deaminated to regenerate the

concentration of the (2S,3R)-3-phenylglycidate (4a), V of

max

apo-enzyme E (Scheme 3). Assembly of this rational,

mechanistic pathway helped us derive a modiﬁed Michaelis−

Menten kinetic equation for the ping-pong mechanism to

support the hypothesis (eqs S3−S5 of the Supporting

Information ).

The phenylserine production rate was measured against the

concentration of the amino donor substrate, (2S)-styryl-α-

alanine, at diﬀerent ﬁxed concentrations of the 3-phenyl-

glycidate substrate (4a or 4b). To dissect the mechanism of

this sequential catalytic reaction pathway, the concentrations of

both substrates, (2S)-styryl-α-alanine and 3-phenylglycidate,

were independently varied to test their eﬀects on the turnover

TcPAM was calculated at ∼1 mM for (2 S )-styryl-α-alanine

(Figure 7A and eq S3 of the Supporting Information).

Kinetic Analyses. Calculating the Michaelis−Menten

kinetics of TcPAM for converting 3-phenylglycidate stereo-

isomers to phenylserine isomers was facilitated by a

quantitative LC/ESI−MRM method to enter values into the

derived rate equations (eq S3−S5 of the Supporting

Information).

The phenylserine production rate increased with increasing

(2S,3R)-3-phenylglycidate (4a) up to 1 mM, following a

4,55

traditional two-substrate ping-pong mechanism.

The

phenylserine production rate was signiﬁcantly reduced when

5079

ACS Catal. 2020, 10, 15071−15082

ACS Catalysis

https://pubs.acs.org/doi/10.1021/acscatal.0c03977.

Research Article

the glycidate 4a was at 2.5 mM (Figure 7A), suggesting

competitive substrate inhibition (Scheme 3, inset 2). Evidence

of 3-phenylglycidate inhibition on TcPAM catalysis was also

supported by a Lineweaver−Burk plot showing convergent Y-

intercepts for concentrations of 4a at 0.5 mM, 1 mM, and 2.5

transfer reaction using (2S)-styryl-α-alanine as the amine

donor and (2S,3R)- and (2R,3S)-3-phenylglycidate substrates

as the amino group acceptors. Under nonenzymatic conditions,

amine nucleophiles intrinsically cleave 3-phenylglycidates at

the electropositive C −O bond to make phenylisoserine more

app

41,42

mM that indicates an unchanged V

and increased K_M

abundantly over the phenylserine isomer.

max

(

Figure 7C), a hallmark of competitive inhibition.

In this study, the TcPAM catalytic space constraints forced

the amino nucleophile resourced from (2S)-styryl-α-alanine to

A similar kinetic trend was observed for substrate 4b, where

app

max

for TcPAM was calculated at ∼1 mM of (2S)-styryl-α-

preferentially cleave the C −O bond of each trans-3-phenyl-

alanine for each of the four concentrations of 4b (Figure 7B).

b also inhibited the phenylserine production rate at 2.5 mM

glycidate enantiomer to produce a single antipode of anti-

phenylserine preferentially. TcPAM prevented nucleophilic

app

max

(Figure 7D). V of TcPAM for (2S)-styryl-α-alanine (S₁),

attack at the C −O bond of the glycidates, and only 1−3%,

calculated from each kinetics plot measured at various ﬁxed

glycidate (S ) concentrations, were plotted against the

concentration of S (4a or 4b) to calculate the intrinsic K_M

and k _c _a _t of Tc PAM for 4a or 4b at steady state (eq S4 of the

Supporting Information ) (Table 1). The catalytic eﬃciency

relative abundance, of anti-phenylisoserines from each of the

trans-3-phenylglycidate enantiomers was made. TcPAM also

turned over each trans-phenylglycidate enantiomer to a single

antipode of syn-phenylserine (∼10% relative abundance),

which were observed as a mixture of syn-diastereoisomers in

an earlier study when racemic phenylglycidate was the

Table 1. Kinetics of TcPAM for the Turnover of 3-

Phenylglycidate Enantiomers to Phenylserine

substrate. The syn-stereoisomers are putatively made via an

oxiranone intermediate, formed from an intramolecular

carboxylate-assisted oxirane-cleavage. Computational modeling

of the proposed oxiranone showed that the intermediate

interacted with active site residues like those seen in other

MIO-enzyme structures with their natural phenylpropanoid

substrates. The modeled structures were poised for amination

entry

K_M(μM)

k_c_a_t(min⁻¹)

k /K (M⁻¹s⁻¹

)

K_I(μM)

cat

230 (127)

190 (54)

0.31 (0.04)

0.20 (0.01)

22.8

17.5

350

520

In addition to phenylserine, 4a and 4b are converted into 1 to 3%

phenylisoserine. Standard deviation in parenthesis (n = 3).

by the NH −MIO to yield the syn-phenylserines.

The biocatalytic approach described here provides an

alternative route to enantioenriched anti-phenylserine from a

trans-3-phenylglycidate enantiomer. This work extends the

transaminase activity of MIO-AMs to a new class of acceptor

molecules to produce bifunctional, hydroxy amino acids.

Equally interesting is that TcPAM converts the (2S,3R)-3-

phenylglycidate to the (2R)-anti-β-hydroxy-α-amino acid

enantiomer, where the amino group is positioned spatially

opposite to the orientation of proteinogenic amino acids. (2R)-

β-Hydroxy-α-amino acids are thus potentially important

building blocks of bioactive compounds that render them

resistant to proteases.

(

k /K ) of TcPAM for 4a appeared to be 1.3-fold higher than

cat M

that for 4b, primarily resulting from the 1.5-fold higher k_c_a_tfor

a over that for 4b. The inhibition constant (K ) for each of

the glycidate substrates was calculated from eq S5 of the

Supporting Information using a reported K value for (2S)-

styryl-α-alanine (105 μM). (2S,3R)-3-Phenylglycidate (4a)

was found to have a K value 1.5-times lower than its (2R,3S)-

enantiomer (4b), indicating that 4a binds TcPAM better than

b during the phenylserine production and likely aﬀects the

reaction turnover at higher concentrations by preventing the

binding of the amino donor (2S)-styryl-α-alanine (S ) to the

This exploration of the TcPAM mechanism also lays the

foundation for further exploration into active site mutagenesis

of TcPAM and other MIO-enzymes. These eﬀorts will focus on

achieving better enantioselectivity and expanding the substrate

selectivity to structurally more complex aryl and aliphatic

glycidates.

active site.

In our previous study with racemic 3-phenylglycidate as the

substrate, TcPAM made a mixture of (2R)-anti- and (2S)-anti-

phenylserine products, with the 2R-isomer (68% relative

abundance) predominating >2-fold over the 2S-isomer (32%

relative abundance), suggesting modest enantioselectivity. In

this study, we wanted to assess whether the observed isomeric

distribution of products resulted only from enantioselectivity,

from stereospeciﬁcity, or other combined eﬀects.

ASSOCIATED CONTENT

sı Supporting Information

■

Stereoisomeric phenylserines were made irreversibly by

TcPAM, which converted (2S,3R)-3-phenylglycidate to (2R)-

anti-phenylserine as the primary product 1.3 times faster than

it converted (2R,3S)-3-phenylglycidate to (2S)-anti-phenyl-

serine. This moderate substrate speciﬁcity (1.5-fold higher k_c_a_t)

calculated for the turnover of the phenylglycidates in this study

accounted principally for the ∼2-fold enantioselectivity

observed earlier for TcPAM with the racemic 3-phenylglycidate

The Supporting Information is available free of charge at

Spectroscopic data, materials and synthesis/character-

ization methods, NMR proﬁles, gas chromatography−

mass spectrometry proﬁles, and enzyme kinetics plots

(PDF )

substrate. This study also revealed that diﬀerences in

AUTHOR INFORMATION

thermodynamic parameters (K ) likely also contribute to the

higher enantioselectivity for the (2S,3R)-glycidate substrate

over its enantiomer.

■

Corresponding Author

Kevin D. Walker − Department of Chemistry and Department

of Biochemistry and Molecular Biology, Michigan State

University, East Lansing, Michigan 48824, United States;

orcid.org/0000-0001-5208-6692; Email: walker@

chemistry.msu.edu

CONCLUSIONS

TcPAM catalyzed the stereoselective amination of 3-phenyl-

glycidate enantiomers through a sequential ping-pong amino-

■

5080

ACS Catal. 2020, 10, 15071−15082

ACS Catalysis

(At PAL2): A Potent MIO-Enzyme for the Synthesis of Non-

Research Article

Authors

Canonical Aromatic α -Amino Acids: Part I: Comparative Character-

ization to the Enzymes From Petroselinum crispum (Pc PAL1) and

Rhodosporidium toruloides (Rt PAL). J. Biotechnol. 2017, 258, 148−

Prakash K. Shee − Department of Chemistry, Michigan State

University, East Lansing, Michigan 48824, United States;

orcid.org/0000-0003-1949-0858

Honggao Yan − Department of Biochemistry and Molecular

Biology, Michigan State University, East Lansing, Michigan

157.

(11) Lovelock, S. L.; Lloyd, R. C.; Turner, N. J. Phenylalanine

Ammonia Lyase Catalyzed Synthesis of Amino Acids by an MIO-

Cofactor Independent Pathway. Angew. Chem., Int. Ed. 2014, 53,

4652−4656.

8824, United States

Complete contact information is available at:

https://pubs.acs.org/10.1021/acscatal.0c03977

(12) Langer, B.; Langer, M.; Retey, J. Methylidene-Imidazolone

Forster, J.; Maury, J.; Borodina, I.; Nielsen, A. T. Highly Active and

MIO) from Histidine and Phenylalanine Ammonia-Lyase. Adv.

Protein Chem. 2001, 58, 175−214.

13) Jendresen, C. B.; Stahlhut, S. G.; Li, M.; Gaspar, P.; Siedler, S.;

Funding

(

The authors are grateful to the Michigan State University

Diversity Research Network: Launch Award Program for

support (GR100324-LAPKW).

Specific Tyrosine Ammonia-Lyases from Diverse Origins Enable

Enhanced Production of Aromatic Compounds in Bacteria and

Saccharomyces cerevisiae . Appl. Environ. Microbiol. 2015, 81, 4458.

(14) Weise, N. J.; Parmeggiani, F.; Ahmed, S. T.; Turner, N. J. The

Bacterial Ammonia Lyase EncP: A Tunable Biocatalyst for the

Synthesis of Unnatural Amino Acids. J. Am. Chem. Soc. 2015, 137,

Notes

The authors declare no competing ﬁnancial interest.

ACKNOWLEDGMENTS

■

(

2977−12983.

We thank the Michigan State University Research Technology

Support Facility Mass Spectrometry and Metabolomics Core

for their technical assistance. Special thanks to Prof. Borhan

and his research group in the Department of Chemistry at

Michigan State University for their technical support and

undergraduate Ariana Saifollahi for her help.

15) Zhou, L.; Wang, Y.; Liu, H.; Han, L.; Zhang, W.; Cui, W.; Liu,

Z.; Zhou, Z. Surface Engineering of a Pantoea agglomerans -Derived

Phenylalanine Aminomutase for the Improvement of (S )-β -Phenyl-

alanine Biosynthesis. Biochem. Biophys. Res. Commun. 2019, 518,

204−211.

(16) Zhu, L.; Feng, G.; Ge, F.; Tao, Y.; Song, P. Immobilized

Phenylalanine Aminomutase and Application in Synthesis of β-

Phenylalanine. CN 108707597 A, 2018.

REFERENCES

■

(17) Zhu, L.; Ge, F.; Li, W.; Song, P.; Tang, H.; Tao, Y.; Liu, Y.; Du,

1) Walker, K. D.; Floss, H. G. Detection of a Phenylalanine

Aminomutase in Cell-Free Extracts of Taxus brevifolia and Preliminary

G. One Step Synthesis of Unnatural β -Arylalanines Using Mutant

Phenylalanine Aminomutase From Taxus chinensis With High β -

Regioselectivity. Enzyme Microb. Technol. 2018, 114, 22−28.

Characterization of Its Reaction. J. Am. Chem. Soc. 1998, 120, 5333−

2) Christenson, S. D.; Liu, W.; Toney, M. D.; Shen, B. A Novel 4-

334.

(18) Wu, B.; Szymanski, W.; Wybenga, G. G.; Heberling, M. M.;

Bartsch, S.; de Wildeman, S.; Poelarends, G. J.; Feringa, B. L.;

Dijkstra, B. W.; Janssen, D. B. Mechanism-Inspired Engineering of

Phenylalanine Aminomutase for Enhanced β -Regioselective Asym-

metric Amination of Cinnamates. Angew. Chem., Int. Ed. 2012, 51,

Methylideneimidazole-5-one-Containing Tyrosine Aminomutase in

Enediyne Antitumor Antibiotic C-1027 Biosynthesis. J. Am. Chem.

Soc. 2003, 125, 6062−6063.

(3) Christianson, C. V.; Montavon, T. J.; Van Lanen, S. G.; Shen, B.;

(

82−486.

Bruner, S. D. The Structure of L-Tyrosine 2,3-Aminomutase from the

C-1027 Enediyne Antitumor Antibiotic Biosynthetic Pathway.

Biochemistry 2007, 46, 7205−7214.

19) Szymanski, W.; Wu, B.; Weiner, B.; de Wildeman, S.; Feringa,

B. L.; Janssen, D. B. Phenylalanine Aminomutase-Catalyzed Addition

of Ammonia to Substituted Cinnamic Acids: A Route to Enantiopure

α - and β -Amino Acids. J. Org. Chem. 2009, 74, 9152−9157.

(4) Huang, S.-X.; Lohman, J. R.; Huang, T.; Shen, B. A New

Member of the 4-Methylideneimidazole-5-One-Containing Amino-

mutase Family From the Enediyne Kedarcidin Biosynthetic Pathway.

Proc. Natl. Acad. Sci. U.S.A. 2013, 110, 8069 −8074.

(20) Wu, B.; Szymanski, W.; Wietzes, P.; de Wildeman, S.;

Poelarends, G. J.; Feringa, B. L.; Janssen, D. B. Enzymatic Synthesis

of Enantiopure α - and β -Amino Acids by Phenylalanine Amino-

mutase-Catalysed Amination of Cinnamic Acid Derivatives. Chem-

BioChem 2009, 10, 338−344.

5) Wanninayake, U.; Walker, K. D. A Bacterial Tyrosine

Aminomutase Proceeds Through Retention or Inversion of Stereo-

chemistry to Catalyze Its Isomerization Reaction. J. Am. Chem. Soc.

(21) Wanninayake, U.; Deporre, Y.; Ondari, M.; Walker, K. D. (S )-

(

013, 135, 11193−11204.

Styryl-α -Alanine Used to Probe the Intermolecular Mechanism of an

Intramolecular MIO-Aminomutase. Biochemistry 2011, 50, 10082−

10090.

(22) Ratnayake, N. D.; Theisen, C.; Walter, T.; Walker, K. D.

Whole-Cell Biocatalytic Production of Variously Substituted β -Aryl-

and β -Heteroaryl-β -Amino Acids. J. Biotechnol. 2016 , 217 , 12 −21.

(23) Lovelock, S. L.; Turner, N. J. Bacterial Anabaena variabilis

phenylalanine ammonia lyase: a biocatalyst with broad substrate

specificity. Bioorg. Med. Chem. 2014, 22, 5555−5557.

6) Yan, J.; Aboshi, T.; Teraishi, M.; Strickler, S. R.; Spindel, J. E.;

Tung, C.-W.; Takata, R.; Matsumoto, F.; Maesaka, Y.; McCouch, S.

R.; Okumoto, Y.; Mori, N.; Jander, G. The Tyrosine Aminomutase

TAM1 Is Required for β -Tyrosine Biosynthesis in Rice. Plant Cell

7) Walter, T.; King, Z.; Walker, K. D. A Tyrosine Aminomutase

015, 27, 1265−1278.

from Rice (Oryza sativa ) Isomerizes (S )-α - to (R )-β -Tyrosine with

Unique High Enantioselectivity and Retention of Configuration.

Biochemistry 2016, 55, 1−4.

(

8) Davidson, V. L. 7.19Protein-Derived Cofactors. In Compre-

(24) Parmeggiani, F.; Lovelock, S. L.; Weise, N. J.; Ahmed, S. T.;

Turner, N. J. Synthesis of D- and L-Phenylalanine Derivatives by

Phenylalanine Ammonia Lyases: A Multienzymatic Cascade Process.

Angew. Chem., Int. Ed. 2015, 54, 4608−4611.

hensive Natural Products II; Liu, H.-W., Mander, L., Eds.; Elsevier:

Oxford, 2010; pp 675-710.

(9) Dressen, A.; Hilberath, T.; Mackfeld, U.; Rudat, J.; Pohl, M.

Phenylalanine Ammonia Lyase From Arabidopsis thaliana (At PAL2):

A Potent MIO-Enzyme for the Synthesis of Non-Canonical Aromatic

α -Amino Acids. Part II: Application in Different Reactor Concepts for

the Production of (S )-2-Chloro-phenylalanine. J. Biotechnol. 2017,

(25) Filip, A.; Nagy, E. Z. A.; Tork, S. D.; Banoczi, G.; Tos a ̧ , M. I.;

́ ́

Irimie, F. D.; Poppe, L.; Paizs, C.; Bencze, L. C. Tailored Mutants of

Phenylalanine Ammonia-Lyase from Petroselinum crispum for the

Synthesis of Bulky L- and D-Arylalanines. ChemCatChem 2018, 10,

2627−2633.

(

58, 158−166.

10) Dressen, A.; Hilberath, T.; Mackfeld, U.; Billmeier, A.; Rudat,

J.; Pohl, M. Phenylalanine Ammonia Lyase From Arabidopsis thaliana

(26) Bartsch, S.; Wybenga, G. G.; Jansen, M.; Heberling, M. M.; Wu,

B.; Dijkstra, B. W.; Janssen, D. B. Redesign of a Phenylalanine

5081

ACS Catal. 2020, 10, 15071−15082

ACS Catalysis

Kokai, E.; Szilagyi, A.; Vertessy, B. G.; Farkas, O.; Paizs, C.; Poppe, L.

Research Article

Aminomutase into a Phenylalanine Ammonia Lyase. ChemCatChem

013, 5, 1797−1802.

27) Weiser, D.; Bencze, L. C.; Ban

Phenylalanine Ammonia-Lyase-Catalyzed Deamination of an Acyclic

Amino Acid: Enzyme Mechanistic Studies Aided by a Novel

Microreactor Filled with Magnetic Nanoparticles. ChemBioChem

Amino Acids Incorporated in Secondary Metabolites. Biopolymers

2010, 93, 802−810.

(

zi, G.; Ender, F.; Kiss, R.;

(45) Rother, D.; Poppe, L.; Viergutz, S.; Langer, B.; Retey, J.

Characterization of the active site of Histidine Ammonia-Lyase from

Pseudomonas putida . Eur. J. Biochem. 2001, 268, 6011−6019.

(46) Ro

ther, D.; Poppe, L.; Morlock, G.; Viergutz, S.; Retey, J. An

Active Site Homology Model of Phenylalanine Ammonia-Lyase from

Petroselinum crispum . Eur. J. Biochem. 2002, 269, 3065−3075.

(

015, 16, 2283−2288.

(47) Schroeder, A. C.; Kumaran, S.; Hicks, L. M.; Cahoon, R. E.;

28) Shee, P. K.; Ratnayake, N. D.; Walter, T.; Goethe, O.;

Halls, C.; Yu, O.; Jez, J. M. Contributions of Conserved Serine and

Tyrosine Residues to Catalysis, Ligand Binding, and Cofactor

Processing in the Active Site of Tyrosine Ammonia Lyase.

Phytochemistry 2008, 69, 1496−1506.

Onyeozili, E. N.; Walker, K. D. Exploring the Scope of an α /β -

Aminomutase for the Amination of Cinnamate Epoxides to

Arylserines and Arylisoserines. ACS Catal. 2019, 9, 7418−7430.

(29) Ratnayake, N. D.; Wanninayake, U.; Geiger, J. H.; Walker, K.

(48) Wybenga, G. G.; Szymanski, W.; Wu, B.; Feringa, B. L.;

D. Stereochemistry and Mechanism of a Microbial Phenylalanine

Janssen, D. B.; Dijkstra, B. W. Structural Investigations into the

Stereochemistry and Activity of a Phenylalanine-2,3-Aminomutase

from Taxus chinensis . Biochemistry 2014 , 53 , 3187 −3198.

Aminomutase. J. Am. Chem. Soc. 2011, 133, 8531−8533.

(30) Strom, S.; Wanninayake, U.; Ratnayake, N. D.; Walker, K. D.;

Geiger, J. H. Insights into the Mechanistic Pathway of the Pantoea

agglomerans Phenylalanine Aminomutase. Angew. Chem., Int. Ed. 2012,

(49) Ulusu, N. N. Evolution of Enzyme Kinetic Mechanisms. J. Mol.

Evol. 2015, 80, 251−257.

31) Wanninayake, U.; Walker, K. D. Assessing the Deamination

1, 2898−2902.

(50) Segel, I. H. Enzyme Kinetics: Behavior and Analysis of Rapid

Equilibrium and Steady-State Enzyme Systems; Wiley & Sons, Inc.:

Hoboken, NJ, 1993.

Rate of a Covalent Aminomutase Adduct by Burst Phase Analysis.

Biochemistry 2012, 51, 5226−5228.

(51) Wybenga, G. G.; Szymanski, W.; Wu, B.; Feringa, B. L.;

(32) Walter, T.; Wijewardena, D.; Walker, K. D. Mutation of Aryl

Janssen, D. B.; Dijkstra, B. W. Structural Investigations into the

Stereochemistry and Activity of a Phenylalanine-2,3-aminomutase

from Taxus chinensis . Biochemistry 2014, 53, 3187−3198.

Binding Pocket Residues Results in an Unexpected Activity Switch in

an Oryza sativa Tyrosine Aminomutase. Biochemistry 2016, 55, 3497−

33) Feng, L.; Wanninayake, U.; Strom, S.; Geiger, J.; Walker, K. D.

503.

(52) Pettersen, E. F.; Goddard, T. D.; Huang, C. C.; Couch, G. S.;

Greenblatt, D. M.; Meng, E. C.; Ferrin, T. E. UCSF Chimera A

Visualization System for Exploratory Research and Analysis. J.

Comput. Chem. 2004, 25, 1605−1612.

Mechanistic, Mutational, and Structural Evaluation of a Taxus

Phenylalanine Aminomutase. Biochemistry 2011, 50, 2919−2930.

(34) Yadav, J. S.; Reddy, M. S.; Rao, P. P.; Prasad, A. R.

(53) Montavon, T. J.; Christianson, C. V.; Festin, G. M.; Shen, B.;

Enantioselective Synthesis of (+)-Sedamine and (-)-Allosedamine.

Bruner, S. D. Design and Characterization of Mechanism-Based

Inhibitors for the Tyrosine Aminomutase Sg TAM. Bioorg. Med. Chem.

Lett. 2008, 18, 3099−3102.

Synthesis 2006, 2006, 4005−4012.

(35) Nandy, J. P.; Prabhakaran, E. N.; Kumar, S. K.; Kunwar, A. C.;

Iqbal, J. Reverse Turn Induced π -Facial Selectivity during Polyaniline-

Supported Cobalt(II) Salen Catalyzed Aerobic Epoxidation of N -

Cinnamoyl L-Proline Derived Peptides. J. Org. Chem. 2003, 68,

54) Goodenough-Lashua, D. M.; Garcia, G. A. tRNA −Guanine

Transglycosylase From E. coli : A Ping-Pong Kinetic Mechanism Is

Consistent With Nucleophilic Catalysis. Bioorg. Chem. 2003, 31, 331−

344.

36) Yadav, J. S.; Raju, A. K.; Rao, P. P.; Rajaiah, G. Highly

679−1692.

(55) Stein, R. L. Kinetics of Two-Substrate Enzymatic Reactions.

Kinetics of Enzyme Action: Essential Principles for Drug Hunters; John

Wiley & Sons, Inc: Hoboken, NJ, 2011; pp 141−168.

Stereoselective Synthesis of Antitumor Agents: Both Enantiomers of

Goniothales Diol, Altholactone, and Isoaltholactone. Tetrahedron:

Asymmetry 2005, 16, 3283−3290.

(37) Fringuelli, F.; Pizzo, F.; Tortoioli, S.; Vaccaro, L. InCl ₃ -

Catalyzed Regio- and Stereoselective Thiolysis of α ,β -Epoxycarboxylic

Acids in Water. Org. Lett. 2005, 7, 4411−4414.

(38) Amantini, D.; Fringuelli, F.; Pizzo, F.; Vaccaro, L. Bromolysis

and Iodolysis of α ,β -Epoxycarboxylic Acids in Water Catalyzed by

Indium Halides. J. Org. Chem. 2001, 66, 4463−4467.

(39) Fringuelli, F.; Pizzo, F.; Rucci, M.; Vaccaro, L. First One-Pot

Copper-Catalyzed Synthesis of α -Hydroxy-β -Amino Acids in Water. A

New Protocol for Preparation of Optically Active Norstatines. J. Org.

Chem. 2003, 68, 7041−7045.

(40) Tabarki, M. A.; Besbes, R. Regioselective Ring Opening of β -

Phenylglycidate and Aziridine-2-Carboxylates With N -Alkylhydroxyl-

amines: Synthesis of Isoxazolidinones. Tetrahedron 2014, 70, 1060−

41) Srivastava, R. P.; McChesney, J. D. A Practical and Inexpensive

064.

Synthesis of the Taxol C-13 Side Chain; N -Benzoyl-(2 R ,3 S )-3-

Phenylisoserine. Nat. Prod. Lett. 1995, 6, 147−152.

(

42) Wilding, B.; Vesela,

A. B.; Perry, J. J. B.; Black, G. W.; Zhang,

M.; Martínkova,

L.; Klempier, N. An Investigation of Nitrile

Transforming Enzymes in the Chemo-Enzymatic Synthesis of the

Taxol Sidechain. Org. Biomol. Chem. 2015 , 13 , 7803 −7812.

(43) Trott, O.; Olson, A. J. AutoDock Vina: Improving the Speed

and Accuracy of Docking with a New Scoring Function, Efficient

Optimization, and Multithreading. J. Comput. Chem. 2010, 31, 455−

44) Cooke, H. A.; Bruner, S. D. Probing the Active Site of MIO-

61.

Dependent Aminomutases, Key Catalysts in the Biosynthesis of β -

5082