Intermediate partitioning kinetic isotope effects for the NIH shift of 4-hydroxyphenylpyruvate dioxygenase and the hydroxylation reaction of hydroxymandelate synthase reveal mechanistic complexity

Article abstract of DOI:10.1021/bi400534q

4-Hydroxyphenylpyruvate dioxygenase (HPPD) and hydroxymandelate synthase (HMS) are similar enzymes that catalyze complex dioxygenation reactions using the substrates 4-hydroxyphenylpyruvate (HPP) and dioxygen. Both enzymes decarboxylate HPP and then hydroxylate the resulting hydroxyphenylacetate (HPA). The hydroxylation reaction catalyzed by HPPD displaces the aceto substituent of HPA in a 1,2-shift to form 2,5-dihydroxyphenylacetate (homogentisate, HG), whereas the hydroxylation reaction of HMS places a hydroxyl on the benzylic carbon forming 3′-hydroxyphenylacetate (S-hydroxymandelate, HMA) without ensuing chemistry. The wild-type form of HPPD and variants of both enzymes uncouple to form both native and non-native products. We have used intermediate partitioning to probe bifurcating steps that form these products by substituting deuteriums for protiums at the benzylic position of the HPP substrate. These substitutions result in altered ratios of products that can be used to calculate kinetic isotope effects (KIE) for the formation of a specific product. For HPPD, secondary normal KIEs indicate that cleavage of the bond in the displacement reaction prior to the shift occurs by a homolytic mechanism. NMR analysis of HG derived from HPPD reacting with enantiomerically pure R-3′-deutero-HPP indicates that no rotation about the bond to the radical occurs, suggesting that collapse of the biradical intermediate is rapid. The production of HMA was observed in HMS and HPPD variant reactions. HMS hydroxylates to form exclusively S-hydroxymandelate. When HMS is reacted with R-3′-deutero-HPP, the observed kinetic isotope effect represents geometry changes in the initial transition state for the nonabstracted proton. These data show evidence of sp³ hybridization in a HPPD variant and sp ² hybridization in HMS variants, suggesting that HMS stabilizes a more advanced transition state in order to catalyze H-atom abstraction.

Full text of DOI:10.1021/bi400534q

Article
pubs.acs.org/biochemistry
Intermediate Partitioning Kinetic Isotope Effects for the NIH Shift of
4‑Hydroxyphenylpyruvate Dioxygenase and the Hydroxylation
Reaction of Hydroxymandelate Synthase Reveal Mechanistic
Complexity
Dhara D. Shah,† John A. Conrad,‡ and Graham R. Moran*,†
†Department of Chemistry and Biochemistry, University of WisconsinMilwaukee, 3210 North Cramer Street, Milwaukee,
Wisconsin 53211-3209, United States
‡Department of Chemistry, University of NebraskaOmaha, 6001 Dodge Street, Omaha, Nebraska 68182-0109, United States
S
* Supporting Information
ABSTRACT: 4-Hydroxyphenylpyruvate dioxygenase (HPPD) and hydroxymandelate synthase (HMS) are similar enzymes that
catalyze complex dioxygenation reactions using the substrates 4-hydroxyphenylpyruvate (HPP) and dioxygen. Both enzymes
decarboxylate HPP and then hydroxylate the resulting hydroxyphenylacetate (HPA). The hydroxylation reaction catalyzed by
HPPD displaces the aceto substituent of HPA in a 1,2-shift to form 2,5-dihydroxyphenylacetate (homogentisate, HG), whereas
the hydroxylation reaction of HMS places a hydroxyl on the benzylic carbon forming 3′-hydroxyphenylacetate (S-
hydroxymandelate, HMA) without ensuing chemistry. The wild-type form of HPPD and variants of both enzymes uncouple to
form both native and non-native products. We have used intermediate partitioning to probe bifurcating steps that form these
products by substituting deuteriums for protiums at the benzylic position of the HPP substrate. These substitutions result in
altered ratios of products that can be used to calculate kinetic isotope effects (KIE) for the formation of a specific product. For
HPPD, secondary normal KIEs indicate that cleavage of the bond in the displacement reaction prior to the shift occurs by a
homolytic mechanism. NMR analysis of HG derived from HPPD reacting with enantiomerically pure R-3′-deutero-HPP
indicates that no rotation about the bond to the radical occurs, suggesting that collapse of the biradical intermediate is rapid. The
production of HMA was observed in HMS and HPPD variant reactions. HMS hydroxylates to form exclusively S-
hydroxymandelate. When HMS is reacted with R-3′-deutero-HPP, the observed kinetic isotope effect represents geometry
changes in the initial transition state for the nonabstracted proton. These data show evidence of sp3 hybridization in a HPPD
variant and sp2 hybridization in HMS variants, suggesting that HMS stabilizes a more advanced transition state in order to
catalyze H-atom abstraction.
4-Hydroxyphenylpyruvate dioxygenase (HPPD) and hydrox-
ymandelate synthase (HMS) catalyze similar reactions using
the same substrates, 4-hydroxyphenylpyruvate (HPP) and
dioxygen. Initially, both enzymes reduce and activate dioxygen
in order to decarboxylate HPP, yielding 4-hydroxyphenylace-
tate (HPA), CO2, and an activated oxo intermediate.1 Both
enzymes then hydroxylate HPA but do so in different positions.
HPPD catalyzes aromatic hydroxylation at the ring C1,
resulting in a migration of the aceto group of HPA to yield
2,5-dihydroxyphenylacetate (homogentisate, HG), whereas
HMS catalyzes hydroxylation of the HPA benzylic methylene
to form S-hydroxymandelate (HMA) (Scheme 1). Despite
modest sequence identity and similarity, the two enzymes
clearly have a common ancestry in that they share numerous
active site residues and adopt a highly similar fold.2 HPPD is in
all likelihood the progenitor because it is ostensibly ubiquitous
in aerobes, where it catalyzes the second step of tyrosine
Received: April 29, 2013
Revised: July 18, 2013
Published: August 13, 2013
© 2013 American Chemical Society
6097
dx.doi.org/10.1021/bi400534q | Biochemistry 2013, 52, 6097−6107

Biochemistry
Article
α-KAOs, so this classification is based solely on their chemical
reactions.7−9 α-KAOs catalyze an extraordinary range of
chemistries; however, each member begins catalysis with the
cleavage of dioxygen and the oxidative decarboxylation of an α-
keto acid (almost universally α-ketoglutarate). These steps are
thought to result in the formation of a common reactive
intermediate (Fe(IV)O) that is then directed to unique
purposes by individual enzymes.10−12 Although these purposes
include an impressive array of oxidation reactions, halogenation,
and even epimerization, this reactive oxo species is most often
used to hydroxylate.6 HPPD and HMS offer respective
examples of aromatic and aliphatic hydroxylation chemistry in
which the requisite α-keto acid is supplied by the pyruvate
substituent of HPP rather than from α-ketoglutarate (Figure 1).
Aromatic hydroxylation by HPPD induces substituent
migration. In the 1960s, Guroff et al. discovered that aromatic
Scheme 1
catabolism, whereas HMS catalyzes the first step in the
biosynthesis of hydroxyphenylglycine, an uncommon tailoring
reaction in a relatively small number of bacteria.3,4
Both HPPD and HMS are Fe(II)-dependent dioxygenases
and, on the basis of the reactions they catalyze, can be grouped
with the α-keto acid dependent oxygenases (α-KAOs).5,6
Structurally, HPPD and HMS have no resemblance to other
Figure 1.
6098
dx.doi.org/10.1021/bi400534q | Biochemistry 2013, 52, 6097−6107

Biochemistry
Article
Scheme 2
hydroxylation by pterin-dependent hydroxylases can induce an
intramolecular migration or shift of the substituent from the
site of hydroxylation.13 Coined the NIH shift, this phenomenon
has since been documented for a variety of metal-dependent
oxygenases.14−17 For essentially all, the shift is observed with
compact substituents such as halides, deuteriums and tritiums
(and presumably protiums), and methyl groups. It has also
been observed that, for substituents larger than a proton, the
shift occurs in only a fraction of total catalysis.13,15,18−24 The
aromatic hydroxylation reaction catalyzed by HPPD induces
the migration of the comparatively large aceto substituent with
high efficiency 25 and so, from an experimental standpoint, is an
important example of this chemistry.
benzylic carbon and subsequent configurational analysis of the
product. It was contended that if a two-electron or carbanion
mechanism was at work, sp3 geometry at the benzylic carbon
would predominate in the cleaved intermediate; therefore, a
chiral substrate would retain the configuration in the product.
Conversely, if the one-electron mechanism were operative, a
symmetrical sp2 in-flight radical could rotate before recombin-
ing with the ring, resulting in racemization at a chiral benzylic
carbon. From these experiments, Leinberger et al. concluded
that the NIH shift of HPPD occurred via a two-electron
mechanism with retention of configuration at the benzylic
methylene C atom.22 More recently, using density functional
approaches to compute the more probable of a comprehensive
set of reaction coordinate possibilities, Borowski et al.
concluded that the NIH shift of HPPD was likely to occur
Initial attempts to define the mechanism of the HPPD shift
used single enantiomer chiral monodeutero-HPP labeled at the
6099
dx.doi.org/10.1021/bi400534q | Biochemistry 2013, 52, 6097−6107

Biochemistry
Article
via a homolytic mechanism and that retention of configuration
at the benzylic position was simply a consequence of a low
barrier to the transition state (TS) for the displaced group to
recombine with the ring relative to the barrier for bond
rotation.26 As such, there is some incongruity in the literature
that is open to be settled by experimental observation that
reports discriminating evidence for the mechanism of the NIH
shift in HPPD.
The hydroxylation reaction catalyzed by HMS is less complex
than that of HPPD and is thought to involve only an initial
hydrogen abstraction by the Fe(IV)O species from the
benzylic position to form a benzylic radical followed by
hydroxyl addition from the resulting Fe(III)OH species.25
Computational methods have provided a comprehensive
description of the basis for hydroxylation regiospecificity in
HMS and have suggested that H-atom abstraction forming a
benzylic radical followed by oxygen rebound is the dominant
mechanism.27,28 Experimental evidence for the mechanism has
come from kinetic isotope effect (KIE) values for the
hydroxylation reaction from intermediate partitioning (vide
infra) and were consistent with H-atom abstraction as the first
step.12,25,29
The fundamental objectives of this study were to obtain
experimental evidence that discriminates between the hetero-
lytic and homolytic NIH-shift mechanisms with HPPD and to
define the geometry of the benzylic radical that forms with H-
atom abstraction in HMS (highlighted steps in Figure 1). The
measurement of KIEs for the relevant catalytic steps would
provide evidence for the mechanism(s), but for both HPPD
and HMS, this observation cannot be made directly because the
hydroxylation and NIH shift steps are kinetically shrouded in
single turnover reactions.30−32 We have recently shown that
HPPD wild-type and also HPPD and HMS variants exhibit
bifurcations that produce multiple products.25 For HPPD,
bifurcation at the NIH-shift step yields either quinolacetic acid
(QAA), a molecule that is hydroxylated but has not undergone
the shift, or the native shift product, HG,25,33 whereas
bifurcation for the H-atom abstraction step that forms HMA
Master Mix and Competent BL21(DE3) Escherichia coli cells
were obtained from New England Biolabs. IPTG was from
Gold Biochemicals. D2O and CDCl3 were obtained from
Cambridge Isotope Laboratories. DMSO-d6 was from Acros
Chemicals. The gene for mouse phenylpyruvate tautomerase
(PPT) optimized for expression in Escherichia coli and including
codons for a His-Tag was purchased from EnzymaX and
subcloned into the Nde I and Xho I restriction sites of pET17b
for expression. Q-Sepharose was from Bio-Rad. Sephacryl S-200
was obtained from Amersham. Quinolacetate (QAA) was made
using the HPPD P243T variant and purified as previously
described.33 3′,3′-Diduetero-HPP was prepared according to
our previously published method.32 Solutions of the enol form
of HPP were prepared by dissolving crystalline enol-HPP (as
purchased) in the aprotic water-miscible solvent methanol to
prevent ketonization. The chiral shift reagent, rhombamine
macrocycle, was a kind gift from Dr. Koichi Tanaka.38 All other
chemicals, buffers, and biological media were obtained from
Fisher Scientific or Sigma-Aldrich Chemicals and were of high
purity.
Mutagenesis. pET17b-derived (Novagen) plasmids carry-
ing the Streptomyces avermitilis HPPD gene (pSAHPPD39) and
the Amycolatopsis orientalis HMS gene (pAOHMS2) were
mutated as previously described to obtain the N245Q, P243T,
S230A, and N245S variants of HPPD and the S201A variant of
HMS.25,33 An additional HPPD variant for this study, F364I,
was obtained using the Stratagene QuickChange protocol using
Phusion High-Fidelity PCR Master Mix and the following
oligonucleotides: TCGGCAAGGGCAACATCAAGGCCCT-
GTTC paired with an oligonucleotide that was its reverse
complement. All variants were screened for spurious mutations
by sequencing the entire dioxygenase genes (Sequetech,
Mountain View, CA).
Expression and Purification. The apo-forms of wild-type
and variants of HPPD and HMS were expressed and purified
using ammonium sulfate fractionation, anion exchange
chromatography, and size exclusion chromatography according
to published methods.2,39
yields HPA, an intermediate that has undergone decarbox-
His-tagged PPT was purified using the following protocol.
Aliquots from frozen cell stocks were plated (200 μL/L of
culture) on LB agar with 100 μg/mL ampicillin. After
incubation for 9−12 h at 37 °C, the cells from two plates
were suspended in ∼20 mL of LB broth and used to inoculate 1
L of LB broth with 100 μg/mL ampicillin. The culture was
grown with vigorous shaking (225 rpm) at 37 °C until the cell
density had reached an OD600 of 1.0. At this time, IPTG was
added to a final concentration of 0.1 mM and incubation was
continued for 2.5 h. The cells were harvested by centrifugation
at 4665g for 30 min and used immediately for protein
purification. Unless otherwise stated, all subsequent purification
procedures were undertaken at 4 °C. Cells were resuspended
using 20 mL of 20 mM HEPES, pH 7.0, per liter of culture and
lysed with sonication (2 × 6 min at 45 W) using a Branson
sonicator fitted with a blunt tungsten tip. The temperature of
the solution was monitored to ensure that it did not exceed 10
°C. The lysed cells were then centrifuged at 15600g for 30 min,
and the pellet was discarded. Streptomycin sulfate in water was
then added to the supernatant to a final concentration of 1.5%
w/v. The mixture was stirred for 10 min and centrifuged at
12900g for 15 min. The supernatant was then loaded directly
onto a Talon metal affinity column pre-equilibrated in 20 mM
HEPES, pH 7.0, at a flow rate of 1 mL/min. The column was
washed with 20 mM HEPES and 10 mM imidazole, pH 7.0, to
25
ylation but is not hydroxylated
(Scheme 2). Thus, these
systems are amenable to the measurement of intermediate-
partitioning-based KIE values. The basis for this method is that
the ratio of products is a direct result of a probability-based
selection that is defined by the decay rate constant for
individual paths bifurcating from some reactive state of the
enzyme.34 As such, substitutions of heavy isotopes for atoms
that are involved in the chemistry of these diverging steps can
alter the balance of the rate constants and thus the product
ratio. Thus, an intrinsic kinetic isotope effect for a specific
reaction step can be calculated simply. This accurate and highly
robust method was originally applied to the P450 hydroxylases
and more recently to mono- and dinuclear Fe(II) dependent
oxygenases.16,35−37 The data we have obtained via these
methods give KIE values that are consistent with the
mechanisms proposed for each of the catalytic steps, which at
least narrows the mechanistic possibilities and suggests that
alternate pathways can be induced that yield the same product
in variant forms of the enzymes.
MATERIALS AND METHODS
■
Materials. HPP and HG were purchased from Acros
Chemicals. HMA and 4-hydroxyphenylacetate (HPA) were
obtained from Sigma-Aldrich. Phusion High-Fiedility PCR
6100
dx.doi.org/10.1021/bi400534q | Biochemistry 2013, 52, 6097−6107

Biochemistry
Article
D
elute nonspecifically bound proteins. The PPT was eluted with
a linear 600 mL gradient from 10 to 250 mM imidazole in 20
mM HEPES, pH 7.0, at a flow rate of 1 mL/min. The fractions
containing PPT were pooled and stored at −80 °C.
In this equation, kobs is determined by measuring P1, the
concentration of product for which the isotope effect is being
determined (HG or HMA for HPPD, and HMA for HMS), and
P2, the concentration of the product that results from
bifurcation in a step that is insensitive to the isotopic insertion
(QAA or HPA for HPPD, and HPA for HMS). In each
measurement, the total consumption of molecular oxygen was
compared to the total concentration of products formed to
ensure that all catalysis was considered.
KIE Values from HPLC Product Analysis. The ratio of
products for wild-type and each variant enzyme were
determined using high-pressure liquid chromatography
(HPLC). Enzyme turnover reactions were undertaken at 25
°C and were composed of 25−30 μM ferrous sulfate, 10−20
μM enzyme (wild-type or variant HPPD or HMS), 1 mM
reductant (either βME or ascorbic acid), 250 μM O2, and 100−
250 μM HPP (either 3′,3′-dideutero or diprotio) in 10 mM
potassium phosphate, pH 7.0. Isotope effects were based on
peak areas for products measured by HPLC as previously
published.25
Preparation of Monodeuterated HG from Wild-Type
HPPD. For the production of HG in sufficient quantity for
NMR analysis, a large-scale turnover reaction (40 mL) was
undertaken in D2O with wild-type HPPD. The reaction mixture
was composed of 420 U of wild-type HPPD, 112 U of PPT, 60
μM ferrous sulfate, and 1 mM ascorbate in 5 mM sodium
phosphate buffer, pD 7.0, and was initiated with the addition of
1.5 mM enol HPP in methanol. The reaction was run at 25 °C
and was equilibrated with 100% O2 by stirring in a vessel with
two small ports so that 100% dioxygen gas could continuously
flush the headspace of the container. Complete conversion of
the available HPP to HG was achieved within ∼40 min, as
determined by analytical reversed-phase HPLC using published
methods.25 The reaction was then brought down to pD 2.0 by
the addition of H2SO4 to stabilize the HG hydroquinone, and
the precipitated protein was removed by centrifugation at
12900g for 20 min. The supernatant was then passed through
10 kDa nominal molecular weight limit centrifugal filter
(Vivaspin15) by centrifuging at 3000g to remove residual
soluble peptide. The filtrate was then lyophilized overnight to a
volume of ∼5 mL, and the sample was loaded on to Sep-Pak
Vac 35 cc C18 column pre-equilibrated with 40 mM sodium
phosphate buffer, pH 7.0. The column was washed with 60 mL
of H2O, and the HG was then eluted with 10% acetonitrile in
water. Each eluted fraction was examined spectrophotometri-
cally for the characteristic λmax absorbance of HG at 290 nm.
Fractions showing significant absorbance at 290 nm were
combined together and lyophilized to obtain dry (powdered)
HG.
KIE values measured with R-3′-monodeutero-HPP were
obtained from coupled turnover reactions in which the
stereospecifically labeled monodeutero substrate was formed
and consumed in situ. R-3′-monodeutero-HPP was prepared
using phenylpyruvate tautomerase (PPT) in deuterium oxide
solvent. PPT catalyzes a stereospecific reaction in which it
interconverts the enol and keto forms of HPP by incorporating
or abstracting a proton at the pro-R position.40 The ratio of
products observed for HPPD or HMS when R-3′-mono-
deutero-HPP was used as a substrate was determined by adding
the enol form of HPP to a reaction that had both PPT and
HMS or HPPD activities. The reaction was designed such that
the combined PPT and nonenzymatic conversion of enol-HPP
to R-3′-monodeutero-keto-HPP was 7-fold faster than the
nonenzymatic conversion of the enol-HPP to the S-
3′monodeutero-keto-HPP, and the HMS or HPPD activity
was 4-fold greater than that of PPT, accounting for both the
solvent kinetic isotope effect (SKIE) on the product release
step of HPPD and HMS and the SKIE for PPT.30,32,40,41
Enzyme turnover reactions were undertaken at 25 °C and were
composed of ferrous sulfate equivalent to the concentration of
HMS or HPPD that gave 1.06 U activity (12.5 μM S201A
HMS or 106 μM F364I HPPD), 0.28 U of PPT, and 1 mM
ascorbate in 5 mM sodium phosphate buffer prepared in
deuterium oxide, pD 7.0 (U is a unit of enzyme activity and is
equal to 1 μmol of substrate consumed or product evolved per
min). The low concentration of buffer used in these reactions
suppresses the observed rate constant for the nonenzymatic
conversion of the enolic HPP to ketoic HPP. The reaction was
initiated by the addition of 50 μM enol HPP in 100% methanol.
After completion of the reaction, all other steps required to
determine the product concentrations were as previously
published.25
Chiral Shift Analysis of Monodeuterated HG. Lyophi-
lized HG was dissolved in CDCl3 containing 10% DMSO-d6,
from which 5 mg of HG was taken for NMR analysis. As a
standard, a comparable amount of unlabeled HG solution in
CDCl3 containing 10% DMSO-d6 was also prepared.
Monodeuterated HG and unlabeled HG solutions were titrated
with 0 to 0.05 equiv rhombamine macrocycle (chiral shift
reagent) by small additions.38 1H NMR spectra were recorded
on a 500 MHz Bruker NMR instrument.
RESULTS
Calculation of KIEs. Kinetic isotope effects were calculated
using eq 1, which computes the factor of perturbation of the
product ratio in the presence of the deuterated substrate for the
case of a bifurcating mechanism and incorporates a correction
■
Product Analysis and Characterization of Variant
Enzymes. All the HPPD and HMS variants except F364I
HPPD and N245D HPPD have been characterized in prior
work.25 The product ratios for wild-type, N245Q, N245S,
P243T, and S230A vary slightly from those previously
published. However, the KIE values obtained by partitioning
methods are calculated from a ratio of ratios and therefore are
largely immune to such variation (eq 1). F364I HPPD and
N245D HPPD could be expressed and purified using the same
methods as wild-type HPPD.39 Interestingly, F364I HPPD
produced four products (Table 2), of which the most
interesting was HMA, the product of the HMS reaction. This
was consistent with the work of O’Hare et al. and Gunsior et al.,
who also observed HMA production with this variant.42,43
D
accounting for the fractional extent by which kobs is reduced
from the intrinsic value (Dk2) by the bifurcation.34
Dk2 + (k2/k3)
1 + (k2/k3)
(P/(P + P ))H
Dkobs
=
=
1
1
2
(P/(P + P ))D
(1)
1
1
2
6101
dx.doi.org/10.1021/bi400534q | Biochemistry 2013, 52, 6097−6107

Biochemistry
Article
a
Table 1. NIH Shift KIEs from Wild-Type HPPD and Variants
variant
wild-type
N245Q
52
N245D
52
N245S
3.4
P243T
8.4
S230A
HG (%)
QAA (%)
HPA (%)
Dkobs HG
replicates
90
1
7
3
21.2
90
45
36
57
9
45
26.8
6.6
46.5
0.990 0.001
2
0.99 0.03
2
1.03 0.01
2
1.39 0.02
2
1.08 0.05
2
1.08 0.01
2
a
N.B. all product percentages are those obtained with the protio substrates. Product percentages for protio and deuteron substrates are available in
the Supporting Information.
Although rationalizing the function of individual residues is
secondary to the purpose of this study, it is interesting to note
that the position occupied by F364 in the structure of HPPD is
filled by I335 in the structure of HMS.44 As such, this mutation
yields one of the only single positional-switching variants for
these two enzymes that has resulted in the alternate activity to
any significant extent. N245D was a largely competent HPPD,
successfully completing the NIH shift to form HG in 50% of
total turnovers with the remaining fractions of turnover yielding
QAA (20%) and HPA (30%).
KIEs for the NIH Shift from Intermediate Partitioning
in Wild-Type and Variant HPPDs. Despite that HPPD
catalysis represents one of the more compelling examples of
shift chemistry, the NIH shift step neither contributes to the
limiting rate of turnover nor is it resolved kinetically in single
turnover reactions, so it is difficult to study by conventional
methods.30,31 Intermediate partitioning is one of the few
experimental methods capable of providing direct evidence for
the chemistry occurring in this step. Moreover, the precision
obtained with this method is sufficient to discern between
mechanisms based on KIEs of small magnitude, as are often
obtained for secondary deuterium isotope effects arising from
subtle hybridization changes at specific carbon atoms. Table 1
shows the utility of the product partitioning in differentiating
between possible pathways for the NIH shift. For HPPD, QAA
acts as an internal reference product for the shift step, having
undergone hydroxylation but not the shift; thus, its formation
rate constant would not be sensitive to deuterons placed at the
benzylic carbon (C3′). To form HG, however, the enzyme
must transiently sever the bond to this carbon, so the rate
constant for the formation of HG will reflect an extent of
influence of benzylic deuterons (Scheme 2). Wild-type HPPD
and variant N245Q show a KIE of 1 for the formation of the
native product HG, which indicates that no significant
geometric change at C3′ occurs in the formation of the
(first) TS for the shift, or more precisely that the sp3 geometry
predominates. The interpretation of this number does not
converge on a single mechanism. A trigonal C3′ in the shift TS
is consistent with an early TS that has retained much of the
initial hybridization of the preshift benzylic methylene. It does
not, however, indicate explicitly that the mechanism is hetero-
or homolytic, as the TS could readily progress to either an in-
flight radical or carbanion intermediate.
In contrast, N245S HPPD showed a distinctly normal KIE of
1.39. This indicates that the (net) rate constant leading the
formation of HG for this variant is slowed by deuterons in the
benzylic position and that a transient geometry change at C3′
must accompany the formation of the shift TS. An sp3 to sp2
hybridization change resulting from homolytic severing of the
C1−C3′ bond would account for the observed effect with this
variant, suggesting a later TS in which the unrestrained in-flight
radical character (sp2) is developed in the TS.
Taken together, and assuming that the reaction traverses the
same barrier for both wild-type and variants, these data support
homolytic cleavage of the bond to the methylene attending the
NIH shift. This conclusion is consistent with what has been
proposed by Borowski et al. who, from calculation, also
proposed a highly unstable biradical intermediate resulting from
the severed bond that collapses to form the new bond to the
ring.26 All other HPPD variants tested in this manner, N245D,
6102
dx.doi.org/10.1021/bi400534q | Biochemistry 2013, 52, 6097−6107

Biochemistry
Article
a
Table 2. Delineation of KIE for the Formation of HMA To Identify Hybridization Geometry of the Radical
variant
F364I HPPD
S201A HMS
HG (%)
47
0
QAA (%)
HPA (%)
HMA (%)
total (%)
19
0
15
3
20
97
100
100
Dkobs, 3′,3′-dideutero-HPP
replicates
3.52 0.1
2.020 0.004
2
3
Dkobs R-3′-monodeutero-HPP
replicates
1.92 0.16
2
1.06 0.04
3
a
N.B. all product percentages are those obtained with the protio substrates. Product percentages for protio and deutero substrates are available in the
Supporting Information.
P243T, and S230A, show small secondary normal KIEs (1.03−
1.08) for the formation of HG. Within error, these values
suggest shift TS geometry similar to the wild-type (C3′-sp3).
NMR Chiral Shift Analysis of 3′-Monodeutero-HG. To
further add to our evidence for the nature of the NIH shift
chemistry, we examined the native product, HG, for evidence of
deuterium scrambling when derived from stereospecifically
labeled HPP (R-3′-monodeutero). This experiment was
undertaken simply to reaffirm the observations of Leinberger
et al., who conceded that derivatization of the product may
have influenced the degree of scrambling observed.22 The
hypothesis for this experiment is that heterolytic cleavage of the
C1−C3′ bond would disfavor racemization of the label at C3′
due to the polar trigonal shape of the in-flight carbanion,
whereas the inherently planar and symmetrical sp2 radical in
this position could permit scrambling. Figure 2 shows NMR
spectra that depict a titration of HPPD-derived monodeu-
terated HG with rhombamine macrocycle chiral shift reagent.38
In this figure, inset spectra in blue are for unlabeled HG. The
lower of these standard spectra is without shift reagent and
shows a single shift for the equivalent benzylic protons. The
upper standard spectrum shows splitting of these two benzylic
protons in an AB spin system with the addition of the shift
reagent. The spectra in black include the titration of the HPPD-
derived HG with the shift reagent. In the absence of the shift
reagent, two HG populations are observed: One resonates at
3.34 ppm and corresponds to the benzylic protons of unlabeled
HG. The second resonance at 3.32 ppm is for R- or S- and/or
RS-3′-monodeutero-HG. The greatest extent of splitting for
unlabeled HG was observed for 0.035 equiv of the shift reagent;
however, HPPD derived monodeutero-HG does not show
evidence of splitting indicating a single enantiomer of
monodeuterated HG. It can now be stated that this lack of
label scrambling is indeed characteristic to the reaction.
However, as noted by Borowski et al., this evidence by itself
is not discriminating for the chemistry of the shift. Such a result
would be predicted for the heterolytic pathway but is also
equally consistent with a homolytic pathway in which the
collapse of a biradical intermediate occurs orders of magnitude
more rapidly than the time taken for a C2′−C3′ bond
rotation.26
Deconstruction of Net KIE Values for the Production
of HMA Using Stereospecifically Deuterated HPP. In
prior work, we have measured the primary deuterium KIE on
6103
dx.doi.org/10.1021/bi400534q | Biochemistry 2013, 52, 6097−6107

Biochemistry
Article
alternate products in some fraction of the total turnover.25,33
Although it may be tempting to rationalize the meaning of the
ratios of products observed with regard to the function of
specific residues, the ratios are dependent on the prevailing
conditions of the reaction (data not shown) and therefore offer
limited opportunities for meaningful conclusions. However,
multiple products indicate that bifurcations (and trifurcations)
are occurring at identifiable stages of catalysis. Specific heavy
isotope substitutions therefore have the potential to alter the
rate constant associated with one arm of the bifurcation; as
such, the KIE for that step can be calculated if the other arm of
the branch is unaffected by the substitution (eq 1, Scheme 2).
Although the data obtained by such methods can be
exceedingly precise, they are also easily overinterpreted. The
inherent kinetic complexity of a multiproduct system dictates
that rationalization of results will likely offer explanations that
are narrow relative to all possible mechanistic and kinetic
possibilities. Herein, we have attempted to explain our data
primarily on the basis of geometric changes in the transition
state of the step affected by the isotope substitution. This
approach assumes that the individual steps are largely
irreversible or that their commitment forward is such that
any reversibility can be ignored. This is broadly reasonable
because oxygenases catalyze highly exothermic reactions for
which most steps will be irreversible or committed forward.
Reversibility is only relevant for those steps that define the ratio
of products of interest (i.e., for the initial steps that lead from
the branch point to the products that define the KIE value).
Reversibility in either of the steps leading away from bifurcating
intermediate will increase the proportion of the other product
independent of the TS geometry. Thus, it should be noted that
the assumptions upon which conclusions are based in this study
cannot, in the absence of additional data, be validated, so what
follows is a confined set of conclusions that attempt to form a
reasoned explanation for the observations but cannot account
exhaustively for the nuanced nature of the interpretation of
individual KIE values for complex systems.
Figure 2. Evidence for a single enantiomer of homogentisate derived
from R-3′-deutero-4-hydroxyphenylpyruvate. HG (5 mg) derived from
HPPD reacting with R-3′-monodeutero-HPP was dissolved in CDCl3
and 10% DMSO-d6. A comparable amount of unlabeled HG solution
in CDCl3 containing 10% DMSO-d6 was also prepared. Monodeu-
terated HG and unlabeled HG solutions were titrated with 0 to 0.05
equiv rhombamine macrocycle (chiral shift reagent) by small
additions. H NMR spectra were recorded on a 500 MHz Bruker
NMR instrument. Blue spectra show the spectra of 3′-diprotic HG in
the presence and absence of the chiral shift reagent.
1
the initial hydrogen abstraction step of the hydroxylation
reaction for HMS using 3′,3′-dideutero-HPP.25 Although the
value measured was modest (2.2−2.6 depending on the enzyme
variant), it includes a multiplicative component from the
geometry at the benzylic carbon for the deuteron not in-flight
(Table 2). Two variants were selected to measure KIEs for the
production of HMA with R-3′-monodeutero-HPP. The first
was S201A HMS and the second was F364I HPPD (that makes
20% HMA). Table 2 includes the percentage of each product
produced by these variants and KIEs measured for the
formation of HMA with 3′,3′-dideutero-HPP as well as 3′-R-
monodeuterated HPP. With 3′,3′-dideutero-HPP as a substrate,
S201A HMS and F364I HPPD show combined primary/
secondary KIEs of 2.02 and 3.52, respectively. For KIEs
measured with 3′-R-monodeuterated HPP as a substrate,
S201A HMS showed a value of 1.06 0.04 for the production
of HMA, indicating the benzylic carbon retains predominant
sp3 character in the TS for the initial hydrogen abstraction.
Conversely for F364I HPPD, the secondary KIE measured with
3′-R-monodeuterated HPP for the formation of HMA was 1.92
0.16, indicating that a geometry change accompanies benzylic
hydrogen abstraction for this HPPD variant. This would be
consistent with a change from sp3 to sp2 geometry in the TS for
this step.
In our earlier work using similar methods, we measured KIE
values on the ring-oxygen insertion step for HPPD using ring-
per-deutero-HPP and, on the hydrogen abstraction step of
HMS, using 3′,3′-dideutero-HPP. We observed an inverse
secondary isotope effect for ring hydroxylation in HPPD that is
consistent with the formation of an epoxide as the initial
product of oxygen addition to the aromatic ring. For HMS,
abstraction of the S-deuteron from HPP dideuterated at the
benzylic position gave a normal primary effect consistent with
25
H-atom abstraction before hydroxylation (Figure 1). In the
present study, we have again investigated one branch point for
each enzyme (Scheme 2). For HPPD, we have studied a branch
point beyond ring-oxygen insertion from which either the
native product HG or QAA is formed. This branch was probed
using 3′,3′-dideutero-HPP to examine if the formation rate
constant for HG is influenced by geometric changes at the
benzylic carbon during the NIH shift. The shift reaction was
further studied using R-3′-monodeutero-HPP to detect
evidence of racemization at the benzylic carbon during the
aceto group migration. For HMS, we have expanded our prior
investigation of the hydroxylation step with the use of R-3′-
monodeutero-HPP. The single enantiomer of the substrate
permitted the measurement of the contribution from the
nonabstracted deuteron to the KIE value measured for HMS
variants in prior work and therefore has provided evidence for
DISCUSSION
■
Both HPPD and HMS have been studied in single turnover
reactions in some detail.30−32 Although both enzymes
accumulate multiple transients, none of the steps observed
show sensitivity to isotopic substitutions at positions of the
substrate that are involved in hydroxylation.30 As such, it was
concluded that the hydroxylation steps of the catalytic cycle
were fast relative to other steps, so the reactive species involved
could not be observed directly by kinetic methods. Recently, we
have shown that active site variants of both enzymes can form
6104
dx.doi.org/10.1021/bi400534q | Biochemistry 2013, 52, 6097−6107

Biochemistry
Article
the hybridization state at the benzylic carbon atom in the TS
for the step that precedes hydroxylation.
both HPPD (F364I) and HMS (S201A) are internally
consistent in that each value determined with the mono-
deuterated substrate is smaller than the value recorded with the
3′,3′-dideutero substrate. For S201A HMS, KIE values with
3′,3′-dideutero-HPP are higher than with R-3′-monodeutero-
HPP by a factor of 2, indicating a primary KIE for abstraction,
which suggests a more optimized benzylic hydroxylation
process in HMS compared to that of the HPPD F364I variant.
However, the KIE obtained with R-3′-monodeutero-HPP is,
within error, unity; this indicates that the benzylic carbon does
not change geometry significantly to reach the hydrogen
abstraction TS. The conclusion in this case is that the TS for
this step has sp3 character at C3′. Interestingly, Wojcik et al.
proposed that the S201 residue is vital for hydroxylation
regiospecificity by tethering the phenol of HPA at unique
angles in HPPD and HMS, so mutation of this residue may
have resulted in a change in geometry in the abstraction TS.27
For F364I HPPD, the KIE values for H-atom abstraction TS
are 1.83 for hydrogen abstraction and 1.92 for geometry
changes at the benzylic carbon in this step, suggesting that the
benzylic carbon develops predominant sp2 character with
relatively little lengthening of the bond to the abstracted
hydrogen in the TS. These values are in line with the prior
modeling studies and raise the possibility for redundancy in the
chemistry, where the dominant chemistry is to form a TS that is
largely sp2 hybridized at C3′, but an HMS variant can be made
to exhibit sp3 character at this position for the TS of this step
and yet yield the same product.
The hydroxylation/shift reaction of HPPD has long been a
biochemical curiosity and has been investigated using
experimental and computational methods. The earliest direct
evidence for the chemistry of the shift in HPPD came from
Leinberger et al., who were able to show that the configuration
of the benzylic carbon was retained during the shift and
proposed that a two-electron process (heterolytic pathway) was
at work.22 The conclusions made here differ from those of
Leinberger. Our KIE data show that no isotope effect is
observed for the formation of HG when 3′,3′-dideutero-HPP is
used as a substrate (Table 1). Although this alone could be said
to be consistent with a TS leading to a benzylic carbanion (sp3)
intermediate, it is also consistent with an sp3 hybridized TS that
exhibits little of the character of an ensuing sp2 hybridized
intermediate (i.e., an early TS). The more definitive
observation is the KIE value measured for this branch with
the N245S variant, which is distinctly normal (1.4). If we accept
that the mechanism is conserved between the wild-type and
variant, this value represents a more mature TS in which the
benzylic methylene is largely departed from the ring C1 and has
adopted the geometry of an untethered radical (sp2) (scheme
in Table 1).
Leinberger et al. showed both that configuration at the
benzylic position was retained and that the product HG was
not racemized to any measurable extent.22 This important work
established a fundamental facet of HPPD chemistry. However,
the methods chosen required derivatization of HG in order to
compare the chirality with synthesized chirality standards,
undermining to some extent the certainty of the observation. In
order to solidify the conclusions regarding racemization during
the NIH shift, we repeated the scrambling experiments of
Leinberger et al. using a chiral shift reagent to induce chirality, a
method that has the advantage of not chemically altering the
product. Our data confirm the earlier observations that the
monodeuterated HG formed in the reaction of HPPD with R-
3′-monodeutero-HPP is a single enantiomer, so we conclude
simply that bond rotation does not occur during the migration
step (Figure 2). Borowski et al. have proposed that the NIH
shift occurs through a homolytic pathway that does not induce
racemization due to a relatively high barrier for bond rotation
compared to that of the radical collapse that forms the new
substituent bond.26 The integrated conclusion is thus in
agreement with that of Borowski and suggests homolytic
cleavage in the first shift TS followed by the facile collapse of a
biradical intermediate.
ASSOCIATED CONTENT
* Supporting Information
■
S
Product ratios from which KIE values were derived. This
material is available free of charge via the Internet at http://
pubs.acs.org.
AUTHOR INFORMATION
Corresponding Author
*G. R. Moran: e-mail, moran@uwm.edu; phone, (414) 229
5031; fax, (414) 229 5530.
■
Funding
This research was supported by a National Science Foundation
grant to G.R.M. (MCB0843619) and by a UWM Research
Growth Initiative grant to G.R.M.
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
Partitioning KIEs measured with deuteriums substituted at
the benzylic carbon of HPP for S201A HMS and F364I HPPD
for the product HMA report the geometry of the H-atom
abstraction TS. Our data indicate that these variants undergo
different geometry changes at the benzylic carbon during the
initial H-atom abstraction. KIE values measured using 3′,3′-
dideutero-HPP are multiplicative for the hydrogen abstracted
and the hydrogen retained and thus are expected to be greater
than KIE values measured with R-3′-monodeuterated HPP
(given that inverse effects are not predicted). Wojcik et al. and
Neidig et al. both concluded from density functional
approaches that the HMS hydroxylation chemistry involves
H-atom abstraction leading to formation of a largely planar
(predominately sp2) benzylic radical.27,28 Our conclusions
regarding this step deviate only slightly from these prior
determinations. The measured KIEs for HMA formation with
We would like to express our gratitude to Dr. Kiochi Tanaka of
Kansai University, Osaka, Japan, for supplying the rhombamine
macrocycle used as a chiral shift reagent for NMR analysis of
homogentiste. These are an exceptional collection of molecules
that allow chirality determinations of aromatic acids. We are
also grateful to Holger Forsterling of the University of
Wisconsin−Milwaukee for his continued assistance with
operation of the 500 MHz NMR spectrometer.
ABBREVIATIONS
■
HPPD, 4-hydroxyphenylpyruvate dioxygenase; HPP, 4-hydrox-
yphenylpyruvate; HG, 2,5-dihydroxyphenylacetate, homogenti-
sate; QAA, quinolacetic acid; HMS, hydroxymandelate
synthase; HMA, hydroxymandelate; HPA, 4-hydroxyphenyla-
cetate; α-KAO, α-keto acid-dependent oxygenase; βME, β-
6105
dx.doi.org/10.1021/bi400534q | Biochemistry 2013, 52, 6097−6107

Biochemistry
Article
II. Synthesis of 3,4-toluene-4−2H oxide and subsequent ″NIH shift″
to 4-hydroxytoluene-3−2H. J. Am. Chem. Soc. 90, 6523−6525.
(19) Jerina, D. M., Daly, J. W., and Witkop, B. (1971) Migration of
Substituents during Hydroxylation of Aromatic Substrates (NIH
Shift). Oxidations with Peroxytrifluoroacetic Acid. Biochemistry 10,
366−372.
(20) Daly, J. W., Jerina, D. M., and Witkop, B. (1972) Arene oxides
and the NIH shift: The metabolism, toxicity and carcinogenicity of
aromatic compounds. Experientia 28, 1129−1264.
(21) Bowman, W. R., Gretton, W. R., and Kirby, G. W. (1980)
Hydroxylation of phenylalanine by Pseudomonas sp.: Measurement of
an isotope effect following the NIH shift. Perkin Trans. 1 (1972−1999)
3, 218−220.
(22) Leinberger, R., Hull, W. E., Simon, H., and Retey, J. (1981)
Steric course of the NIH shift in the enzymic formation of
homogentisic acid. Eur. J. Biochem. 117, 311−318.
mercaptoethanol; NIH, National Institutes of Health; KIE,
kinetic isotope effect; HPLC, high-pressure liquid chromatog-
raphy; PPT, phenylpyruvate tautomerase; NMR, nuclear
magenetic resonance
REFERENCES
■
(1) Diebold, A. R., Brown-Marshall, C. D., Neidig, M. L., Brownlee, J.
M., Moran, G. R., and Solomon, E. I. (2011) Activation of alpha-Keto
Acid-Dependent Dioxygenases: Application of an {FeNO}(7)/
{FeO(2)}(8) Methodology for Characterizing the Initial Steps of
O(2) Activation. J. Am. Chem. Soc. 133, 18148−18160.
(2) Brownlee, J., He, P., Moran, G. R., and Harrison, D. H. (2008)
Two roads diverged: the structure of hydroxymandelate synthase from
Amycolatopsis orientalis in complex with 4-hydroxymandelate.
Biochemistry 47, 2002−2013.
(23) Moran, G. R., Phillips, R. S., and Fitzpatrick, P. F. (1999)
Influence of steric bulk and electrostatics on the hydroxylation
regiospecificity of tryptophan hydroxylase: characterization of
methyltryptophans and azatryptophans as substrates. Biochemistry 38,
16283−16289.
(24) Hillas, P. J., and Fitzpatrick, P. F. (1996) A mechanism for
hydroxylation by tyrosine hydroxylase based on partitioning of
substituted phenylalanines. Biochemistry 35, 6969−6975.
(25) Shah, D. D., Conrad, J. A., Heinz, B., Brownlee, J. M., and
Moran, G. R. (2011) Evidence for the mechanism of hydroxylation by
4-hydroxyphenylpyruvate dioxygenase and hydroxymandelate synthase
from intermediate partitioning in active site variants. Biochemistry 50,
7694−7704.
(26) Borowski, T., Bassan, A., and Siegbahn, P. E. M. (2004) 4-
hydroxyphenylpyruvate dioxygenase: A hybrid density functional study
of the catalytic reaction mechanism. Biochemistry 43, 12331−12342.
(27) Wojcik, A., Broclawik, E., Siegbahn, P. E., and Borowski, T.
(2012) Mechanism of benzylic hydroxylation by 4-hydroxymandelate
synthase. A computational study. Biochemistry 51, 9570−9580.
(28) Neidig, M. L., Decker, A., Choroba, O. W., Huang, F., Kavana,
M., Moran, G. R., Spencer, J. B., and Solomon, E. I. (2006)
Spectroscopic and electronic structure studies of aromatic electrophilic
attack and hydrogen-atom abstraction by non-heme iron enzymes.
Proc. Natl. Acad. Sci. U. S. A. 103, 12966−12973.
(29) Price, J. C., Barr, E. W., Hoffart, L. M., Krebs, C., and Bollinger,
J. M., Jr. (2005) Kinetic Dissection of the Catalytic Mechanism of
Taurine:alpha-Ketoglutarate Dioxygenase (TauD) from Escherichia
coli. Biochemistry 44, 8138−8147.
(30) Johnson-Winters, K., Purpero, V. M., Kavana, M., and Moran, G.
R. (2005) Accumulation of Multiple Intermediates in the Catalytic
Cycle of (4-Hydroxyphenyl)pyruvate Dioxygenase from Streptomyces
avermitilis. Biochemistry 44, 7189−7199.
(31) Purpero, V. M., and Moran, G. R. (2006) Catalytic, noncatalytic,
and inhibitory phenomena: kinetic analysis of (4-hydroxyphenyl)-
pyruvate dioxygenase from Arabidopsis thaliana. Biochemistry 45,
6044−6055.
(32) He, P., Conrad, J. A., and Moran, G. R. (2010) The rate-limiting
catalytic steps of hydroxymandelate synthase from Amycolatopsis
orientalis. Biochemistry 49, 1998−2007.
(3) Hubbard, B. K., Thomas, M. G., and Walsh, C. T. (2000)
Biosynthesis of L-p-hydroxyphenylglycine, a non-proteinogenic amino
acid constituent of peptide antibiotics. Chem. Biol. 7, 931−942.
(4) Li, T. L., Choroba, O. W., Charles, E. H., Sandercock, A. M.,
Williams, D. H., and Spencer, J. B. (2001) Characterisation of a
hydroxymandelate oxidase involved in the biosynthesis of two unusual
amino acids occurring in the vancomycin group of antibiotics. Chem.
Comm. (Cambridge, England), 1752−1753.
(5) Hausinger, R. P. (2004) FeII/alpha-ketoglutarate-dependent
hydroxylases and related enzymes. Crit. Rev. Biochem. Mol. Biol. 39,
21−68.
(6) Purpero, V., and Moran, G. R. (2007) The diverse and pervasive
chemistries of the alpha-keto acid dependent enzymes. JBIC 12, 587−
601.
(7) Armstrong, R. N. (2000) Mechanistic diversity in a metal-
loenzyme superfamily. Biochemistry 39, 13625−13632.
(8) He, P., and Moran, G. R. (2011) Structural and mechanistic
comparisons of the metal-binding members of the vicinal oxygen
chelate (VOC) superfamily. J. Inorg. Biochem. 105, 1259−1272.
(9) Zhang, Z., Ren, J., Harlos, K., McKinnon, C. H., Clifton, I. J., and
Schofield, C. J. (2002) Crystal structure of a clavaminate synthase-
Fe(II)-2-oxoglutarate- substrate-NO complex: evidence for metal
centred rearrangements. FEBS Lett. 517, 7−12.
(10) Price, J. C., Barr, E. W., Glass, T. E., Krebs, C., and Bollinger, J.
M., Jr. (2003) Evidence for hydrogen abstraction from C1 of taurine
by the high-spin Fe(IV) intermediate detected during oxygen
activation by taurine:alpha-ketoglutarate dioxygenase (TauD). J. Am.
Chem. Soc. 125, 13008−13009.
(11) Galonic, D. P., Barr, E. W., Walsh, C. T., Bollinger, J. M., Jr., and
Krebs, C. (2007) Two interconverting Fe(IV) intermediates in
aliphatic chlorination by the halogenase CytC3. Nat. Chem. Biol. 3,
113−116.
(12) Hoffart, L. M., Barr, E. W., Guyer, R. B., Bollinger, J. M., Jr., and
Krebs, C. (2006) Direct spectroscopic detection of a C-H-cleaving
high-spin Fe(IV) complex in a prolyl-4-hydroxylase. Proc. Natl. Acad.
Sci. U. S. A. 103, 14738−14743.
(13) Guroff, G., Daly, J. W., Jerina, D. M., Renson, J., Witkop, B., and
Udenfriend, S. (1967) Hydroxylation-induced migration: The NIH
shift. Science 157, 1524−1530.
(14) Hillas, P. J., and Fitzpatrick, P. F. (1996) A mechanism for
hydroxylation by tyrosine hydroxylase based on partitioning of
substituted phenylalanines. Biochemistry 35, 6969−6975.
(15) Moran, G. R., Derecskei-Kovacs, A., Hillas, P. J., and Fitzpatrick,
P. F. (2000) On the Catalytic Mechanism of Tryptophan Hydroxylase.
J. Am. Chem. Soc. 122, 4535−4541.
(16) Mitchell, K. H., Rogge, C. E., Gierahn, T., and Fox, B. G. (2003)
Insight into the mechanism of aromatic hydroxylation by toluene 4-
monooxygenase by use of specifically deuterated toluene and p-xylene.
Proc. Natl. Acad. Sci. U. S. A. 100, 3784−3789.
(33) Brownlee, J. M., Heinz, B., Bates, J., and Moran, G. R. (2010)
Product analysis and inhibition studies of a causative Asn to Ser variant
of 4-hydroxyphenylpyruvate dioxygenase suggest a simple route to the
treatment of Hawkinsinuria. Biochemistry 49, 7218−7226.
(34) Fitzpatrick, P. F. (2006) in Isotope Effects in Chemsitry and
Biology (Kohen, A., and Limbach, H.-H., Eds.), pp 861−873, CRC
Taylor & Francis, Boca Raton, FL.
(35) Korzekwa, K. R., Trager, W. F., and Gillette, J. R. (1989) Theory
for the observed isotope effects from enzymatic systems that form
multiple products via branched reaction patyways: Cytochrome P-450.
Biochemistry 28, 9012−9018.
(36) Darbyshire, J. F., Iyer, K. R., Grogan, J., Korzekwa, K. R., and
Trager, W. F. (1996) Substrate probe for the mechanism of aromatic
(17) Sono, M., Roach, M. P., Coulter, E. D., and Dawson, J. H.
(1996) Heme-containing oxygenases. Chem. Rev. 96, 2841−2888.
(18) Jerina, D. M., Daly, J. W., and Witkop, B. (1968) The role of
arene oxide-oxepin systems in the metabolism of aromatic substrates.
6106
dx.doi.org/10.1021/bi400534q | Biochemistry 2013, 52, 6097−6107

Biochemistry
Article
hydroxylation catalyzed by cytochrome P450. Drug Metab. Dispos. 24,
1038−1045.
(37) Frantom, P. A., and Fitzpatrick, P. F. (2003) Uncoupled forms
of tyrosine hydroxylase unmask kinetic isotope effects on chemical
steps. J. Am. Chem. Soc. 125, 16190−16191.
(38) Tanaka, K., Nakai, Y., and Takahashi, H. (2011) Efficient NMR
chiral discrimination of carboxylic acids using rhombamine macro-
cycles as chiral shift reagent. Tetrahedron: Asymmetry 22, 178−184.
(39) Johnson-Winters, K., Purpero, V. M., Kavana, M., Nelson, T.,
and Moran, G. R. (2003) 4-Hydroxyphenyl)pyruvate Dioxygenase
from Streptomyces avermitilis: The Basis for Ordered Substrate
Addition. Biochemistry 42, 2072−2080.
(40) Retey, J., Bartl, K., Ripp, E., and Hull, W. E. (1977)
Stereospecificity of phenylpyruvate tautomerase. A convenient method
for the preparation of chirally labelled phenylpyruvates. Euro. J.
Biochem. 72, 251−257.
(41) Johnson, W. H., Jr., Czerwinski, R. M., Stamps, S. L., and
Whitman, C. P. (1999) A kinetic and stereochemical investigation of
the role of lysine-32 in the phenylpyruvate tautomerase activity
catalyzed by macrophage migration inhibitory factor. Biochemistry 38,
16024−16033.
(42) O’Hare, H. M., Huang, F., Holding, A., Choroba, O. W., and
Spencer, J. B. (2006) Conversion of hydroxyphenylpyruvate
dioxygenases into hydroxymandelate synthases by directed evolution.
FEBS Lett. 580, 3445−3450.
(43) Gunsior, M., Ravel, J., Challis, G. L., and Townsend, C. A.
(2004) Engineering p-hydroxyphenylpyruvate dioxygenase to a p-
hydroxymandelate synthase and evidence for the proposed benzene
oxide intermediate in homogentisate formation. Biochemistry 43, 663−
674.
(44) Brownlee, J., Johnson-Winters, K., Harrison, D. H. T., and
Moran, G. R. (2004) The Structure of the Ferrous Form of (4−
Hydroxyphenyl)pyruvate Dioxygenase from Streptomyces avermitilis
in Complex with the Therapeutic Herbicide, NTBC. Biochemistry 43,
6370−6377.
6107
dx.doi.org/10.1021/bi400534q | Biochemistry 2013, 52, 6097−6107