A Tyrosine Aminomutase from Rice (Oryza sativa) Isomerizes (S)-α- to (R)-β-Tyrosine with Unique High Enantioselectivity and Retention of Configuration

Article abstract of DOI:10.1021/acs.biochem.5b01331

A recently discovered 3,5-dihydro-5-methylidene-4H-imidazol-4-one (MIO)-dependent tyrosine aminomutase (OsTAM) from rice [Yan, J., et al. (2015) Plant Cell 27, 1265] converts (S)-α-tyrosine to a mixture of (R)- and (S)-β-tyrosines, with high (94%) enantiomeric excess, which does not change with pH, like it does for two bacterial TAMs. The K_M of 490 μM and the k_cat of 0.005 s^-1 are similar for other TAM enzymes. OsTAM is unique and also catalyzes (R)-β- from (S)-α-phenylalanine. OsTAM principally retains the configuration at the reactive C_α and C_β centers during catalysis much like the phenylalanine aminomutase on the Taxol biosynthetic pathway in Taxus plants.

Full text of DOI:10.1021/acs.biochem.5b01331

Rapid Report
pubs.acs.org/biochemistry
A Tyrosine Aminomutase from Rice (Oryza sativa) Isomerizes (S)‑α- to
(R)‑β-Tyrosine with Unique High Enantioselectivity and Retention of
Configuration
Tyler Walter,† Zayna King,§ and Kevin D. Walker*,†,‡
‡
†Department of Chemistry and Department of Biochemistry and Molecular Biology, Michigan State University, East Lansing,
Michigan 48824, United States
§Medgar Evers College, Brooklyn, New York 11225, United States
S
* Supporting Information
ABSTRACT: A recently discovered 3,5-dihydro-5-meth-
ylidene-4H-imidazol-4-one (MIO)-dependent tyrosine
aminomutase (OsTAM) from rice [Yan, J., et al. (2015)
Plant Cell 27, 1265] converts (S)-α-tyrosine to a mixture
of (R)- and (S)-β-tyrosines, with high (94%) enantiomeric
excess, which does not change with pH, like it does for two
bacterial TAMs. The KM of 490 μM and the kcat of 0.005
s−1 are similar for other TAM enzymes. OsTAM is unique
Figure 1. Exchange of the NH2/H pair by retention of configuration
and also catalyzes (R)-β- from (S)-α-phenylalanine.
OsTAM principally retains the configuration at the reactive
Cα and Cβ centers during catalysis much like the
phenylalanine aminomutase on the Taxol biosynthetic
pathway in Taxus plants.
(ROC, top route) and inversion of configuration (IOC, bottom route).
tyrosine was observed in the earlier study, which is atypical for
other TAMs and indicated a potential fourth subclass of MIO-
AMs. Considering the preponderance of TAM catalysts that
produce a mixture of enantiomers of β-tyrosine, herein, we
reevaluated the product stereochemistry of OsTAM and
reassessed whether p-coumarate is made as a byproduct.
Also, an earlier study showed that the enantiomeric ratio of
β-tyrosine products changed with pH during catalysis by
CcTAM from Chondromyces crocatus.8 We describe the effects
of pH on the enantiomeric ratio catalyzed by OsTAM. In
addition, we calculated the intrinsic kinetic parameters of the
recombinantly expressed enzyme and dissected the stereo-
chemistry of the isomerization mechanism using stereospecifi-
class I lyase-like family of aminomutases (AMs) uses a
A3,5-dihydro-5-methylidene-4H-imidazol-4-one (MIO) ac-
tive site moiety to catalyze α- to β-amino acid isomerization
reactions. These catalysts are found on biosynthetic pathways
to biologically active natural products that contain a β-amino
acid building block.1−5 The MIO facilitates the α,β-elimination
of the NH2/H pair from the β-aryl-α-amino acids and forms an
NH2-MIO adduct, a tightly bound acrylate intermediate, and
transfers a proton to a catalytic tyrosine.6 As our knowledge of
the stereochemical mechanism of MIO enzymes improves,1,3,7,8
three subclasses begin to emerge. One class rotates the acrylate
intermediate and places the si or re face toward the NH2-MIO.
One rotamer rebounds back to the (S)-α-amino acid substrate,
and the other commits to stereoselective catalysis of the (R)-β-
amino acid. After the acrylate rotates, the NH2/H pair rebound
to the face of the acrylate opposite the one from which it was
removed, resulting in retention of configuration (ROC) (Figure
1). A second subclass impedes the rotation of the acrylate, and
the NH2/H pair rebounds to the face of the acrylate from
which it was originally attached, proceeding with inversion of
configuration (IOC) to make the (S)-β-amino acid (Figure 1).
Members of the third subclass, usually tyrosine aminomutases
(TAMs), use both ROC and IOC mechanisms, making a
mixture of (R)- and (S)-β-tyrosines, typically with one
enantiomer predominating.5
2
cally H-labeled tyrosines.
Enzyme Expression and Assay. OsTAM was cloned and
expressed as an N-terminal His6 fusion from the pET-28a(+)
vector in Escherichia coli BL21 (DE3) cells. The enzyme (690
amino acids, 75 kDa) was purified (Ni-affinity and gel-filtration
chromatographies) and concentrated by size-selective filtration
(30 kDa molecular weight cutoff) to 1.8 mg/mL (∼80% pure
by SDS−PAGE and Coomassie Blue staining). After OsTAM
(0.2 mg) had been incubated with (S)-α-tyrosine (1 mM) for 3
h, the reaction mixture was basified to pH 12, the tyrosines
were converted to their 4′-O,3-N-di(ethoxycarbonyl) ethyl
esters, and the p-coumarate byproduct was derivatized to its 4′-
O-ethoxycarbonyl ethyl ester by the ethyl chloroformate and
ethanol byproduct in the reaction. The derivatives were
analyzed by gas chromatography/mass spectrometry (GC/EI-
MS). The derivatized biosynthetic products had retention times
and fragment ion abundances identical to those of authentic
In a recent study, a tyrosine aminomutase from the monocot
Japanese rice Oryza sativa (OsTAM) was shown to make
exclusively (R)-β-tyrosine.9 No coumarate byproduct or (S)-β-
Received: December 10, 2015
Published: December 28, 2015
© 2015 American Chemical Society
1
DOI: 10.1021/acs.biochem.5b01331
Biochemistry 2016, 55, 1−4

Biochemistry
Rapid Report
standards derivatized equivalently (Figures S1−S5 and details
in the Supporting Information). The kcat of OsTAM was similar
to those of other TAMs (SgTAM, CcTAM, and KedY4) yet
∼10-fold slower than those of all phenylalanine aminomutases
(PAMs) studied so far (Table S1). The KM of OsTAM was
similar in magnitude to those of CcTAM and KedY4 and
approximately 10-fold higher than those of all PAMs and one
TAM (SgTAM). The equilibrium constant (Keq ≈ 1.1 at 29 °C)
at pH 9 is similar to that reported for three other MIO-
aminomutases,3,10 slightly favoring the β-isomer. Here, we
found OsTAM does make p-coumarate, although slowly at
0.0017 s−1; this byproduct interestingly was not observed in a
recent, independent study of OsTAM.9
gauche to the β-acetamido in a staggered anti conformation
when the phenyl ring and carboxy ester groups are farthest
apart. This infers that the δ 2.89 (dd) resonance of the
perprotio (R)-β-tyrosine corresponds to the pro-2R hydrogen
and the one at δ 2.78 corresponds to the pro-2S hydrogen
(Figure 3A).
To assess the stereochemistry of the hydrogen rebound,
[3,3-2H2]-α-tyrosine (5 mM) was incubated with OsTAM (5
mg) in 20 mM phosphate buffer (pH 9.0) at 29 °C for 6 days.
The isotopomers in the mixture were derivatized to their (4′-O-
methyl,N-acetyl) methyl esters, dissolved in CDCl3, and
1
2
analyzed by H NMR. One derivatized biosynthetic H-labeled
β-tyrosine isotopomer had proton resonances at δ 2.89 (d, 2J =
2
1
Stereochemistry of the OsTAM-Catalyzed Reaction.
OsTAM was incubated with (2S)-α-tyrosine to assess the
absolute stereochemistry and ratio of the β-tyrosine enan-
tiomers produced by the biocatalyst. The amino acids in the
reaction mixture were derivatized with chiral auxiliary (2S)-
methylbutyric anhydride and diazomethane. The resulting 4′-
O,3-N-di((2S)-methylbutyric) methyl esters were analyzed by
GC/EI-MS (Figure S7). The biosynthetic β-tyrosines com-
prised a 97:3 mixture of R and S enantiomers; thus, OsTAM,
like other TAMs, makes both enantiomers of β-tyrosine.
Stereochemistry of NH2/H Rebound. We used (2S,3R)-
[3-2H]- and (2S,3S)-[2,3-2H2]-α-tyrosine8 to assess the
enantiospecific removal of the prochiral hydrogens at Cβ
during OsTAM catalysis. OsTAM (0.2 mg) was incubated
separately with each stereospecifically 2H-labeled tyrosine for 3
h. Pyridine and ethyl chloroformate treatment gave the
di(ethoxycarbonyl) 2H-labeled amino acid ethyl esters that
were analyzed by GC/EI-MS. Diagnostic fragment ions (Table
S2) showed that the (3R)-2H of the 2S,3R substrate remained
at the benzylic carbon of the β-product while the (3S)-2H of the
2S,3S substrate was transferred during the OsTAM reaction
(Figure 2). Because the absolute configuration of the
biosynthetic β-tyrosine is 3R, the NH2 group thus adds to Cβ
with ROC.
−16 Hz) and δ 2.78 (d, J = −16 Hz) (Figure 3C). The H
NMR data are consistent with a diprotio-CH2 of a [3-2H]-(R)-
β-tyrosine isotopomer, present at 77% relative to a second
isotopomer, which we identified as (2S,3R)-[2,3-2H2]-β-
tyrosine (23% abundant). The latter had a single resonance
at δ 2.88 (singlet) for the single proton on the methylene
carbon. GC/EI-MS analysis of the derivatized biosynthetic β-
tyrosine confirmed the ratio of mono- and dideuterio
isotopomers (Figure S21). Together, these 1H NMR data
suggest that OsTAM migrates the deuterium from Cβ of
[3,3-2H2]-α-tyrosine to the pro-S position at Cα of β-tyrosine
with 23% deuterium retention and 77% D-to-H exchange
(Figure 3C). This D-to-H exchange was similarly observed for a
TcPAM3 and CcTAM.8 Thus, OsTAM catalyzes its isomer-
ization reaction with ROC at each reactive carbon center
(Figure 1).
Effect of pH on the Stereochemistry of β-Tyrosine
Catalyzed by OsTAM. In previous studies of other tyrosine
aminomutases, SgTAM11 and CcTAM,8,12 the R:S ratio of β-
tyrosine enantiomers biosynthesized by each catalyst changed
as the reactions progressed.5 The earlier study of CcTAM
showed that at pH 7 the amount of (S)-β-tyrosine relative to
the R isomer increased from 10 to 20% as the conversion of α-
to β-tyrosine progressed from 13 to 55%, respectively. The
latter study also showed that as the pH increased from 7 to 9
the (S)-antipode increased from 10 to 17% after 13%
conversion of the α- to β-amino acid and from 20 to 27%
after 55% conversion, respectively.8 These earlier data strongly
suggested that the enantiomeric ratio of the products catalyzed
by TAMs is likely dependent on pH and reaction progress. To
assess the effects of pH on OsTAM, (2S)-α-tyrosine was
incubated with OsTAM for 24 h at pH 7, 8, 9, and 10. OsTAM
interestingly maintained an ∼97:3 R:S ratio as the reaction
reached different maximal conversions (9−52% conversion of
α- to β-tyrosine) for each reaction at pH 7−10 over 24 h
(Figure 4). The dearth of (3S)-antipode (∼3%) catalyzed by
OsTAM might have contributed to it being overlooked in an
earlier study that used HPLC separation and fluorescence
detection of the product.9
Figure 2. (2S,3R)-[3-2H]- and (2S,3S)-[2,3-2H2]-α-tyrosine substrates
and the 2H labeling of biosynthetic products inferred from GC/EI-MS
analysis of N,O-diacyl ethyl ester derivatives. D-to-H exchange shown
as a percentage.
1
Comparing the Active Sites of MIO-Dependent
Enzymes. Similar to those of other MIO-dependent AMs,
the OsTAM active site comprises residues from three
monomeric subunits of the active homotetramer.4,13,14 An
Ala-Ser-Gly triad forms the MIO group; a catalytic Tyr98 is
poised to transfer a Cβ hydrogen of the substrate, and a
conserved Arg344 is positioned to form a carboxylate salt
bridge with the substrate. OsTAM catalyzes its product with
similar R enantioselectivity (∼97%) as TcPAM, catalyzing its
reaction with a 99.9% preference for (R)-β-phenylalanine.15
The H NMR chemical shifts of the prochiral hydrogens of
the derivatized, authentic (R)-β-tyrosine appear as an ABX spin
system at δ 2.89 (dd, 2J = −15 Hz; 3J = 6.0 Hz) and δ 2.78 (dd,
2J = −15 Hz; 3J = 6.0 Hz) (Figure 3A). (2S,3R/2R,3S)-
[2,3-2H2]-β-Tyrosine (81% dideuterio; 88% D at Cα; 92% D at
Cβ) was synthesized by Pd-catalyzed syn addition of D2 gas to
methyl (Z)-N-acetamido-(4-methoxyphenyl)acrylate. The
chemical shift singlet for the lone Cα-H of the derivatized
racemic mixture of [2,3-2H2]-β-tyrosine appeared at δ 2.88
(Figure 3B). The Cα-H of the latter dideuterio racemate is
2
DOI: 10.1021/acs.biochem.5b01331
Biochemistry 2016, 55, 1−4

Biochemistry
Rapid Report
1
Figure 3. Partial H NMR spectra of 4′-O-methyl,N-acetyl methyl ester derivatives of β-tyrosine isotopomers. Chemical shifts of (A) the prochiral
hydrogens of derivatized authentic β-tyrosine, (B) the Cα hydrogen of derivatized authentic (2S,3R/2R,3S)-[2,3-2H2]-β-tyrosine (81% dideuterio),
and (C) the Cα hydrogens of derivatized biosynthetic β-tyrosine catalyzed by OsTAM. (a) (i) OsTAM; (ii) base, acetic anhydride; (iii) HCl; (iv)
CH2N2. Numbers in NMR spectrum are relative peak areas.
Figure 5. Active sites of (A) OsTAM structure model and (B) TcPAM
(PDB entry 3NZ4). Residues common to both enzymes (green), the
catalytic Arg and Tyr residues, and the MIO (purple) are depicted.
Asn446 and Tyr125 (yellow) near the para carbon of the coumarate
intermediate modeled in OsTAM are depicted; equivalently positioned
residues (orange) near the para carbon of cinnamate in complex with
TcPAM are also shown. Heteroatoms: oxygen (red), nitrogen (blue),
and sulfur (yellow).
Figure 4. Maximal conversion of α-tyrosine to the 3R β-isomer
catalyzed by OsTAM over 24 h from pH 7 to 10 (black bars). Percent
for tyrosine is supported by an earlier study of a TcPAM
of biosynthetic (3R)-β-tyrosine made by OsTAM relative to the (3S)-
homologue (TchPAM from Taxus chinensis).16 In the earlier
○
antipode ( ).
study, the further-reaching Cys107His mutation allowed
TcPAM to turn over tyrosine with high enantioselectivity,16
suggesting that His107 formed a selective interaction with the
4′-hydroxyl of the tyrosine substrate.
Comparing a structural model of OsTAM and the TcPAM
structure (PDB entry 3NZ4) shows several shared residues
surrounding their active sites (Figure 5). The active site
residues of these plant-derived aminomutases differ by only two
residues. These striking similarities likely explain their shared
enantioselective bias for the R enantiomer. However, OsTAM
uses a Tyr125/Asn446 pair, while TcPAM uses Cys107/Lys427
partners positioned similarly near the para carbon of the
phenylpropanoid substrates. These functionally distinct resi-
dues likely define their respective substrate selectivities. The
Tyr125(OsTAM) residue likely engages in a H-bonding
interaction with the 4′-hydroxyl of the tyrosine substrate. The
longer reach of the H-bonding Tyr125(OsTAM) residue over
the shorter Cys107(TcPAM) likely allows OsTAM to interact
better with its tyrosine substrate. By contrast, Cys107 of
TcPAM likely H-bonds with its Lys427 partner, helping to
construct a hydrophobic active site pocket to instead bind
phenylalanine. The use of an extended length, polar residue
(such as Tyr125) by OsTAM to confer its substrate selectivity
Substrate Selectivity of OsTAM. All known TAMs use a
3-(4-hydroxyaryl)alanine substrate, while PAMs use phenyl-
alanine derivatives and preclude tyrosine as a substrate. TAMs
typically use polar residues (Ser, His, Glu/Thr, and Tyr) near
the 4′-hydroxyphenyl ring of the tyrosine substrate (Figure 6).
In contrast, PAMs use a mixture of variously hydrophobic and
polar residues (Ala/Cys, Val/Leu, Met/Lys, and Ile/Phe) near
the phenyl ring of the eponymous substrate.
OsTAM uses a combination of polar and hydrophobic
residues (Tyr125, Leu126, Asn446, and Val450) around the
aryl ring of the substrate that resemble those found in PAMs
more than in TAMs. Therefore, (S)-α-phenylalanine was
incubated with OsTAM to test its PAM activity. The
biosynthesized products were N,O-derivatized and analyzed
app
by GC/EI-MS as before. OsTAM slowly converted (k
=
cat
0.00017 s−1) (S)-α- to (3R)-β-phenylalanine with a KaMpp of 9
mM (Figures S32 and S33), thus marking the first instance of a
3
DOI: 10.1021/acs.biochem.5b01331
Biochemistry 2016, 55, 1−4

Biochemistry
Rapid Report
Figure 6. Partial sequence alignment (GeneDoc) of MIO-dependent enzymes. Conserved residues (highlighted in purple) are catalytic Tyr, Arg
(binds the carboxylate of the substrate), and the Ala/Thr-Ser-Gly triad (form the MIO cofactor). Isosterically similar residues in the active site of
MIO catalysts (highlighted in green) and residues near the para carbon of the substrate in TcPAM (highlighted in red) and OsTAM (highlighted in
yellow). Residues proposed as being key for substrate selectivity are bold and boxed. Identical residues shared in 10 of the 13 sequences (highlighted
in black). Residue numbering based on OsTAM. Accession numbers are in the Supporting Information.
Notes
wild-type aminomutase that uses both α-phenylalanine and α-
tyrosine as substrates.
The authors declare no competing financial interest.
This study adds information about the mechanisms of
catalysis of MIO-dependent aminomutases. Here, we find, like
two other TAMs (CcTAM and SgTAM), OsTAM catalyzes a
mixture of (R)- and (S)-β-tyrosine, but with very high
enantioselectivity (∼97%) for the R isomer. By contrast,
three other bacterial TAMs are reported to catalyze exclusively
the (S)-β-amino acid.5,17 The enantiomeric ratio catalyzed by
OsTAM is virtually unaltered by pH and reaction progress,
unlike those of bacterial SgTAM and CcTAM, whose R:S ratios
vary over time.7,8,12 This investigation provides a basis to begin
inquiry into how the rice OsTAM restricts its enantioselectivity.
Further, after defining the absolute stereochemistry and using
2H-labeled tyrosines, we showed that the OsTAM mechanism
principally retains the configuration at Cα and Cβ after the
isomerization. This reaction pathway is similar to the
mechanism followed by TcPAM, also from a plant. The unique
active site residues of OsTAM are more permissive than those
in TcPAM and allow OsTAM to turn over two aromatic amino
acids, although with a substantial preference for tyrosine. As
more MIO aminomutases are discovered and their cryptic
stereochemistries examined, a new subdivision of MIO-
dependent enzymes that distinguishes plant aminomutases
from those in bacteria might emerge.
ACKNOWLEDGMENTS
We thank undergraduate Devinda Wijewardena for his
technical assistance.
■
REFERENCES
■
(1) Ratnayake, N. D., Wanninayake, U., Geiger, J. H., and Walker, K.
D. (2011) J. Am. Chem. Soc. 133, 8531−8533.
(2) Magarvey, N. A., Fortin, P. D., Thomas, P. M., Kelleher, N. L.,
and Walsh, C. T. (2008) ACS Chem. Biol. 3, 542−554.
(3) Mutatu, W., Klettke, K. L., Foster, C., and Walker, K. D. (2007)
Biochemistry 46, 9785−9794.
(4) Christianson, C. V., Montavon, T. J., Van Lanen, S. G., Shen, B.,
and Bruner, S. D. (2007) Biochemistry 46, 7205−7214.
(5) Krug, D., and Muller, R. (2009) ChemBioChem 10, 741−750.
̈
(6) Wanninayake, U., and Walker, K. D. (2012) Biochemistry 51,
5226−5228.
(7) Christianson, C. V., Montavon, T. J., Festin, G. M., Cooke, H. A.,
Shen, B., and Bruner, S. D. (2007) J. Am. Chem. Soc. 129, 15744−
15745.
(8) Wanninayake, U., and Walker, K. D. (2013) J. Am. Chem. Soc.
135, 11193−11204.
(9) 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., and Jander, G. (2015) Plant Cell 27, 1265−
1278.
ASSOCIATED CONTENT
(10) Chesters, C., Wilding, M., Goodall, M., and Micklefield, J.
■
(2012) Angew. Chem., Int. Ed. 51, 4344−4348.
S
* Supporting Information
(11) Christenson, S. D., Wu, W., Spies, M. A., Shen, B., and Toney,
M. D. (2003) Biochemistry 42, 12708−12718.
The Supporting Information is available free of charge on the
ACS Publications website at DOI: 10.1021/acs.bio-
chem.5b01331.
(12) Rachid, S., Krug, D., Weissman, K. J., and Muller, R. (2007) J.
Biol. Chem. 282, 21810−21817.
(13) Feng, L., Wanninayake, U., Strom, S., Geiger, J., and Walker, K.
D. (2011) Biochemistry 50, 2919−2930.
Experimental methods, GC/MS and NMR data, and
gene accession numbers (PDF)
(14) Strom, S., Wanninayake, U., Ratnayake, N. D., Walker, K. D.,
and Geiger, J. H. (2012) Angew. Chem., Int. Ed. 51, 2898−2902.
(15) Klettke, K. L., Sanyal, S., Mutatu, W., and Walker, K. D. (2007)
J. Am. Chem. Soc. 129, 6988−6989.
AUTHOR INFORMATION
■
Corresponding Author
*E-mail: walker@chemistry.msu.edu.
(16) Wu, B., Szymanski, W., Wijma, H. J., Crismaru, C. G., de
Wildeman, S., Poelarends, G. J., Feringa, B. L., and Janssen, D. B.
(2009) Chem. Commun. 46, 8157−8159.
(17) Van Lanen, S. G., Oh, T.-j., Liu, W., Wendt-Pienkowski, E., and
Shen, B. (2007) J. Am. Chem. Soc. 129, 13082−13094.
Funding
The authors are grateful to the MSU Office for Inclusion and
Intercultural Initiatives for supporting the MSU Chemistry 4-
Plus Bridge to the Doctorate Program (Z.K., GA017081-
CIEG).
4
DOI: 10.1021/acs.biochem.5b01331
Biochemistry 2016, 55, 1−4