Catalytic mechanism of a heme and tyrosyl radical-containing fatty acid α-(di)oxygenase

Article abstract of DOI:10.1021/ja104180v

The steady-state catalytic mechanism of a fatty acid α-(di)oxygenase is examined, revealing that a persistent tyrosyl radical (Tyr379 ^?) effects O₂ insertion into C_α-H bonds of fatty acids. The initiating C_α-H homolysis step is characterized by apparent rate constants and deuterium kinetic isotope effects (KIEs) that increase hyperbolically upon raising the concentration of O ₂. These results are consistent with H^? tunneling, transitioning from a reversible to an irreversible regime. The limiting deuterium KIEs increase from ～30 to 120 as the fatty acid chain is shortened from that of the native substrate. In addition, activation barriers increase in a manner that reflects decreased fatty acid binding affinities. Anaerobic isotope exchange experiments provide compelling evidence that Tyr379 ^? initiates catalysis by H^? abstraction. C _α-H homolysis is kinetically driven by O₂ trapping of the α-carbon radical and reduction of a putative peroxyl radical intermediate to a 2(R)-hydroperoxide product. These findings add to a body of work which establishes large-scale hydrogen tunneling in proteins. This particular example is novel because it involves a protein-derived amino acid radical.

Full text of DOI:10.1021/ja104180v

Published on Web 12/17/2010
Catalytic Mechanism of a Heme and Tyrosyl
Radical-Containing Fatty Acid r-(Di)oxygenase
Arnab Mukherjee, Alfredo M. Angeles-Boza, Gregory S. Huff, and Justine P. Roth*
Department of Chemistry, Johns Hopkins UniVersity, 3400 North Charles Street, Baltimore,
Maryland 21218, United States
Received May 14, 2010; E-mail: jproth@jhu.edu
Abstract: The steady-state catalytic mechanism of a fatty acid R-(di)oxygenase is examined, revealing
that a persistent tyrosyl radical (Tyr379•) effects O2 insertion into CR-H bonds of fatty acids. The initiating
CR-H homolysis step is characterized by apparent rate constants and deuterium kinetic isotope effects
(KIEs) that increase hyperbolically upon raising the concentration of O2. These results are consistent with
H• tunneling, transitioning from a reversible to an irreversible regime. The limiting deuterium KIEs increase
from ∼30 to 120 as the fatty acid chain is shortened from that of the native substrate. In addition, activation
barriers increase in a manner that reflects decreased fatty acid binding affinities. Anaerobic isotope exchange
experiments provide compelling evidence that Tyr379• initiates catalysis by H• abstraction. CR-H homolysis
is kinetically driven by O2 trapping of the R-carbon radical and reduction of a putative peroxyl radical
intermediate to a 2(R)-hydroperoxide product. These findings add to a body of work which establishes
large-scale hydrogen tunneling in proteins. This particular example is novel because it involves a protein-
derived amino acid radical.
diates in photosystem II,2,10 ribonucleotide reductase,11 cyto-
Introduction
chrome c oxidase,12 and several heme-containing peroxidases13
as well as dioxygenases.3,4b,14-16 The fatty acid R-(di)-
oxygenase from rice (RRO) is the focus of the present study.
In RRO, a Tyr• is produced as a result of oxidizing ferric
protoporphyrin IX, FeIII(Por), with H2O2. Coordination of H2O2
followed by O-O bond heterolysis generates a ferryl π-cation
radical, FeIVdO(Por•+), which may react by intraprotein proton-
coupled electron transfer,3 producing FeIVdO(Por), a protonated
base, and Tyr379• as shown in eq 1.
There is substantial evidence that amino acid radicals act as
cofactors in a number of redox-active metalloenzymes,1-4 yet
detailed mechanistic studies have been somewhat limited.5 This
is particularly true of heme proteins where spectroscopic features
of the radical are obscured by the Soret absorption band of the
prosthetic group, making it a challenge to evaluate catalytic
competence.6,7 Examination of the kinetics and thermodynamics
that characterize amino acid radical-mediated hydrogen atom
transfer is needed to characterize this seemingly general strategy
in enzyme catalysis.1-6,8
RRO is a plant-derived pathogen-inducible (di)oxygenase
(PIOX).15-17 Members of this family of heme proteins possess
a conserved Tyr and exhibit roughly 30% structural homology
Tyrosyl radicals (Tyr•) are particularly widespread, having
been spectroscopically observed9 and/or implicated as interme-
(10) (a) Jenson, D. L.; Barry, B. A. J. Am. Chem. Soc. 2009, 131, 10567–
10573. (b) Kuehne, H.; Brudvig, G. W. J. Phys. Chem. B 2002, 106,
8189–8196.
(1) (a) Stubbe, J.; van der Donk, W. A. Chem. ReV. 1998, 98, 705–762.
(b) Pesavento, R. P.; van der Donk, W. A. AdV. Protein Chem. 2001,
58, 317.
(11) Stubbe, J. Curr. Opin. Chem. Biol. 2003, 7, 183–188.
(12) (a) Siletsky, S. A.; Han, D.; Brand, S.; Morgan, J. E.; Fabian, M.;
Geren, L.; Millett, F.; Durham, B.; Konstantinov, A. A.; Gennis, R. B.
Biochim. Biophys. Acta, Bioenerg. 2006, 1757, 1122–1132. (b) Bu,
Y.; Cukier, R. I. J. Phys. Chem. B 2005, 109, 22013–22026. (c)
Proshlyakov, D. A.; Pressler, M. A.; Babcock, G. T. Proc. Natl. Acad.
Sci. U.S.A. 1998, 95, 8020–8025.
(2) Diner, B. A.; Britt, R. D. AdV. Photosynth. Respir. 2005, 22, 207–
233.
(3) Rouzer, C. A.; Marnett, L. J. Chem. ReV. 2003, 103, 2239–2304.
(4) Evidence of catalytic tyrosyl radicals in plant-derived heme-containing
dioxygenases:(a) Su, C.; Sahlin, M.; Oliw, E. H. J. Biol. Chem. 1998,
273, 20744–20751. (b) Gupta, A.; Mukherjee, A.; Matsui, K.; Roth,
J. P. J. Am. Chem. Soc. 2008, 130, 11274–11275.
(13) (a) Furtmueller, P. G.; Zederbauer, M.; Jantschko, W.; Helm, J.;
Bogner, M.; Jakopitsch, C.; Obinger, C. Arch. Biochem. Biophys. 2006,
445, 199–213. (b) Fielding, A. J.; Singh, R.; Boscolo, B.; Loewen,
P. C.; Ghibaudi, E. M.; Ivancich, A. Biochemistry 2008, 47, 9781–
9792. (c) Divi, R. L.; Doerge, D. R. Biochemistry 1994, 33, 9668–74.
(d) Ehrenshaft, M.; Mason, R. P. Free Radical Biol. Med. 2006, 41,
422–30.
(5) (a) Licht, S.; Gerfen, G. J.; Stubbe, J. Science 1996, 271, 477–81. (b)
Aubert, C.; Vos, M. H.; Mathis, P.; Eker, A. P. M.; Brettel, K. Nature
2000, 405, 586–590. (c) Mozziconacci, O.; Williams, T. D.; Kerwin,
B. A.; Schoneich, C. J. Phys. Chem. B 2008, 112, 15921–15932.
(6) Tsai, A.-L.; Kulmacz, R. J. Arch. Biochem. Biophys. 2010, 493, 103–
24.
(14) Kulmacz, R. J.; van der Donk, W. A.; Tsai, A.-L. Prog. Lipid Res.
2003, 42, 377–404.
(7) Wu, G.; Rogge, C. E.; Wang, J.-S.; Kulmacz, R. J.; Palmer, G.; Tsai,
A.-L. Biochemistry 2007, 46, 534–542.
(15) Hamberg, M.; Ponce de Leon, I.; Sanz, A.; Castresana, C. Prostag-
landins Other Lipid Mediators 2002, 68/69, 363–374.
(16) Koeduka, T.; Matsui, K.; Akakabe, Y.; Kajiwara, T. J. Biol. Chem.
2002, 277, 22648–22655.
(8) Himo, F.; Siegbahn, P. E. M. Chem. ReV. 2003, 103, 2421–2456.
(9) (a) Svistunenko, D. A.; Jones, G. A. Phys. Chem. Chem. Phys. 2009,
11, 6600–6613. (b) Svistunenko, D. Biochim. Biophys. Acta 2005,
1707, 127–155.
(17) Smith, W. L.; DeWitt, D. L.; Garavito, R. M. Annu. ReV. Biochem.
2000, 69, 145–182.
9
10.1021/ja104180v  2011 American Chemical Society
J. AM. CHEM. SOC. 2011, 133, 227–238 227

A R T I C L E S
Mukherjee et al.
metalloporphyrin edge, have been constructed on the basis of a
COX-1 structure.27 Given the inherent uncertainty of homology
models, it is unclear whether a nearby histidine, His311, is close
enough to influence formation of Tyr379• through hydrogen
bonding28 or whether it is strategically placed, remote from the
catalytic radical, where it can anchor the carboxylate tail of the
fatty acid.17,29
Fatty acid oxidation by RRO produces 2(R)-hydroperoxide
compounds (eq 2).16 These species are only moderately stable
at low temperatures and undergo spontaneous (non-enzymatic)
decarboxylation to form Cn-1 aldehydes, as depicted in eq 2.
Although the aldehyde is the only detectable product after
workup,4b,16 a small fraction of the initially formed 2(R)-
hydroperoxide is likely to react with FeIII-RRO by 2e- oxidation.
Such a reaction, where the product activates the enzyme,
accounts for the ability of RRO to catalyze fatty acid oxidation
in the absence of added radical initiators, such as H2O2.
to the mammalian prostaglandin H2 synthases, also known as
the cyclooxygenases (COX-1 and COX-2).16 The cyclooxyge-
nases catalyze the first dedicated step in prostaglandin biosyn-
thesis from arachidonic acid, thereby controlling the formation
of prostanoids that effect a range of physiological and patho-
physiological functions; inflammation is perhaps the most
recognized.17-19 Analogously, the PIOX family of proteins
produces oxylipoxins that participate in cell signaling, wound
healing, and the plant’s immune response.15
Studies employing site-directed mutagenesis and electron
paramagnetic resonance (EPR) spectroscopy have provided
evidence that the Tyr379• is required for catalytic functioning
of RRO.4b,16,20 The mutant Tyr379Phe-RRO is devoid of fatty
acid dioxygenase activity as well as the persistent EPR signal
found in samples of the wild-type (wt) RRO at ambient
temperature. The stability of the FeIVdO(Por•+) is enhanced in
Tyr379Phe-RRO relative to wt-RRO and other heme proteins.21
The quantitative formation of this species can be observed,
followed by its slow decay to FeIVdO(Por) and eventually back
to FeIII(Por) over several seconds.
RRO provides an ideal model system for understanding COX
and related enzymes that catalyze reactions using amino acid
radicals.1-8 Though mutagenesis studies have implicated the
conserved Tyr• in dioxygenase activity,4b,22,23 questions remain
as to how the radical is formed and endowed with the
thermodynamic and kinetic propensity to oxidize fatty acids.
RRO reacts with its substrates specifically at the acidic CR-H
bond adjacent to the carboxylate. Detailed studies of RRO
catalysis have been facilitated by the large quantities of pure
protein that can be isolated, together with the protein’s stability
with respect to oxidative damage.
In this paper we describe the steady-state catalytic mechanism
of RRO revealed by studies of initial rates, substrate deuterium
kinetic isotope effects (KIEs), and solvent isotope exchange.
Saturated fatty acids (FAs) of variable chain length, hexade-
canoic acid (16:0), dodecanoic acid (12:0), and decanoic acid
(10:0), are used to establish the kinetic mechanism as well as
the limiting rate constants and KIEs that reflect homolytic
cleavage of the CR-H bond. The molecular oxygen concentra-
tion is shown to affect the reversibility of initial H• abstraction
from the fatty acid. Analysis of the rate-limiting step at low O2
concentrations, where CR-H homolysis is reversible, is the
subject of a separate study.30
Under conditions of O2 saturation, CR-H homolysis is
characterized by large deuterium KIEs which range from ∼30
to 120. These effects are consistent with extensive nuclear
tunneling in the turnover-controlling step, as proposed in
preliminary studies with the native substrate, 16:0.4b Here it is
demonstrated that the deuterium KIEs become larger upon
shortening the fatty acid chain. Inflation of the deuterium KIE
is accompanied by a marked decrease in kcat/KM(FA), by 2 orders
of magnitude, but little variation in kcat. Thus, while nuclear
tunneling is an inherent feature of H• abstraction from C-H
bonds, it is not obvious to what extent the probability of this
reaction is influenced by RRO,31 nor is it evident how the
enzyme controls O2 addition to the incipient R-carbon radical,
RRO is recombinantly expressed in Escherichia coli, largely
in the form of the FeIII(Por)-containing holoprotein.16,24,25
Though a crystal structure is not yet available,26 homology
models4b,17 featuring a conserved tyrosine, ∼6 Å from the
(18) Choi, S.-H.; Aid, S.; Bosetti, F. Trends Pharmacol. Sci. 2009, 30,
174–181.
(25) Typical Rz values are g0.4 for the isolated holoproteins. The addition
of neither hemin nor hematin results in a discernible increase in specific
activity.
(26) Lloyd, T.; Krol, A.; Campanaro, D.; Malkowski, M. Acta Crystallogr.,
Sect. F 2006, F62, 365–367.
(19) Marnett, L. J. Annu. ReV. Pharmacol. Toxicol. 2009, 49, 265–290.
(20) Koszelak-Rosenblum, M.; Krol, A. C.; Simmons, D. M.; Goulah, C. C.;
Wroblewski, L.; Malkowski, M. G. J. Biol. Chem. 2008, 283, 24962–
24971.
(21) Dunford, H. B. Heme Peroxidases; Wiley: New York, 1999.
(22) (a) Karthein, R.; Dietz, R.; Nastainczyk, W.; Ruf, H. H. Eur.
J. Biochem. 1988, 171, 313–20. (b) Dietz, R.; Nastainczyk, W.; Ruf,
H. H. Eur. J. Biochem. 1988, 171, 321–8.
(27) Malkowski, M. G.; Ginell, S. L.; Smith, W. L.; Garavito, R. M. Science
2000, 289, 1933–1937.
(28) Wilson, J. C.; Wu, G.; Tsai, A.-L.; Gerfen, G. J. J. Am. Chem. Soc.
2005, 127, 1618–1619.
(23) Mukherjee, A.; Brinkley, D. W.; Chang, K.-M.; Roth, J. P. Biochem-
istry 2007, 46, 3975–3989.
(29) Bhattacharyya, D. K.; Lecomte, M.; Rieke, C. J.; Garavito, R. M.;
Smith, W. L. J. Biol. Chem. 1996, 271, 2179–84.
(30) Huff, G. S.; Doncheva, I. S.; Angeles-Boza, A. M.; Mukherjee, A.;
Brinkley, D. W. Cramer, C. J.; Roth, J. P. Manuscript in preparation.
(31) (a) Benkovic, S. J.; Hammes-Schiffer, S. Science 2006, 312, 208–
209. (b) Hammes-Schiffer, S.; Benkovic, S. J. Annu. ReV. Biochem.
2006, 75, 519–541.
(24) A homologue from Arabidopsis thaliana has also been described: (a)
Liu, W.; Rogge, C. E.; Bambai, B.; Palmer, G.; Tsai, A.-L.; Kulmacz,
R. J. J. Biol. Chem. 2004, 279, 29805–29815. (b) Liu, W.; Wang,
L.-H.; Fabian, P.; Hayashi, Y.; McGinley, C. M.; van der Donk, W. A.;
Kulmacz, R. J. Plant Physiol. Biochem. 2006, 44, 284–293.
9
228 J. AM. CHEM. SOC. VOL. 133, NO. 2, 2011

Catalytic Mechanism of a Fatty Acid R-(Di)oxygenase
A R T I C L E S
Removal of these additives allowed for control experiments that
tested the possibility of kinetic artifacts arising from the presence
of salt or detergent.37
resulting in the formation of the 2(R)-hydroperoxide product
with high regio- and stereochemical fidelity.
Experimental Section
The concentrations of wt-RRO and Tyr379Phe-RRO were
determined by UV-vis spectrophotometry using the molar extinc-
tion coefficients of the FeIII(Por) Soret absorption band. The pyridine
hemochrome assay38 was performed under oxidizing and reducing
conditions to obtain average values of ε412 nm ) 123 000 M-1 cm-1
for wt-RRO and ε410 nm ) 157 000 M-1cm-1 for the Tyr379Phe
mutant. The wt enzyme’s specific activity, optimized by addition
of H2O2,37 correlated to the concentration of FeIII(Por) determined
spectrophotometrically. This relationship, which held for multiple
preparations of the wt protein, was used to develop a standard assay
for the determination of the active enzyme concentration. Under
no conditions was fatty acid dioxygenase activity detectable in the
Tyr379Phe mutant.
General Procedures. Chemical reagents were obtained from
Sigma-Aldrich in the highest purity available (>98%) and used
without further purification unless noted. Deuterium-labeled materi-
als were obtained from Cambridge Isotope Laboratories and used
as received. Hydrogen peroxide (30%) was obtained from Fisher
and its concentration determined using the well-known optical
extinction coefficient ε240 nm ) 43.6 M-1 cm-1. Substrate-containing
solutions were prepared just prior to use by dilution of the fatty
acids dissolved in ethanol into the reaction buffer and stirring under
an atmosphere of air, He/O2 or N2/O2.
Electronic absorption measurements were performed using an
Agilent 8453 diode array or OLIS stopped-flow RSM 1000
spectrophotometer. EPR spectra were recorded on a Bruker X-band
(9.395 GHz) spectrometer as previously described.4b The persistent
Tyr379• was further probed by power saturation experiments at 15
K. The power was varied from 20.1 µW to 20.1 mW while the
modulation amplitude was maintained at 10 G. P1/2, corresponding
to the power at half-saturation, was derived from the standard
equation: log(S/P1/2) ) -(b/2) log(P1/2 + P) + (b/2) log(P1/2) +
log K, where P is the power, S is the peak to trough amplitude of
the EPR signal, b is the inhomogeneity parameter (∼1), and K is
a floating parameter.
A 1:1 relationship between FeIII(Por) and active wt-RRO contain-
ing Tyr379• was assumed in the steady-state kinetic analysis. The
formation of the radical appears to require two-electron oxidation
of the FeIII prosthetic group to the FeIVdO(Por•+) state (cf. eq 1).
Only one oxidizing equivalent can be accounted for, however, by
formation of Tyr379•. Prior to treatment of wt-RRO with H2O2, a
trace amount of this radical with an EPR signature at g ) 2.0054,
is detectable at 295 K.4b Following the addition of 10-100 equiv
of H2O2, the EPR signal intensity increases as does the enzyme’s
fatty acid dioxygenase activity.4b,37
EPR spectroscopy, at ambient and at liquid He temperatures,
has been used to characterize the species formed upon manually
mixing H2O2 with wt- and Tyr379Phe-RRO.4b,37 In experiments
with the wt protein, the yield of the persistent radical at g ) 2.0054
is typically 25-30% on the basis of the concentration of FeIII(Por)
bound to the holoprotein. No EPR signal is detectable when the
Tyr379Phe-RRO is treated in an identical manner; however, the
mutant does undergo the expected two-electron oxidation by H2O2.
A stabilized FeIVdO(Por•+) is formed and then decays to
FeIVdO(Por) before returning to FeIII(Por) after several hundred
seconds at 295 K.4b Despite the marked differences in stability
associated with the ferryl states in the two proteins, wt-RRO and
Tyr379Phe-RRO exhibit comparable catalase-like activities; i.e.,
the H2O2 disproportionation reactions progress quantitatively,
producing 1 equiv of O2 and H2O for every 2 equiv of H2O2
consumed.37
Steady-State Kinetics. Initial rates were recorded at 295 K and
pH 7.2 in solutions containing 50 mM sodium phosphate (µ ) 0.1
M) unless noted otherwise. Though wt-RRO is able to catalyze
fatty acid oxidation in the absence of added H2O2, initial rates can
be 2-3 times less than the maximum for the slower substrates 12:0
and 10:0.37 In contrast, preincubation with H2O2 has a smaller effect
upon the initial rate of 16:0 oxidation. In most experiments, the
enzyme was pretreated with H2O2 (10-100 equiv) to optimize the
dioxygenase activity.37 In other instances, active enzyme concentra-
tions were determined by the standard assay, where V/[E]t ) 9.0
s-1 for air-saturated solutions containing 270 µM O2 and 100 µM
16:0 at 295 K.
Due to the importance of kinetic trends, a number of control
experiments were undertaken to probe for artifacts that could
manifest as inhibition or systematic changes in enzyme activity.
Prior to each initial rate determination, a background “drift” rate
of the O2 electrode was recorded and used to correct the experi-
mentally observed rate when necessary. In experiments with H2O2,
a short time was allotted so that O2 evolution due to RRO’s catalase-
like activity had ceased prior to initiation of the dioxygenase
reaction by addition of the fatty acid. In other experiments, which
were initiated by the addition of H2O2 pre-treated enzyme, the same
rate was obtained. It was regularly verified that O2 consumption
rates varied in direct proportion to the wt-RRO concentration. In
A Clark-type O2 electrode32 (Yellow Springs Inc.) was used to
record changes in the concentration of O2 with time. Readings from
the voltmeter were transmitted to a PC workstation with a 1 s
sampling time. The electrode probe, equipped with a 10 µL injection
port, was fitted inside a transparent water-jacketed reaction chamber,
connected to a temperature-controlled recirculating bath (VWR 116-
0S), and used atop a magnetic stirrer.
Density functional theory (DFT) calculations were performed
using Gaussian03.33 A calibrated method,30 employing the modified
Perdew-Wang (mPWPW91) functional and the following basis
set 6-311G* (O), 6-31G (C), and STO-3G (H), was used to obtain
energy-minimized structures which were subject to vibrational
frequency analyses.34 Calculations of deuterium equilibrium isotope
effects were performed by analyzing isotopic zero-point energy
differences associated with the symmetric C-H(D) and O-H(D)
stretches of butanoate and 4-hydroxyphenylacetate, which serve as
models for the longer chain fatty acids and reduced tyrosine.35 It
was confirmed that no other vibrational frequencies changed
significantly upon this isotopic substitution.
Protein Preparation, Characterization, and Quantification.
The genes encoding wt-RRO and Tyr379Phe-RRO were se-
quenced36 and overexpressed to obtain N-terminal His-tagged
proteins.4b,16 Affinity chromatography afforded proteins of g98%
purity as indicated by polyacrylamide gel electrophoresis and amino
acid sequencing (Texas A&M University Protein Chemistry Facil-
ity). A slight modification of the published procedure16 involved
dialysis of the concentrated protein against 50 mM sodium
phosphate buffer (pH 7.2) to remove most of the NaCl and Nonidet
P-40 or Nonidet P-40 substitute used during protein isolation.
(32) Corrections for “dampening effects” were deemed unnecessary: Tsai,
A.-L.; Wu, G.; Kulmacz, R. J. Biochemistry 1997, 36, 13085–13094.
The concentrations of O2 determined using the O2 electrode cor-
responded closely to the amount of aldehyde product formed as
indicated by GC-MS as well as the determination of CO2 by
manometry.
(33) Frisch, M. J.; et al. Gaussian 03W, revision C.02; Gaussian, Inc.:
Pittsburgh, PA, 2003. The complete citation appears in the Supporting
Information.
(34) Cramer, C. J. Essentials of Computational Chemistry, 2nd ed.; Wiley:
New York, 2004.
(35) Melander, L.; Saunders, W. H., Jr. Reaction Rates of Isotopic
Molecules, 2nd ed.; Wiley: New York, 1980.
(36) Sequencing results are provided in the Supporting Information of ref
4b.
(37) See the Supporting Information for details.
(38) Berry, E. A.; Trumptower, B. L. Anal. Biochem. 1987, 161, 1–15.
9
J. AM. CHEM. SOC. VOL. 133, NO. 2, 2011 229

A R T I C L E S
Mukherjee et al.
addition, the introduction of NaCl (e30 mM) and detergents such
as Nonidet P-40 or Nonidet P-40 substitute (e0.02%) was found
to have no effect on the initial rates.37
All kinetic data were analyzed using OLIS Globalworks or
Kaleidagraph 4.0 (Synergy Software). Accordingly, all rate con-
stants are reported with (2 standard errors. The apparent turnover
rate constant (apkcat) and the apparent Michaelis constant (apKM) were
routinely obtained by nonlinear curve fitting of initial rate data to
the Michaelis-Menten expression. The limiting steady-state kinetic
parameters kcat and kcat/KM were derived from the hyperbolic
ap
dependences of
k
cat
and apkcat/KM upon increasing the concentration
Figure 1. UV-vis absorption spectra acquired by rapid-mixing, stopped-
flow spectrophotometry upon reaction of RRO (4.5 µM) with 100 equiv of
H2O2. The initial spectrum is shown in red and the final spectrum in blue.
Spectra collected over the first 0.25 s are shown in the inset.
of the corresponding cosubstrate.
Deuterium Kinetic Isotope Effects. KIEs were determined for
the reactions of deuterium-labeled and unlabeled fatty acids under
competitive as well as noncompetitive conditions. As previously
described,4b,39,40 competitive KIEs were determined by reacting
wt-RRO with 5:1 or 10:1 mixtures of perdeutero to perprotio fatty
acids. Following acid quench and workup, the ratios of the aldehyde
products at conversions from 1% to 5% were compared to the same
ratio at 100% conversion. Measurements were performed by gas
chromatography-mass spectrometry (GC-MS) on a Shimadzu
GC17A/QP5050A equipped with a DB-5 ms column.
Noncompetitive KIEs were extracted from Michaelis-Menten
plots obtained independently for the protio and deutero fatty acids.
These experiments were performed at disparate enzyme concentra-
tions that were varied by 10-50-fold to achieve similar rates
for the labeled and unlabeled fatty acid substrates. Because
perdeuterated and R,R-dideuterated fatty acids were used in this
D
D
study, the observed kcat and kcat/KM(FA) consist of primary as
well as secondary KIE contributions. The latter is expected to be
e1.25 on the basis of precedents from related studies.41,42
Isotope Exchange. Isotope exchange experiments were con-
ducted under an atmosphere of N2, inside an MBraun glovebox, or
in the presence of air. The incorporation of a single proton (from
H2O) into the R-position of the deuterium-labeled fatty acid was
demonstrated under anaerobic conditions. A typical experiment
involved preincubating protein solutions (1-10 µM) with 10 equiv
of H2O2 and then stirring under N2. The protein was added to a
rigorously anaerobic solution of the deuterium-labeled fatty acid
present at 100 µM. The reaction was quenched at time intervals
between 0.5 and 8 h and the unreacted fatty acid isolated and
analyzed. All samples were prepared for analysis by dichlo-
romethane extraction of the acidified reaction mixture followed by
dilution into a 50:50 methanol/0.1% ammonium acetate buffer.
Control experiments, performed with solutions of deuterium-labeled
fatty acids in the absence of protein, were treated identically to
solutions containing either wt-RRO or the Tyr379Phe mutant. The
fatty acid isotope composition was quantified by electrospray
ionization mass spectrometry (ESI-MS) using a ThermoFinnigan
LCQ Deca operating in negative ion mode.
Figure 2. Power saturation of the EPR signal obtained upon reaction of
wt-RRO (40 µM) with 10 equiv of H2O2. Data were collected at 15 K and
analyzed as described in the Experimental Section. The inset shows the
fitted data used to determine the power at half-saturation (P1/2).
wt-FeIIIRRO with excess H2O2, in the absence of fatty acid, by
monitoring the reaction using stopped-flow spectrophotometry.
At concentrations of 1-5 µM wt-RRO, reactions with 50-500
equiv of H2O2 caused only slight spectral changes, as shown in
Figure 1. An average H2O2 independent rate constant of 62 (
12 s-1 was determined at 295 K. Optical changes on the
millisecond time scale were indistinguishable from those
acquired on longer time scales using conventional UV-vis
spectrophotometry.4b The spectrum of the product is character-
ized by a slight bathochromic (red) shift and bleach of the Soret
band. No evidence for FeIVdO(Por•+) or FeIVdO(Por) was
obtained; yet these species are readily observed when Tyr379Phe-
RRO is treated with H2O2 under the same conditions.4b
Results
Enzyme Activation. Attempts have been made to observe
EPR power saturation experiments corroborated the assign-
ment of the persistent Tyr379•.4b Solutions of wt-RRO and
Tyr379Phe-RRO (40 µM) were prepared in air-saturated sodium
phosphate buffer (pH 7.2, µ ) 0.1 M) containing 25% glycerol
and placed in sample tubes where they were manually mixed
with 10 equiv of H2O2 at ambient temperature. Samples were
frozen first at 193 K and then at 77 K. No EPR signal was
discernible in the sample containing the Tyr379Phe mutant,
whereas the expected signal at g ) 2.0054 was observed in the
sample containing wt protein.37 Diminution of the signal
amplitude associated with the Tyr379 · in wt-RRO was observed
at variable powers, under the conditions outlined in the
Experimental Section, indicating P1/2 ) 0.6 mW at 15 K (Figure
2). Previous studies of heme proteins have shown this P1/2 to
formation of the oxidized prosthetic group upon treatment of
(39) The competitive KIE was calculated using the following equation,
where [P-H]/[P-D] is the ratio of the integrated peak areas for the
aldehyde products at ∼1-5% conversion and [S0-D]/[S0-H] is the
initial ratio of substrate isotopologues, which should be equivalent to
the ratio of products at 100% conversion:
Dk
cat(FA) )
KM
[S0-D]
[P-H]
[P-D] [S0-H]
A similar method is outlined in ref 40. By definition, the competitive
KIE reflects the isotope effect on the second-order rate constant.
(40) Lewis, E. R.; Johansen, E.; Holman, T. R. J. Am. Chem. Soc. 1999,
121, 1395–1396.
(41) Grant, K. L.; Klinman, J. P. Bioorg. Chem. 1992, 20, 1–7.
(42) Rickert, K. W.; Klinman, J. P. Biochemistry 1999, 37, 12218–28.
9
230 J. AM. CHEM. SOC. VOL. 133, NO. 2, 2011

Catalytic Mechanism of a Fatty Acid R-(Di)oxygenase
A R T I C L E S
be characteristic of protein-derived radicals rather than the
FeIVdO(Por•+), which is also EPR-active.43
Upon oxidation of resting wt-RRO, Tyr379• is presumably
formed concomitant with the reduction of FeIVdO(Por•+) to
FeIVdO(Por). The absence of a detectable ferryl intermediate
in the stopped-flow spectrophotometric experiments suggests
that the latter process, along with the reduction of FeIVdO(Por)
back to FeIII(Por), occurs more rapidly than the initial oxidation
of FeIII-RRO by H2O2. These observations are consistent with
Tyr oxidation rate constants as large as 460 ( 50 s-1 under
similar conditions in COX-1.44 Different behavior occurs in the
structurally related heme peroxidases that do not use amino acid
radicals for catalysis. In these cases, Tyr oxidation over distances
comparable to those found in COX and RRO has been associated
45-47
with rate constants of e5 s-1
.
In light of its similarity to COX, it is surprising that RRO
does not exhibit a significant peroxidase activity.4b,24 Instead
the enzyme exhibits a catalase-like activity that may protect it
from overoxidation at high concentrations of H2O2.48 Despite
the lack of peroxidase activity in the presence of prototypical
reducing agents,49 the dioxygenase activity is affected upon
exposure of RRO to H2O2, concomitant with an increase in the
persistent EPR signal assigned to Tyr379•. Pretreatment with
H2O2 has been observed to increase initial rates of O2 consump-
tion in a manner that depends on the preparation of protein and
the identity of the fatty acid substrate. Such results are provided
in the Supporting Information, where it is shown how the
optimal concentration of H2O2 was determined for use in the
steady-state kinetic experiments.
Steady-State Kinetics. Steady-state experiments employed
16:0, 12:0, and 10:0 to probe the effects of varying the O2 and
FA concentrations upon the apparent kinetic parameters.
Because a large number of Michaelis-Menten plots were
obtained in the course of these studies, only the salient results
are presented below while the remainder are provided as Sup-
porting Information.
Figure 3. Apparent second-order rate constants for (protio) fatty acid
oxidation in response to varying the concentration of the cosubstrate.
referred to as 18:2 or linoleic acid), because of the very low
KM(FA) values which approach the detection limit of the O2
electrode.
Another objective was to determine the limiting kinetic
parameters so that the energetics of the Tyr379•-mediated H•
abstraction step could be assessed. In the present study,
experiments were most feasible with the native substrate, 16:0,
because kinetic saturation readily occured. Comparisons to 12:0
were made possible by working at high O2 and fatty acid
concentrations.
Limiting values for kcat, kcat/KM(FA), and kcat/KM(O2) were
obtained from the analysis of apparent kinetic parameters for
all substrates except 10:0. In the latter case, KM(h19-10:0) is
estimated to be in excess of 750 µM, which is above the highest
concentration that could reliably be used in the experiments,
compromising the determination of kcat. As depicted in Figure 3,
the apparent second-order rate constants ap{kcat/KM(h23-12:0)},
ap{kcat/KM(h19-10:0)} and the associated ap{kcat/KM(O2)} increased
upon raising the cosubstrate concentration.
A primary objective was to determine the kinetic mechanism
of RRO. This analysis required evaluating the apparent rate
ap
ap
ap
constants kcat, kcat/KM(FA), and kcat/KM(O2) in response to
variable cosubstrate concentrations. Use of 12:0 and 10:0
facilitated these analyses by exposing the changes in apparent
rate constants at low fatty acid and O2 concentrations. Such
measurements are more challenging with longer chain fatty
acids, such as 16:0 and 9(Z),12(Z)-octadecadienoic acid (also
Substrate Deuterium Kinetic Isotope Effects on kcat
/
KM(FA). Substrate deuterium KIEs provide the critical informa-
tion needed to test kinetic mechanisms and assess the degree
of rate limitation by H• abstraction from the fatty acid CR-H
bond. Competitive and noncompetitive measurements were
(43) (a) Yeh, H.-C.; Gerfen, G. J.; Wang, J.-S.; Tsai, A.-L.; Wang, L.-H.
Biochemistry 2009, 48, 917–928. (b) Schunemann, V.; Lendzian, F.;
Jung, C.; Contzen, J.; Barra, A.-L.; Sligar, S. G.; Trautwein, A. X.
J. Biol. Chem. 2004, 279, 10919–30.
D
performed in this study to obtain the limiting KIEs kcat and
Dkcat/KM(FA), which are defined at concentrations g6KM(O2).
The noncompetitive measurements employed disparate enzyme
concentrations, whereas the competitive measurements were
performed with the same enzyme concentration by analyzing
solutions containing a mixture of the deuterium-labeled and
unlabeled fatty acids. Therefore, the deuterium KIEs were
confirmed to be independent of the RRO concentration over a
wide range.
(44) Liu, J.; Seibold, S. A.; Rieke, C. J.; Song, I.; Cukier, R. I.; Smith,
W. L. J. Biol. Chem. 2007, 282, 18233–18244. The temperature used
for stopped-flow measurements was not specified and is therefore
assumed to be ambient.
(45) Miller, V. P.; Goodin, D. B.; Friedman, A. E.; Hartmann, C.; Ortiz
de Montellano, P. R. J. Biol. Chem. 1995, 270, 18413–18419.
(46) (a) Kimura, S.; Yamazaki, I. Arch. Biochem. Biophys. 1979, 198, 580–
588. (b) Furtmu¨ller, P. G.; Jantschko, W.; Regelsberger, G.; Jakopitsch,
C.; Arnhold, J.; Obinger, C. Biochemistry 2002, 41, 11895. (c)
Zederbauer, M.; Furtmueller, P. G.; Ganster, B.; Moguilevsky, N.;
Obinger, C. Biochem. Biophys. Res. Commun. 2007, 356, 450–6.
(47) Yeh, H.-C.; Gerfen, G. J.; Wang, J.-S.; Tsai, A.-L.; Wang, L.-H.
Biochemistry 2009, 48, 917–928.
In Figure 4, typical Michaelis-Menten plots are shown for
two protio and two deutero fatty acids that exhibit saturation
kinetics. The Dkcat values derived for 16:0 and 12:0 are 27 ( 3
and 102 ( 8, respectively. The former KIE is in good agreement
with that previously reported (31 ( 5);4b however, the non-
competitively determined Dkcat/KM(16:0) is inaccurate due to the
very small KM(d31-16:0) = 2.4 ( 1.2 µM. As a result,
(48) Zamocky, M.; Furtmueller, P. G.; Obinger, C. Antioxid. Redox
Signaling 2008, 10, 1527–1548.
(49) No peroxidase activity was observed with the following reductants:
phenol, 2-methoxyphenol, tetramethylphenylenediamine, or 2,2′-
azinobis(3-ethylbenzothiazoline-6-sulfonic acid) (ABTS). For similar
findings see ref 24a.
9
J. AM. CHEM. SOC. VOL. 133, NO. 2, 2011 231

A R T I C L E S
Mukherjee et al.
oxidation step (Vide infra). In the case of the native substrate,
D
D
similar values of kcat and kcat/KM(16:0) suggest an intrinsic
D
deuterium KIE of ∼30. Indistinguishable values of kcat and
Dkcat/KM(16:0) have also been observed at pH 10.0,4b where the
deuterium KIEs (of ∼50) as well as the rate constants are
elevated relative to those at pH 7.2.
In contrast to 16:0, the oxidation of 12:0 is characterized by
kinetic complexity. The Dkcat of 102 is significantly larger than
the Dkcat/KM(12:0) of 59. This comparison suggests that substrate
binding partially masks the KIE upon the second-order rate
constant. A similar situation may obtain for 10:0; however, the
inflated KM for this fatty acid precludes the comparison of Dkcat/
D
KM(10:0) to kcat. If fatty acid binding were partially rate
limiting, Dkcat/KM(10:0) ) 120 would actually represent a lower
limit to the intrinsic KIE.
The limiting rate constants and deuterium KIEs extracted from
Figures 3-5 are summarized in Table 1. The kcat/KM(O2) values
were determined similarly, by extrapolation of ap{kcat/KM(O2)}
to saturating levels of each protio and deutero fatty acid. These
data will be further analyzed in a forthcoming study.30 Here it
is simply noted that kcat/KM(O2) approximates 3 × 105 M-1 s-1
for h31-16:0 and h23-12:0 and that the rate constants are of
the same order of magnitude for d31-16:0 and d23-12:0. The
deuterium KIE upon kcat/KM(O2) is, therefore, much smaller
than those observed upon kcat and kcat/KM(FA). Although Dkcat
Figure 4. Michaelis-Menten plots used to determine noncompetitive
deuterium KIEs upon the oxidation of 16:0 (a) and 12:0 (b) at saturating
levels of O2 (1.3 mM).
D
and kcat/KM(FA) contain primary as well as secondary con-
tributions, the values are well outside the semiclassical limits,
even after correction for typical secondary deuterium KIEs.41,42
Therefore, the competitive and noncompetitive deuterium
KIEs determined to reflect rate-limiting CR-H homolysis are
clearly indicative of extensive hydrogen tunneling. Interest-
ingly, this radical-mediated reaction in RRO gives rise to
some of the largest deuterium KIEs that have been reported
to date.51
Isotope Exchange Experiments. Tyr379•-containing RRO
catalyzes isotope exchange of the pro-R hydrogen in the
2-position of the fatty acid with hydrogen from the solvent under
anaerobic conditions.52 Experiments were conducted by pre-
treating wt-RRO and Tyr379Phe-RRO with optimal concentra-
tions of H2O2 followed by deoxygenation under N2. The protein
was then introduced into the deoxygenated solution of deuterium-
labeled fatty acid. Control experiments were performed simul-
taneously using all components except for the protein. These
solutions were stirred under N2 at 298 K and samples collected
for analysis at time intervals between 0.5 and 8 h. For
comparison, experiments were also performed in air-saturated
solutions, and the progress was monitored until 20-50% of the
O2 was consumed. All solutions were quenched and prepared
for ESI-MS analysis following the procedure described in the
Experimental Section.
Figure 5. Variation of apparent deuterium KIEs with O2 concentration.
Measurements were performed competitively for 12:0 (a) and noncompeti-
tively for 10:0 (b).
competitive KIE measurements were used to obtain a reliable
D
estimate of kcat/KM(16:0) ) 33 ( 4. This KIE is within the
D
error limits of the average kcat ) 29 ( 4.
Competitive as well as noncompetitive measurements indi-
D
cated kcat/KM(12:0) ) 59 ( 5 and 51 ( 8, respectively. The
competitive KIE is considered more reliable than the noncom-
petitive KIE, which was determined at a single saturating O2
concentration (∼1.3 mM) versus extrapolation to O2 saturation,
as shown in Figure 5. As described in the Experimental section,
the precision of the competitive KIE measurement derives from
the resolution of GC-MS signals corresponding to the perprotio
and perdeutero aldehyde products.4b Unfortunately, the precision
is compromised for 10:0 due to significant overlap of the isotopic
nonaldehyde products. Noncompetitive measurements at variable
The results in Figure 6 clearly demonstrate that wt-RRO
catalyzes exchange of the deuterium from the CR-D bond with
protons from the H2O solvent in the absence of O2. When O2 is
present, isotope exchange into the “unreacted” fatty acid is
apparently inhibited; i.e., the deuterium does not wash out into
the solvent. Importantly, no exchange occurs when the fatty
acid is incubated with the Tyr379Phe mutant under identical
conditions. These experiments provide compelling evidence that
Tyr379• initiates fatty acid oxidation by H•/D• abstraction. The
D
O2 concentrations were, therefore, performed to obtain kcat/
KM(10:0) ) 120 ( 8.
Evidence that the observed KIEs approach intrinsic values
comes from the comparison of Dkcat to Dkcat/KM(FA). Though it
would be ideal to address this issue by comparing steady-state
KIEs to pre-steady-state KIEs determined in the absence of O2,50
such measurements have not been undertaken due to the
expected difficulty in directly observing changes in the Tyr379•
concentration because of a potentially endothermic CR-H
(51) Klinman, J. P. Chem. Phys. Lett. 2009, 471, 179–193.
(52) Similar results were observed using D2O and analyzing deuterium
incorporation into the perprotio fatty acids.
(50) Cook, P. F.; Cleland, W. W. Enzyme Kinetics and Mechanism; Garland
Science: New York, 2007.
9
232 J. AM. CHEM. SOC. VOL. 133, NO. 2, 2011

Catalytic Mechanism of a Fatty Acid R-(Di)oxygenase
A R T I C L E S
Table 1. Rate Constants and Substrate Deuterium KIEs upon Fatty Acid CR-H Oxidationa
a Values quoted are the average of at least two independent measurements. b Competitive (c) and noncompetitive (nc) KIEs were determined at O2
saturation, i.e., g6KM(O2). c From ref 4b. d Lower limit from measurements at 750 µM h31-10:0. e Estimated by varying O2 at the limit of h19- and
d19-10:0 solubility.
Figure 6. ESI-MS analysis of solvent isotope exchange into the unreacted fatty acid. All experiments were conducted with N2-saturated solutions at 298
K unless noted: (a) d19-10:0 in H2O, (b) d19-10:0 + wt-RRO in H2O, (c) d19-10:0 + wt RRO + O2 (∼258 µM) in H2O (d) R,R-d2-16:0 in H2O, (e) R,R-
d2-16:0 + wt-RRO in H2O, (f) R,R-d2-16:0 + Tyr379Phe-RRO in H2O.
inability of H2O2-activated Tyr379Phe-RRO to mediate hydro-
gen isotope exchange also argues convincingly against the
involvement of FeIVdO(Por•+) and FeIVdO(Por) as well as other
amino acid radicals that may be formed upon exposure of the
heme proteins to H2O2.
with the absence of significant solvent KIEs on any of the
steady-state kinetic parameters.4b,30
Discussion
Enzyme Activation. RRO undergoes 2e- oxidation by H2O2
at the FeIII(Por) prosthetic group to generate the Tyr379•
cofactor. FeIVdO(Por•+) is a likely intermediate; however, it
does not accumulate to a detectable level under the experimental
conditions. Therefore, it is uncertain whether FeIVdO(Por•+) is
the direct precursor to the Tyr379• intermediate, as observed in
the homologous cyclooxygenases.3 In RRO, reduction of
FeIVdO(Por) is also facile, resulting in FeIII(Por) as the only
observable species on the millisecond to second time scale. The
results are consistent with rapid ferryl formation in wt-RRO
followed by amino acid oxidation at a rate much greater than
that seen in the Tyr379Phe-RRO mutant.4b Stopped-flow spec-
As expected for stereospecific enzyme catalysis, only a single
deuterium atom is lost to the solvent and a single protium is
incorporated into the unreacted fatty acid. Thus, d18-10:0 and
R-d1-16:0 are the only detectable products from solutions
initially containing d19-10:0 and R,R-d2-16:0. That wt-RRO does
not mediate loss of the deuterium label from fatty acids under
aerobic conditions is attributed to slow exchange of H2O with
Tyr379-OD. Therefore, under normal catalytic conditions, the
deuterium label is transferred from the fatty acid via Tyr379•
into the 2(R)-hydroperoxide product. These results are consistent
9
J. AM. CHEM. SOC. VOL. 133, NO. 2, 2011 233

A R T I C L E S
Mukherjee et al.
trophotometry suggests Tyr379• formation with a H2O2-
independent rate constant of g62 s-1 in the wt enzyme.
The proposed catalytic Tyr• is conserved in RRO and in COX,
making comparisons of the two fatty acid oxidizing enzymes
of interest. Whereas COX reacts with hydroperoxides to form
multiple tyrosyl radicals and observable ferryl intermediates that
are implicated in “suicide deactivation” of the enzyme,6 RRO
reacts with hydroperoxides in a controlled manner, generating
a persistent Tyr• as the only observable intermediate during
catalysis. Also in contrast to COX, RRO can turn over for
several hours under ambient conditions, oxidizing all fatty acid
present in solution.4b,37 Regarding similarities of the two
enzymes, COX exhibits a protective peroxidase activity that is
independent of its dioxygenase activity,3 while RRO catalyzes
H2O2 disproportionation. That this reaction is independent of
fatty acid oxidation is evidenced by the observation that wt-
RRO and Tyr379Phe-RRO react with H2O2 at similar rates and
efficiencies.
At this stage, it is unclear whether the absence of peroxidase
activity in RRO is due to an intrinsic property of the enzyme
or to its specificity for an exogenous reductant that has yet to
be identified.4a,24 The absence of peroxidase activity could reflect
inaccessibility of reductants to the oxidized prosthetic group,
which appears to be buried below the protein surface to a greater
extent in RRO than in COX.3,17 The structural issues notwith-
standing, RRO’s catalase-like activity48 protects it from over-
oxidation by excess H2O2 while allowing for enzyme activa-
tion.37
RRO is isolated with a trace amount of a spectroscopically
observable radical resembling the catalytic species.4b An organic
radical also appears in the structurally related fatty acid
R-dioxygenase from Arabidopsis thaliana, but there the Tyr•
has not been implicated in catalysis.24 The presence of trace
levels of Tyr379• in RRO could explain why addition of
hydroperoxide initiators is not required for optimal dioxygenase
activity.4b A significant rate enhancement is seen upon addition
of H2O2 to 10:0 and 12:0 but not so much for 16:0. The
difference is presumably due to accumulation of the 2(R)-
hydroperoxide product, resulting in enzyme activation. The
reaction is faster with the native substrate than with the shorter
chain analogues, which exhibit lower apparent turnover rates
due to their elevated KM(FA).37
noncovalent interactions within the enzyme-substrate (Michae-
lis) complex could assist in positioning the fatty acid so that
the reactive CR-H(D) approaches the Tyr379•. The weaker fatty
acid binding affinity, which is at least partially manifest in the
increasing KM(FA), should impact the distance over which the
hydrogen must transfer. Increasing this distance is expected to
result in diminution of the rate constant and an inflation of the
deuterium KIE.
Given the degree of rate limitation by CR-H(D) homolysis,
comparisons of Dkcat to Dkcat/KM(FA) can be used to infer intrinsic
KIEs.53 That kcat ) 29 ( 4 is indistinguishable from kcat/
KM(16:0) ) 33 ( 4 indicates that the intrinsic KIE is fully
expressed at pH 7.2. A similar relationship has been reported
D
D
D
D
at pH 10.0, where kcat ) 54 ( 7 and kcat/KM(16:0) )53 (
5.4b In view of the results from the present study, the increased
intrinsic KIE for 16:0 under basic conditions is attributed to a
change in positioning of the bound fatty acid in the active site.
Partial rate limitation by fatty acid binding is indicated for the
D
medium chain length fatty acid, where kcat ) 102 ( 8 differs
D
D
significantly from kcat/KM(12:0) ) 59 ( 5. Although kcat
cannot be determined for the shortest chain substrate because
of its elevated KM(10:0), a value of >90 is estimated from the
apparent turnover rate constants at the limit of fatty acid
solubility. Despite the inability to observe saturation kinetics,
the results suggest a large degree of rate limitation by CR-H(D)
bond cleavage. If this contribution to the second-order rate
D
constant were partially masked by fatty acid binding, kcat/
KM(10:0) ) 120 would actually represent a lower limit to the
intrinsic KIE.
The closeness of Dkcat and Dkcat/KM(FA) to the intrinsic KIEs
indicates that KM(FA) should be similar for the protio and
deutero fatty acids. Furthermore, KM(FA) should be comparable
to Kd, the dissociation constant for the enzyme-fatty acid
complex.53 Assuming a simple two-step mechanism, where the
fatty acid binds reversibly prior to CR-H homolysis, Kd is
estimated to range from 4 µM (16:0) to ∼40 µM (12:0) to g750
µM (10:0). These estimates correspond to free energies of -7.3
kcal mol-1 (16:0), -5.9 kcal mol-1 (12:0), and e-4.2 kcal
mol-1 (10:0), which reflect stabilization of the Michaelis
complex with respect to the free oxidized enzyme and fatty acid.
To briefly summarize, the hydrogen transfer distance is
expected to be a primary determinant of the deuterium KIE.
This conclusion is supported by the observed trends which result
in the largest deuterium KIE being observed for the least tightly
bound fatty acid. Elongating this distance in the Michaelis
complexes is expected to diminish the probability of protium
tunneling to a lesser extent than the probability of deuterium
tunneling. There is also likely to be a significant increase in
thermal activation energy required to achieve the configuration
of heavy nuclei conducive to deuterium tunneling, as observed
in numerous studies where semiclassical descriptions of hydro-
gen transfer are insufficient.51
Fatty Acid Oxidation. Studies of RRO reveal a hyperbolic
ap
O2-dependent increase in the apparent rate constants,
k
and
cat
ap{kcat/KM(FA)}, as well as in the apparent deuterium KIEs,
ap{Dkcat} and ap{Dkcat/KM(FA)}, to limiting values for CR-H(D)
homolysis. The latter results are indicative of reversible H• and
D• tunneling at subsaturating O2 concentrations and a transition
to irreversible tunneling at saturating concentrations g6KM(O2).
The limiting kcat values for protio fatty acids are relatively
indistinguishable, varying from 9.0 (16:0) to 13.3 (12:0) to >5.5
D
(10:0) s-1. The limiting kcat values exhibit greater variation,
A modified version of quantum mechanical Marcus Theory54
has been implemented to explain a wide range of large and small
deuterium KIEs on proton-coupled electron transfer/hydrogen
atom transfer reactions. These nonadiabatic formalisms separate
the electronic coupling and Franck-Condon overlap associated
increasing from 29 (16:0) to 102 (12:0) to >90 (10:0). The
turnover rate constants for the deuterated fatty acids are more
sensitive to varying the chain length. This effect is ostensibly a
reflection of how the fatty acid is bound, relative to Tyr379•, in
the active site. Dkcat/KM(FA) also increases with decreasing chain
length, from 33 (16:0) to 58 (12:0) to 120 (10:0). Shortening
the fatty acid chain also has a dramatic effect upon kcat/KM(FA),
decreasing from 2.2 × 106 (h31-16:0) to 3.3 × 105 (h23-12:0) to
1.1 × 104 (h19-10:0) M-1 s-1 and from 6.7 × 104 (d31-16:0) to
5.7 × 103 (d23-12:0) to 9.2 × 101 (d19-10:0) M-1 s-1. In addition
to the interaction between fatty acid carboxylate and His311,
(53) Klinman, J. P.; Matthews, R. G. J. Am. Chem. Soc. 1985, 107, 1058–
60.
(54) (a) Marcus, R. A.; Sutin, N. Biochim. Biophys. Acta 1985, 811, 265–
322. (b) Kamerlin, S, C. L.; Warshel, A. Proteins: Structure, Function,
Bioinf. 2010, 78, 1339–1375. (c) Hammes-Schiffer, S. Acc. Chem.
Res. 2006, 39, 93–100.
9
234 J. AM. CHEM. SOC. VOL. 133, NO. 2, 2011

Catalytic Mechanism of a Fatty Acid R-(Di)oxygenase
A R T I C L E S
in other enzymes, such as galactose oxidase56 and possibly
soybean lipoxygenase,57 where retention of hydrogen from
the substrate gives rise to a significant deuterium KIE upon
Scheme 1. Proposed Mechanism of Anaerobic Isotope Exchange
Mediated by RRO
D
kcat/KM(O2). In the latter case, the reported kcat/KM(O2) of
∼4 may fall within the limits of error. Measurements of this
type are complicated by artifacts encountered when KM(O2)
for the deuterated substrate is greatly suppressed to offset a
very large deuterium KIE upon kcat. The proposal that the
deuterium label from the fatty acid is retained at Tyr379 in
RRO, due to slow solvent isotope exchange with active site
residues, is consistent with the absence of solvent KIEs or a
discernible change in Dkcat or Dkcat/KM(FA) when experiments
are conducted in D2O instead of H2O.
with the hydrogen isotopes’ nuclear wave functions from the
free energy barrier ∆Gq. In the future, it will be interesting to
determine whether such treatments can reproduce the rate
constants and very large deuterium KIEs reported here for Tyr•-
mediated CR-H homolysis in RRO.
Analysis of the Kinetic Mechanism. The cumulative results
obtained in this study are consistent with the minimal mecha-
nism depicted in Scheme 2, where Eox and Ered reflect the redox
•
state of Tyr379. The FA, FAO2 , and P correspond to the fatty
Effect of Molecular Oxygen. Studies of the mechanism by
which RRO reacts with molecular oxygen are due to appear in
a forthcoming paper.30 The results presented here are discussed
within the context of the kinetic mechanism. The data in Figure
3 reveal a dependence of ap{kcat/KM(O2)} upon the fatty acid
concentration as well as a dependence of ap{kcat/KM(FA)} upon
the O2 concentration in the reactions of 12:0 and 10:0. Such
results frequently point to sequential kinetic mechanisms;23,50
however, these mechanisms can be excluded for RRO because
of the results in Figure 5, which reveal increases in the apparent
deuterium KIEs, ap{Dkcat/KM(FA) and ap{Dkcat}, upon raising the
concentration of O2.55
Additional mechanistic insight is derived from the limiting
kinetic parameters associated with the oxidation of the protio
and deutero fatty acids. kcat/KM(O2), determined for both 12:0
and 16:0 isotopologues, is on the order of 105 M-1 s-1. Thus,
Dkcat/KM(O2) is quite small compared to the very large Dkcat and
Dkcat/KM(FA). These results, together with the trends in apparent
second-order rate constants with increasing cosubstrate con-
centration, are consistent with a mechanism involving CR-H
cleavage prior to O2 trapping of the R-carbon radical. At
concentrations in the physiological range, i.e., ∼50 µM O2, the
rate-limiting step occurs after O2 enters the catalytic cycle. On
the basis of the magnitude of kcat/KM(O2), this step most likely
involves an activation-controlled rather than diffusion-controlled
process, such as reduction of the putative 2(R)-peroxyl radical.
A detailed analysis of this mechanism, employing competitively
determined oxygen-18 KIEs and DFT calculations performed
on structures of potential transition states and intermediates, is
the subject of a forthcoming paper.30
acid substrate, 2(R)-peroxyl radical, and the 2-(R)(hydro)per-
oxide product. A prime symbol is used to indicate that the
protonation state of the products formed in the first irreversible
step is unclear. In this description, cleavage of the fatty acid
CR-H bond occurs reversibly prior to reversible O2 radical
trapping and rate-limiting reduction of the 2(R)-peroxyl radical
to eventually form the 2(R)-hydroperoxide product. The scheme
accounts for the observation that CR-H homolysis (k3) is
reversible at low concentrations of O2 and that reduction of the
2(R)-peroxyl radical intermediate (k7) becomes the first irrevers-
ible step.
The mechanism corresponding to steps after O2 enters the
catalytic cycle is the subject of a separate study.30 In the present
context, retention of the hydrogen from the fatty acid within
the active site is expected to result in a small deuterium KIE
upon kcat/KM(O2). Only an upper limit has been estimated
because KM(O2) for the deuterated substrate is close to the limit
of detection. Importantly, kinetic expressions derived for the
mechanism in Scheme 2 can be used to predict the hyperbolic
dependence of the apparent kcat/KM(O2) and kcat/KM(FA) on the
cosubstrate concentrations, together with the O2 depedence of
the apparent Dkcat/KM(FA). Explicit consideration of a small KIE
on k7 should have little effect.
The kinetic mechanism resembles one considered by Glickman
and Klinman in studies of soybean lipoxygenase (LOX),57 which
oxidizes linoleic acid using an FeIII-OH cofactor.58 The oxidation
of a weak bisallylic C-H bond was proposed to be irreversible
on the basis of the very large Dkcat of ∼80. Subsequent trapping of
the delocalized fatty acid radical by O2 was proposed to be the
first irreversible step and thus the origin of regio- and stereo-
specificity.59 The situation is different in RRO, where reactions of
the fatty acid and O2 are linked by a reversible CR-H homolysis
step.
Kinetic equations were derived for the observed deuterium
KIEs in RRO following Cleland’s method of net rate
constants.60 For the purposes of this analysis, k3 and k4 were
considered to be the only hydrogen isotope sensitive steps
in Scheme 2. Although there is likely to be a small KIE upon
The reversibility of the initial hydrogen transfer during RRO
catalysis is further evidenced by isotope exchange experiments
under anaerobic conditions, revealing that wt-RRO catalyzes
replacement of a single deuterium from R,R-d2-16:0 and d19-
10:0 by protium from the H2O solvent. Isotope exchange into
the unreacted fatty acid is surprisingly sensitive to the presence
of O2. Inhibition of exchange, observed when O2 is present,
suggests a slow rate of solvent isotope incorporation into the
active site Tyr379. Apparently, O2 trapping of the fatty acid-
derived radical and reduction of the putative 2(R)-peroxyl radical
intermediate out-compete the Tyr O-D exchange with H2O
illustrated in Scheme 1.
(56) Humphreys, K. J.; Mirica, L. M.; Wang, Y.; Klinman, J. P. J. Am.
Chem. Soc. 2009, 131, 4657–4663.
(57) Glickman, M. H.; Klinman, J. P. Biochemistry 1996, 35, 12882–12892.
(58) Scarrow, R. C.; Trimitsis, M. G.; Buck, C. P.; Grove, G. N.; Cowling,
R. A.; Nelson, M. J. Biochemistry 1994, 33, 15023–35.
(59) Knapp, M. J.; Klinman, J. P. Biochemistry 2003, 42, 11466–11475.
(60) (a) Cook, P. F.; Yoon, M. Y.; Hara, S.; McClure, G. D., Jr.
Biochemistry 1993, 32, 1795. (b) Cleland, W. W. Kinet. Biochem.
1975, 14, 3220–4.
That solvent isotope exchange lags behind fatty acid
R-dioxygenation in RRO resembles the behavior proposed
(55) Miller, S. M.; Klinman, J. P. Methods Enzymol. 1982, 87, 711–32.
9
J. AM. CHEM. SOC. VOL. 133, NO. 2, 2011 235

A R T I C L E S
Mukherjee et al.
Scheme 2. Proposed Minimal Mechanism of Fatty Acid Oxidation by RRO
k7 detectable under certain conditions,30 its neglect has little
impact upon the following analysis, which concentrates upon
the steps before O2 enters the catalytic cycle. The apparent
deuterium KIE on kcat/KM(FA) is expressed according to eq
3. Terms include the intrinsic deuterium KIE and equilibrium
isotope effect on CR-H homolysis, k3 and (k3/k4), respec-
tively, as well as the forward and reverse commitment factors,
Cf ) k3/k2 and Cr ) k4/k5[O2] (1 + k6/k7). It is further noted
that the influence of k7 is limited to the Cr term and that it is
moderated by the magnitude of k6.
combination of the R-carbon radical with O2, the observed kcat/
KM(O2) of ∼105 M-1 s-1 divided by the O2 association constant
(Ka ) 103 M-1) gives k7 ) 102 s-1 for rate-limiting 2(R)-peroxyl
radical reduction. In comparison, rate constants of ∼106 M-1
s-1 have been reported for bimolecular H• transfer from
R-tocopherol to lipid-derived peroxyl radicals.62b
D
D
Considering the above-mentioned rate constants, the contribu-
tion of k6/k7 to the reverse commitment factor is on the order
of 104, such that Cr ) k4/k5[O2](1 + 104). Thus, a small KIE on
D
k7 would have little effect upon Cr. That the apparent kcat/
KM(FA) approaches a limiting value independent of the O2
concentration (as shown in Figure 5) is consistent with Cr
approaching 0 as the O2 concentration is raised to saturating
levels. The result is consistent with an expression for the limiting
Dkcat/KM(FA) given by eq 4. Here Dk3 is moderated only by the
forward commitment to catalysis. Because k3 is the only term
D
k3
Dk3 + Cf +
Cr
ap
Dkcat
KM
(k )
4
(
)
FA
)
(3)
{
}
1 + Cf + Cr
The term D(k3/k4), representing the equilibrium isotope effect
on H• transfer from the fatty acid CR-H bond to Tyr379•, was
calculated from the relevant stretching frequencies obtained
using the DFT method described in the Experimental Section.
Energy-minimized structures of 4-hydroxyphenylacetate and
butanoate were used as models for Tyr379 and the longer chain
fatty acids. The equilibrium isotope effect derives from the zero-
point energy differences calculated from the normal mode
stretching frequencies: ν(O-H) ) 3528.6 cm-1, ν(O-D) )
2568.4 cm-1, ν(CR-H) ) 3008.5 cm-1, and ν(CR-D) ) 2189.9
cm-1. The difference in zero-point energy is found to be 70.8
cm-1. Neglecting entropic contributions, this energy difference
can be equated to ∆∆G° ) 0.202 kcal mol-1, which corresponds
D
Dkcat/KM(FA) has in common with kcat, defined by eq 5, the
comparison of KIEs can reveal the extent to which CR-H
cleavage is the first irreversible/rate-limiting step.
k3
Dk3 +
Dkcat
KM
k2
k3
(
)
FA )
(4)
(5)
1 +
k2
k3
Dk3 +
1 +
k7
k3
k7
Dkcat
)
D
to (k3/k4) ) 1.4 at 295 K.
The following analyses are intended to address how the
proposed mechanism in Scheme 2 accommodates reversible
CR-H cleavage in light of the steady-state KIE expressions.
First, it is noted that, in solution, O2 trapping of carbon radicals
typically occurs with rate constants that approach the diffusion
The following analysis was performed to test how the
expressions derived from the mechanism in Scheme 2 quanti-
tatively reproduce the KIEs upon 12:0 and 10:0 oxidation upon
lowering the O2 concentration. The transition from reversible
to irreversible CR-H homolysis upon raising the concentration
of O2, suggests that H• abstraction by Tyr379• ranges from
slightly endothermic to thermoneutral. Assuming ∆G° ) 0 kcal
mol-1 for this step in Scheme 2 and taking the average kcat for
the protio fatty acids to be k3 ) 10 s-1, k4 is constrained to be
limit, i.e., with values on the order of 109 M-1 s-1 61 A smaller
.
rate constant, on the order of 107 M-1 s-1, has been proposed
for linoleic acid oxidation in soybean lipoxygenase and attributed
to irreversible O2 trapping of the delocalized pentadienyl radical;
however, reversible O2 trapping followed by rate-limiting
peroxyl radical reduction could not be excluded.59 Rate constants
for O2 dissociation from related peroxyl radicals, corresponding
to k6 in Scheme 2, have been examined and correlated to C-O
bond dissociation energies.62 k6 is expected to be ∼103 s-1 for
the stable conjugated peroxyl radicals produced by lipoxygenase,
whereas k6 values of ∼106 s-1 are characteristic of nonconju-
gated peroxyl radicals with weaker C-O bonds, resembling the
2(R)-peroxyl radical intermediate formed in RRO.
10 s-1
.
First, the oxidation of 12:0 is considered under the premise
that Dkcat ) 102 is equivalent to the intrinsic deuterium KIE on
the initial hydrogen atom transfer step, Dk3. This calculation
employs eq 3 along with the computed D(k3/k4) ) 1.4, Cf )
0.76 from eq 4, and Cr ) (10-4 M-1)/[O2] derived from k4 )10
s-1, k5 ) 109 M-1 s-1, k6 ) 106 s-1, and k7 ) 102 s-1. As
D
observed in Figure 5a, the limiting kcat/KM(12:0) is predicted
In light of results from the literature,61,62 the 2(R)-peroxyl
radical intermediate in RRO is considered to form with k5 on
the order of 109 M-1 s-1 and undergo O2 dissociation with k6
on the order of 106 s-1. These estimates suggest a favorable
equilibrium constant of 103 M-1 for 2(R)-peroxyl radical
formation. Assuming a simple two-step process, involving initial
to drop from its observed maximum of 59 to an ap{Dkcat/KM(12:
0)} of 16 at 20 µM O2.
The apparent KIE for 10:0 oxidation at subsaturating O2
concentrations can also modeled under the assumption that the
measured Dkcat/KM(10:0) ) 120 is equivalent to Dk3. Using this
value, together with D(k3/k4) ) 1.4, Cf ) 0, and a somewhat
smaller Cr ) (3.0 × 10-5 M-1)/[O2] in eq 3, predicts a drop
from the maximum of 120 to an ap{Dkcat/KM(10:0)} of 45 at 20
µM O2. The experimental results in Figure 5b are reproduced,
thus validating the proposed mechanism and the derived kinetic
expressions.
(61) Maillard, B.; Ingold, K. U.; Scaiano, J. C. J. Am. Chem. Soc. 1983,
105, 5095–5099.
(62) (a) Tallman, K. A.; Roschek, B., Jr.; Porter, N. A. J. Am. Chem. Soc.
2004, 126, 9240–9247. (b) Pratt, D. A.; Mills, J. H.; Porter, N. A.
J. Am. Chem. Soc. 2003, 125, 5801–5810.
9
236 J. AM. CHEM. SOC. VOL. 133, NO. 2, 2011

Catalytic Mechanism of a Fatty Acid R-(Di)oxygenase
A R T I C L E S
Scheme 3. Proposed Tyr•-Mediated Fatty Acid R-Dioxygenase Mechanism
Thermodynamics of Tyr•-Mediated Cr-H Homolysis. The
present studies reveal that FeIII-wt-RRO is activated by H2O2.37
The form of the enzyme resulting from this reaction contains a
persistent Tyr379• and enhanced fatty acid dioxygenase
activity.4b In contrast, the Tyr379Phe mutant does not exhibit
such reactivity. Treatment with H2O2 merely oxidizes the
prosthetic group to the FeIVdO(Por•+) state, which decays to
FeIVdO(Por) and back to FeIII(Por) using two reducing equiva-
lents of unknown origin.
is formed concomitant with regeneration of the catalytic Tyr•.
Mechanisms of concerted or sequential proton-coupled
electron transfer (PCET) are possible, as depicted in Scheme
3.
Concerted PCET is thermodynamically favored over the
sequential mechanism where a high-energy intermediate is
necessarily formed, as depicted in brackets in the scheme. In
this study, hydrogen atom transfer is seen as a specific type of
concerted proton-coupled electron transfer occurring with ap-
proximately the same ∆G°. Like the initiating CR-H homolysis,
reoxidation of Tyr379 by the 2(R)-peroxyl radical is expected
to be close to thermoneutral on the basis of the fatty acid
hydroperoxide (OO-H) bond dissociation energies of 85-90
kcal mol-1 66-68 and tyrosine (O-H) bond dissociation energies
in the same range.64,65 In general, bimolecular rate constants
on the order of 106 M-1 s-1 are seen for concerted PCET/
hydrogen atom transfer reactions of hydroxylic species and oxyl
radicals in the solution phase.68 k7 of ∼102 s-1, corresponding
to the first irreversible step in Schemes 2 and 3, suggests that
the thermodynamic barrier can be as high as 14.5 kcal mol-1
as expected for a sequential PCET, where either electron or
proton transfer is partially rate-limiting.
Despite the accessibility of the high oxidizing FeIVdO(Por•+)
and FeIVdO(Por) in wt-RRO and in the Tyr379Phe mutant, these
intermediates do not mediate fatty acid CR-H oxidation. Instead
a protein-derived radical initiates this transformation. Structural
factors facilitate binding of the fatty acid in proximity to the
persistent Tyr379•, a radical that is stabilized relative to the ferryl
prosthetic group yet still maintains the thermodynamic ability
to abstract a hydrogen atom from the R-position of various fatty
acids. Typical CR-H bond dissociation energies are between
63
90 and 94 kcal mol-1
,
whereas phenolic bond dissociation
energies span a larger range, from 80 to 90 kcal mol-1
,
depending on the environment.64,65 On these grounds, the initial
H• transfer step in RRO is expected to be somewhat endothermic
or close to thermoneutral, consistent with the kinetic analyses
described above.
Conclusions
Scheme 3 depicts the detailed chemical mechanism proposed
for RRO in which reversible H• abstraction is kinetically driven
by reduction of O2 to the fatty acid 2(R)-hydroperoxide product.
As mentioned above, kcat/KM(O2) of ∼105 M-1 s-1 reflects steps
beginning with combination of O2 and the fatty acid derived
radical up to 2(R)-peroxyl radical reduction. The small magni-
tude of the observed rate constant is ostensibly more consistent
with an activation-controlled process than a diffusion-controlled
process, such as O2 trapping of a radical or product release.
Presently, it is unclear whether the 2(R)-hydroperoxide product
The kinetic mechanism of rice R-(di)oxygenase has been
examined with C10, C12, and C16 saturated fatty acids. Catalysis
is most likely initiated by H• abstraction from the fatty acid by
Tyr379•. The reversibility of this step is established by the
dependence of apparent second-order rate constants and at-
tendant deuterium KIEs upon the O2 concentration as well as
solvent isotope exchange experiments that reveal inhibition by
O2. The observation that isotope exchange is mediated by wt-
RRO but not the Tyr379Phe mutant reinforces the proposal that
Tyr379• is the initiator of catalytic fatty acid R-dioxygenation.
Another intriguing finding concerns the dependence of
deuterium KIEs upon the fatty acid chain length. Shortening
the carbon chain from 16:0 to 10:0 reduces the fatty acid binding
(63) (a) Wenthold, P. G.; Squires, R. R. J. Am. Chem. Soc. 1994, 116,
11890–7. (b) Rauk, A.; Yu, D.; Taylor, J.; Shustov, G. V.; Block,
D. A.; Armstrong, D. A. Biochemistry 1999, 37, 9089–9096.
(64) (a) Jovanovic, S. V.; Harriman, A.; Simic, M. G. J. Phys. Chem. 1986,
90, 1935–9. (b) Harriman, A. J. Phys. Chem. 1987, 91, 6102–4.
(65) (a) Mulder, P.; Korth, H.-G.; Pratt, D. A.; DiLabio, G. A.; Valgimigli,
L.; Pedulli, G. F.; Ingold, K. U. J. Phys. Chem. A 2005, 109, 2647–
2655. (b) Nara, S. J.; Valgimigli, L.; Pedulli, G. F.; Pratt, D. A. J. Am.
Chem. Soc. 2010, 132, 863–872. (c) Blomberg, M. R. A.; Siegbahn,
P. E. M.; Styring, S.; Babcock, G. T.; Aakermark, B.; Korall, P. J. Am.
Chem. Soc. 1997, 119, 8285–8292.
(66) Blanksby, S. J.; Ramond, T. M.; Davico, G. E.; Nimlos, M. R.; Kato,
S.; Bierbaum, V. M.; Lineberger, W. C.; Ellison, G. B.; Okumura,
M. J. Am. Chem. Soc. 2001, 123, 9585–9596.
(67) Das, T. N.; Dhanasekaran, T.; Alfassi, Z. B.; Neta, P. J. Phys. Chem.
A 1998, 102, 280–284.
(68) Huang, M. L.; Rauk, A. J. Phys. Org. Chem. 2004, 17, 777–786.
9
J. AM. CHEM. SOC. VOL. 133, NO. 2, 2011 237

A R T I C L E S
Mukherjee et al.
affinity and, most likely, accounts for the decrease in kcat/KM(FA)
by 2 orders of magnitude. Elongation of the hydrogen tunneling
distance may be inferred from the escalating deuterium KIEs
associated with the less tightly bound substrates. Despite
differences in the fatty acid chain length, kcat corresponding to
CR-H homolysis is relatively invariant with values of 9.0 (
similar to thermally activated “gating” or “distance sampling”,51,69
derives from the shorter distance required for deuterium transfer
because of its heavier mass and poorer wave function overlap
in the ground vibrational state than that associated with
protium.69b
In conclusion, rice R-(di)oxygenase presents an ideal model
system for investigating a growing class of heme and tyrosyl
radical-containing enzymes. The present studies have allowed
detailed mechanistic insights involving the observation of re-
versible hydrogen atom tunneling upon oxidation of a relatively
robust C-H bond by a persistent tyrosyl radical. New questions
have emerged concerning how to explain the very large deute-
rium kinetic isotope effects on such reactions and their apparent
correlation to substrate binding affinity. It will be interesting to
determine whether the nonadiabatic formalisms established to
treat concerted proton-coupled electron transfer69b can be used
to model the tyrosyl radical-mediated reactions uncovered in
this study.
1.0 (h31-16:0), 13.3 ( 0.9 (h23-12:0) and >5.5 (h19-10:0) s-1
.
Somewhat larger variations in kcat are indicated for CR-D
homolysis, where values range from 0.31 ( 0.04 (d31-16:0) to
0.13 ( 0.01 (d23-12:0) to >0.06 (d19-10:0) s-1. The results
D
suggest an increase in kcat as the fatty acid chain length and
binding affinity decrease. The origins of these effects remain
to be more thoroughly tested through studies of the temperature-
dependent intrinsic deuterium kinetic isotope effects.
D
A dependence of kcat upon the fatty acid chain length may
be expected on the basis of intrinsic barriers, where that ac-
companying C-D homolysis is greater than that accompanying
C-H homolysis.51 Such barriers can be determined from studies
of primary kinetic isotope effects as a function of temperature69
and from studies of isotopic rate constants as a function of
∆G°.70 Within the terminology of Marcus Theory,54a the
reorganization energy required to attain the activated configu-
ration (or transition state) conducive to hydrogen tunneling may
be greater for deuterium than for protium. This effect, which is
Acknowledgment. J.P.R. acknowledges grants from the Na-
tional Science Foundation (MCB-0919898) and the Department of
Energy (DE-FG02-09ER16094) as well as support from an Alfred
P. Sloan Fellowship and Camille Dreyfus Teacher-Scholar Award.
Supporting Information Available: Additional kinetic and
spectroscopic characterization. This material is available free
of charge via the Internet at http://pubs.acs.org.
(69) (a) Knapp, M. J.; Rickert, K.; Klinman, J. P. J. Am. Chem. Soc. 2002,
124, 3765–3774. (b) Hatcher, E.; Soudackov, A. V.; Hammes-Schiffer,
S. J. Am. Chem. Soc. 2004, 126, 5763–5775.
(70) Brinkley, D. W.; Roth, J. P. J. Am. Chem. Soc. 2005, 127, 15720–
1572.
JA104180V
9
238 J. AM. CHEM. SOC. VOL. 133, NO. 2, 2011