Peroxidase Self-inactivation in Prostaglandin H Synthase-1
Pretreated with Cyclooxygenase Inhibitors or Substituted with Mangano
Protoporphyrin IX*
Gang
Wu
,
Jennifer L.
Vuletich§,
Richard J.
Kulmacz
,
Yoichi
Osawa§, and
Ah-Lim
Tsai
¶
From the
Division of Hematology, Department of
Internal Medicine, University of Texas Health Science Center at
Houston, Houston, Texas 77030 and the § Department of
Pharmacology, University of Michigan Medical School,
Ann Arbor, Michigan 48109
Received for publication, January 23, 2001, and in revised form, March 13, 2001
 |
ABSTRACT |
Self-inactivation imposes an upper limit
on bioactive prostanoid synthesis by prostaglandin H synthase (PGHS).
Inactivation of PGHS peroxidase activity has been found to begin with
Intermediate II, which contains a tyrosyl radical. The structure of
this radical is altered by cyclooxygenase inhibitors, such as
indomethacin and flurbiprofen, and by replacement of heme by manganese
protoporphyrin IX (forming MnPGHS-1). Peroxidase
self-inactivation in inhibitor-treated PGHS-1 and MnPGHS-1 was
characterized by stopped-flow spectroscopic techniques and by
chromatographic and mass spectrometric analysis of the
metalloporphyrin. The rate of peroxidase inactivation was about 0.3 s
1 in inhibitor-treated PGHS-1 and much
slower in MnPGHS-1 (0.05 s
1); as with PGHS-1
itself, the peroxidase inactivation rates were independent of peroxide
concentration and structure, consistent with an inactivation process
beginning with Intermediate II. The changes in metalloporphyrin
absorbance spectra during inactivation of inhibitor-treated PGHS-1 were
similar to those observed with PGHS-1 but were rather distinct in
MnPGHS-1; the kinetics of the spectral transition from Intermediate II
to the next species were comparable to the inactivation kinetics in
each case. In contrast to the situation with PGHS-1 itself, significant
amounts of heme degradation occurred during inactivation of
inhibitor-treated PGHS-1, producing iron chlorin and heme-protein
adduct species. Structural perturbations at the peroxidase site
(MnPGHS-1) or at the cyclooxygenase site (inhibitor-treated PGHS-1)
thus can influence markedly the kinetics and the chemistry of PGHS-1
peroxidase inactivation.
 |
INTRODUCTION |
Prostaglandin H synthase
(PGHS)1 catalyzes a key step
in prostaglandin biosynthesis, the conversion of arachidonic acid to prostaglandin G2/H2 (1). PGHS undergoes
irreversible self-inactivation during catalysis, thus limiting the
overall number of turnovers (2-6). A recent mechanistic study of
PGHS-1 peroxidase self-inactivation yielded two important findings:
(a) the inactivation rate is independent of both peroxide
and enzyme concentrations; and (b) a new spectral intermediate, Intermediate III, accumulates during the
self-inactivation process after formation of Intermediate II and before
the appearance of a terminal complex (7). Peroxidase inactivation thus
does not occur by decomposition of Intermediate I (or Compound I), which contains a porphyrin radical. Instead, the branch point between peroxidase catalysis and irreversible self-inactivation is probably at Intermediate II, which contains a tyrosyl radical (8,
9). Factors that change the structure of Intermediate II might thus be
expected to modify the self-inactivation mechanism.
Pretreatment of PGHS-1 with cyclooxygenase inhibitors, such as
indomethacin, flurbiprofen, or aspirin, is known to alter the tyrosyl
radical structure in Intermediate II (10, 11), and the altered radical
fails to oxidize arachidonic acid to initiate cyclooxygenase activity
(12, 13). A second approach to altering Intermediate II structure
involves replacement of heme by mangano protoporphyrin IX (forming
MnPGHS-1). The steady-state peroxidase activity of MnPGHS-1 is only
~4% that of the iron enzyme because of very slow formation of
Intermediate I (14), but essentially full cyclooxygenase activity is
preserved (15-17). The peroxide-induced radical species in MnPGHS-1
displays EPR characteristics that are different from those of the iron
enzyme, but the radical remains capable of oxidizing arachidonate to
initiate the cyclooxygenase cycle (13). PGHS-1 treated with
cyclooxygenase inhibitors and MnPGHS-1 thus provide useful systems
to examine the relationship between reactive enzyme intermediates and
peroxidase self-inactivation.
We have evaluated the peroxidase self-inactivation kinetics in MnPGHS-1
and in PGHS-1 pretreated with indomethacin or flurbiprofen. The
inactivation mechanisms were similar to that of native PGHS-1, although
key intermediates showed different heme structures, and the overall
rates were much slower for MnPGHS-1. Thus, modification of the
Intermediate II structure has a strong influence on the self-inactivation process in PGHS-1.
 |
EXPERIMENTAL PROCEDURES |
Materials--
Hemin and MnPPIX were obtained from Porphyrin
Products, Inc. (Logan, UT). Tween 20 and
Tris(2-carboxyethyl)phosphine were from Pierce. Guaiacol,
hydrogen peroxide, and indomethacin were from Sigma. Flurbiprofen was
from Upjohn Company (Kalamazoo, MI). Arachidonic acid was from Nuchek
Preps (Elysian, MN). Peracetic acid and 3-chloroperbenzoic acid were
from Aldrich; EtOOH was from Polysciences Inc. (Warrington, PA), and
PPHP was from Cayman Chemical Co. (Ann Arbor, MI). 15-HPETE was
prepared according to Graff et al. (18) and was purified by
HPLC using a Dynamax Microsorb silicic acid column (4.6 × 250 mm,
5 µM). The other peroxides were quantified
calorimetrically using excess PGHS-1 and
N,N,N',N'-tetramethyl-p-phenylenediamine but were
not further purified. X-Omat film was from Kodak (Rochester, NY).
PGHS-1 apoenzyme was isolated from ram seminal vesicles as described
previously (19). Holoenzyme was obtained by reconstitution of PGHS-1
apoenzyme with heme and subsequent removal of unbound heme by
DEAE-cellulose (19). The concentration of PGHS-1 was determined by
absorbance at 410 nm (Am = 165 mM
1
cm
1). MnPGHS-1 was prepared by adding
equimolar Mn-PPIX to PGHS-1 apoenzyme and incubating for 30 min at room
temperature. Inhibitor-treated PGHS-1 was prepared by incubating PGHS-1
with 1-1.5 eq of either indomethacin or flurbiprofen (in ethanol) for
1-2 h on ice; the final ethanol concentration was less than 1%. The
residual cyclooxygenase activity in inhibitor-treated PGHS-1 was
usually around 4% of the control value.
Measurements of Peroxidase Inactivation during Reaction of PGHS-1
with Peroxide--
Peroxidase inactivation kinetics for MnPGHS-1 and
inhibitor-treated PGHS-1 were determined at 24 °C using a sequential
mixing protocol on a Bio-Sequential DX-18MV stopped-flow instrument
(Applied Photophysics, Leatherhead, U. K.) as described previously
(7). EtOOH or PPHP was used as the peroxidase substrate in the first stage for inhibitor-treated PGHS-1. 15-HPETE or PPHP was used as the
substrate in the first-stage for MnPGHS-1 because of the low reaction
rate with EtOOH (14). The buffer used for these reactions was 0.1 M KPi, pH 7.2, with 10% glycerol and 0.1%
Tween 20. The assay mixture in the second stage was 10 mM
guaiacol, 10 mM H2O2 for
inhibitor-treated enzyme, and 10 mM guaiacol, 5 mM peracetic acid for MnPGHS-1. Peracetic acid was
substituted for H2O2 because the former gave a
workable peroxidase rate for MnPGHS-1. Peracetic acid was used only in
the second stage to measure surviving peroxidase activity; the effects
of contaminants in the peracetic acid thus should be constant
regardless of the first-stage reaction time.
Kinetics of PGHS Heme Spectral Changes during Reaction with
Peroxide--
Heme spectral changes were followed using the DX-18MV
stopped-flow instrument with either photodiode array detection or with the kinetic scan utility. Reconstructed spectra were analyzed by
singular value decomposition to remove background noise signals and
simplify the data matrices (20) and were then subjected to the global
analysis routine. A linear, three-species, two-step model (A
B
C) was the basic mechanistic scheme used to deconvolute the spectra of
intermediate species (7). Fitting to alternative models, involving four
species and three steps (A
B
C
D) or two species and one
step (A
B), was used to check the suitability of the two-step
model. The kinetic scan spectra and deconvoluted spectra for reactions
with MnPGHS-1 were smoothed using a fast Fourier transform /reverse
fast Fourier transform treatment (14).
HPLC Analysis of Metalloporphyrin Modification--
Covalent
changes in heme or MnPPIX during peroxide-induced PGHS
self-inactivation were examined by HPLC analysis of rapid quench
samples (7). In brief, inhibitor-treated PGHS-1 or MnPGHS-1 was reacted
with the desired peroxide for a defined period at 24 °C and then
quenched by mixing with a 60% acetonitrile, 1.2% TFA solution using
an Update Instrument (Madison, WI) System 1000 chemical/freeze quench
apparatus. HPLC analysis was performed using a Waters model 600S
controller, a model 717 Plus autosampler, and a model 996 photodiode
array detector (Waters Corp., Milford, MA). Samples were injected onto
a Vydac C4 column (5 µm; 0.21 × 15 cm) equilibrated with
solvent A (0.1% TFA) at a flow rate of 0.3 ml/min. Beginning at 2 min,
a linear gradient was run to 75% solvent B (0.1% TFA in acetonitrile)
over 30 min, and then another linear gradient was run to 100% solvent
B over the next 3 min.
Electrospray Ionization Liquid Chromatography Mass Spectrometry
Analysis--
Electrospray ionization LC-MS was accomplished using a
Thermoquest LCQ/LC-MS system (Finnigan, San Jose, CA), connected to a
Hewlett Packard series 1100 degasser, binary pump, and autosampler (Hewlett Packard, Wilmington, DE). Chromatographic conditions were
optimized for heme using myoglobin as a standard. The sheath gas and
the auxiliary gas were set at 90 and 30 (arbitrary units), respectively. The spray voltage was 4.2 kV, and the capillary temperature was 200 °C. Samples were purified by HPLC described above, except that the column was equilibrated with 25% solvent B at a
flow rate of 0.3 ml/min. After 15 min, linear vamp to 75 and then to
100% solvent B were run over 20 and 5 min, respectively. The effluent
was infused directly into the mass spectrometer. Standard iron chlorin
eluted at 16 min followed by heme at 24.7 min.
Luminescent detection of covalently bound heme was conducted using a
published procedure (21). Quenched PGHS-1 samples (20 µl) were added
to 20 µl of a mixture containing 5% SDS, 20% glycerol, 0.02%
bromphenol blue, and 100 mM
Tris(2-carboxyethyl)phosphine in 125 mM Tris-HCl, pH 6.8, and incubated at room temperature for 30 min. Bands separated by
electrophoresis on 7.5% SDS-polyacrylamide gel (22) were transferred
to a nitrocellulose membrane (0.2 µm; Bio-Rad). A chemiluminescence
reagent (Super Signal, Pierce) was used as described by the
manufacturer to detect the peroxidase activity of metalloporphyrin
irreversibly bound to the protein. Films (X-Omat) were exposed for 40 min.
 |
RESULTS |
Peroxidase Self-inactivation of Indomethacin- and
Flurbiprofen-treated PGHS-1
Kinetics of Peroxidase Inactivation--
PGHS-1 pretreated with
indomethacin or flurbiprofen was first reacted with hydroperoxide for
different lengths of time, and the surviving peroxidase activity was
quantified by reaction with excess guaiacol and
H2O2, using the sequential mixing protocol described under "Experimental Procedures." The surviving peroxidase activity was found to decline with single-exponential kinetics at all
initial PPHP and EtOOH levels (data not shown). The observed first-order rate constants for decay of peroxidase activity are shown
in Fig. 1 as a function of initial
peroxide concentration. The inactivation rates for indomethacin-treated
PGHS-1 varied between 0.13 and 0.29 s
1 with
an average value of 0.19 ± 0.06 s
1
(Fig. 1, A and B). Similar results were obtained
for flurbiprofen-treated PGHS-1, but with a larger range, from 0.25 to
0.71 s
1; the average was 0.37 ± 0.14 s
1 (Fig. 1, C and D).
The data in Fig. 1 make it clear that the rate of peroxidase
inactivation in both inhibitor-treated PGHS-1 preparations was
essentially independent of peroxide concentration with either EtOOH or
PPHP. Varying the enzyme concentration from 0.1 to 2.4 µM
also led to little change in the inactivation rate. In the case of
flurbiprofen-treated PGHS-1 reacted with EtOOH, there was a trend
toward higher decay rates with increased peroxide concentration for
0.36 and 0.9 µM PGHS-1, but this trend is unlikely to be
significant because it was not observed at either higher or lower
enzyme concentrations (Fig. 1D).

View larger version (30K):
[in this window]
[in a new window]
|
Fig. 1.
Effects of peroxide concentration and
structure on the rate of peroxidase self-inactivation in
indomethacin-treated PGHS-1 (panels A and
B) and flurbiprofen-treated PGHS-1 (panels
C and D). Peroxidase inactivation
kinetics were determined using the sequential stopped-flow method
described under "Experimental Procedures" at various levels of PPHP
(panels A and C) or EtOOH (panels B
and D). Enzyme levels are indicated at the upper
left of each panel.
|
|
Heme Spectral Changes during Peroxidase Self-inactivation of
Inhibitor-treated PGHS-1--
The Soret region spectra acquired by
photodiode array detector during the reaction of indomethacin- and
flurbiprofen-treated PGHS-1 with EtOOH were similar to those observed
for native PGHS-1 (Fig. 2). Data from the
first 100 ms reflect primarily the transient formation of Intermediate
I and its subsequent conversion to Intermediate II. These spectral
changes had rate constants of at least 200 s
1, much faster than the observed
inactivation kinetics (Fig. 1). To simplify analysis of spectral
changes during peroxidase inactivation, attention was focused on
spectral changes after 100 ms. Reconstructed spectra for the 0.1-1-s
time range showed an isosbestic point at 403-404 nm and thus appeared
to reflect a single transition between two species (Fig. 2. spectra
b and c). Spectra obtained over the later 1-50-s
period showed a simple decline in amplitude with little shift in
wavelength (Fig. 2, spectrum d). Data between 0.1 and
50 s were found to fit better to a linear three-species, two-step
model than to models with more or fewer steps (not shown), and so rate
constants for a two-step mechanism were evaluated. The rate constant
for the first step was 0.50 s
1 for the native
PGHS-1 (Fig. 2A). Similar values were found for indomethacin- and flurbiprofen-treated PGHS-1, 0.28 and 0.58 s
1, respectively. These rate constants for
the initial spectral transition in the 0.1-1-s range were of the same
order as the corresponding peroxidase inactivation rate constants for
inhibitor-treated PGHS-1, 0.2 and 0.4 s
1
(Fig. 1). The fitted rate constant for the second spectral step for the
native PGHS-1 was 0.055 s
1. This is
similar to those obtained for the two inhibitor-treated samples (0.043 and 0.050 s
1) and about an order of magnitude
slower than the first step. The three spectral species deconvoluted
from data acquired after 0.1 s are shown in Fig. 2 (spectra
b, c and d in both panels). The qualitatively different nature of spectral changes in the early
phase (increasing above and decreasing below 403-404 nm) and in the
later phase (decreasing both above and below the isosbestic point at
403-404 nm) (Fig. 2) confirms that spectra b and
c reflect distinct intermediates even though their spectral
line shapes are similar. The spectral changes observed during
conversion from Intermediate II (spectrum b) to IIIa
(spectrum c) for inhibitor-treated PGHS-1 (Fig. 2,
B and C) differed only marginally in the position of the isosbestic point and the direction of amplitude changes on
either side of the isosbestic point from those observed for the
Intermediate II to III transition in native PGHS-1 (Fig.
2A). The presence of cyclooxygenase inhibitor thus did not
markedly alter the transitions in heme optical spectrum during
peroxidase self-inactivation.

View larger version (23K):
[in this window]
[in a new window]
|
Fig. 2.
Spectral changes during peroxidase
inactivation of native PGHS-1 (panel A) and PGHS-1
treated with indomethacin (panel B) or flurbiprofen
(panel C). 0.58 µM native PGHS-1,
1.58 µM indomethacin-treated PGHS-1, or 1.46 µM flurbiprofen-treated PGHS-1 was reacted with 32 µM EtOOH at 24 °C, and spectral changes were obtained
from diode array stopped-flow measurements. Spectra were first analyzed
by singular value decomposition between 0.1 and 1 s and between 1 and 50 s. Spectra of resting enzyme (a) are presented
in solid lines for comparison. Fitting of the diode array
data to a three-species, two-step linear mechanism was used to
deconvolute spectra for individual reaction intermediates, with data
from the initial 100 ms excluded from the fitting to minimize
contamination from Intermediate I. Spectra are shown for resting enzyme
(a), Intermediate II (b), Intermediate III
(c), and reaction product at 50 s (d).
Spectrum d in each panel was similar to the
spectrum of the terminal complex after a 10-min reaction.
|
|
Diode array stopped-flow data for reaction between indomethacin-treated
PGHS-1 and PPHP gave results similar to those shown for reaction with
EtOOH (data not shown), indicating that the spectral transitions
observed are independent of the peroxide structure.
Heme Structural Changes during Peroxidase Inactivation of
Inhibitor-treated PGHS-1--
Flurbiprofen- and indomethacin-treated
PGHS-1 reacted for various lengths of time with 10 eq of EtOOH were
chemically quenched and analyzed by HPLC, with heme-containing
components monitored at 400 nm and protein-containing components
monitored at 220 nm (Fig. 3). The
chromatographic profile for the flurbiprofen-PGHS-1 control (mixing
with buffer instead of EtOOH) in panel A showed a prominent
A220 peak corresponding to PGHS-1 protein (peak
3) and a prominent A400 peak (peak 2)
corresponding to native heme dissociated from the protein. For the
sample reacted with EtOOH for 20 s shown in panel B,
there was a ~50% decrease in the amount of free heme (peak 2) and an
increase in two heme products (peaks 1 and 3). Peak 1 had a shorter
retention time (~22 min) than heme itself (~24 min) and had little
220 nm absorbance, suggesting that it represented a heme derivative not
associated with the protein. Peak 3 had significant absorbance at both
400 and 220 nm, indicating that it represented a heme adduct to the
protein. Chromatographic profiles from a parallel experiment with
indomethacin-treated PGHS-1 (Fig. 3, panels C and
D) were very similar to those for flurbiprofen-treated
PGHS-1.

View larger version (26K):
[in this window]
[in a new window]
|
Fig. 3.
HPLC analysis of heme structural changes
during peroxidase self-inactivation of inhibitor-treated PGHS-1.
Flurbiprofen-treated (panels A and B) or
indomethacin-treated (panels C and D) PGHS-1 (5.8 µM in 0.1 M potassium phosphate, pH 7.2, containing 0.1% Tween 20 and 10% glycerol) was mixed at 24 °C with
an equal volume of either buffer (panels A and C)
or 60 µM EtOOH in buffer (panels B and
D) and quenched after 20 s by a solution of 60%
acetonitrile and 1.2% TFA. The final concentrations of protein,
peroxide, and acetonitrile/TFA in the quenched samples were one-third
those of the original. Heme-containing chromatographic eluents were
monitored by A400; those containing protein were
monitored by A220. Numbered peaks are
discussed under "Results." Note the difference in the
A400 axis scale in panels B and
D compared with panels A and C.
|
|
The fact that peak 1 had a shorter retention time than intact heme,
along with its Soret maximum at 398 nm (data not shown), was
reminiscent of an iron chlorin product formed during reaction of
myoglobin with H2O2 (23). Thus, a further
comparison between peak 1 and heme chlorin was conducted. Metmyoglobin
was treated with hydrogen peroxide at pH 4.7 to generate the chlorin
product (23) for analysis by HPLC. The retention time of the chlorin standard peak at about 22 min (Fig.
4B) coincided with that for peak 1 from self-inactivated flurbiprofen-treated PGHS-1 (Fig. 4A), as confirmed by the increased peak height upon
coinjection of the two samples (Fig. 4C).

View larger version (13K):
[in this window]
[in a new window]
|
Fig. 4.
Chromatographic comparison of peak 1 with
authentic chlorin. HPLC profiles are shown for
flurbiprofen-treated PGHS-1 inactivated by reaction with EtOOH for
5 s (panel A), metmyoglobin reacted with
H2O2 at pH 4.7 for 1 h (panel
B), and coinjection of equal volumes of samples used in
panels A and B (panel C). The peak
labeled with an asterisk (*) in panel B was
identified previously as an iron chlorin product (22). Details are
described under "Experimental Procedures."
|
|
LC-MS analysis was conducted to determine unambiguously the chemical
structure of the peak 1 compound(s), as done previously for myoglobin
inactivated by hydrogen peroxide (23). Parallel mass analyses of peak 1 HPLC fractions from myoglobin/H2O2 and flurbiprofen-treated PGHS-1/EtOOH samples are shown in Fig.
5. A dominant species, with an
m/z ratio of 632.2, is present in peroxide-treated samples derived from both myoglobin and
flurbiprofen-treated PGHS (panels A and B in Fig.
5). This m/z 632 ion compound was shown
previously to derive from a compound characterized by NMR as a
hydroxychlorin (chemical structure shown in Fig. 5) (23). The major
modified heme compound in peak 1 found in both flurbiprofen- and
indomethacin-treated PGHS-1 after reaction with peroxide is thus a
chlorin. There are other ions of lower intensity in the mass spectra of
the chlorins which most likely represent impurities. The present
results do not exclude the possibility that the heme product identified
from PGHS is chromatographically indistinguishable isomer of the
chlorin identified from myoglobin.

View larger version (21K):
[in this window]
[in a new window]
|
Fig. 5.
Mass spectrometric analysis of the major
dissociable heme product formed during inactivation of
flurbiprofen-treated PGHS-1 by peroxide. 100-µl aliquots of
inhibitor-treated PGHS-1 reacted with EtOOH or myoglobin reacted with
H2O2 were fractionated by HPLC. Material
corresponding to peak 1 in Fig. 4 in each case was analyzed by
electrospray ionization LC-MS as described under "Experimental
Procedures." Panel A, mass spectrum of peak 1 material
from H2O2-treated myoglobin. Panel
B, mass spectrum of peak 1 material from the peroxide-treated
flurbiprofen complex of PGHS-1.
|
|
LC-MS analysis of HPLC peak 3 from inhibitor-treated PGHS-1 was not
successful. To confirm that the material eluting in peak 3 represented
heme adducts to PGHS-1 protein, we utilized a chemiluminescence assay
developed for detection of heme adducts to myoglobin (21). This method
is specific for protein-heme adducts because native heme dissociates
from the protein during electrophoresis and migrates at the dye front,
whereas heme irreversibly bound to the protein gives a
chemiluminescence signal. As shown in Fig.
6, luminescence indicative of
protein-bound heme was found in both flurbiprofen- and
indomethacin-treated PGHS-1 reacted with EtOOH but not in the controls,
confirming that heme-protein adduct was formed during peroxidase
inactivation.

View larger version (64K):
[in this window]
[in a new window]
|
Fig. 6.
Analysis of protein-heme adduct in
self-inactivated flurbiprofen-treated PGHS-1 (panel A)
and indomethacin-treated PGHS-1 (panel B).
Samples of enzyme inactivated by reaction with peroxide were separated
using SDS-polyacrylamide gel electrophoresis and transferred to a
nitrocellulose membrane for detection of heme-protein adducts by
chemiluminescence assay as described under "Experimental
Procedures."
|
|
The intensities of peaks 1 (modified heme), 2 (native heme), and 3 (heme-protein adduct) are shown as functions of the reaction time with
peroxide for both indomethacin- and flurbiprofen-treated PGHS-1 (Fig.
7). The changes in peaks 1 and 2 approximated single-exponential kinetics, with rate constants of 0.20 s
1 and 0.40 s
1,
respectively, for indomethacin-PGHS-1 (Fig. 7A) and 0.25 s
1 and 0.072 s
1,
respectively, for flurbiprofen-PGHS-1 (Fig. 7B). The
kinetics of changes in peak 3 (400 nm) were more complex, with an
abrupt increase at the earliest time point compared with the control, but only slight further increase at later reaction times. peak 2 was
the predominant A400 species in all HPLC profiles,
accounting for at least half of the integrated absorbance. The peak 2 decay was slightly faster for indomethacin-treated PGHS-1 (0.40 s
1) than for flurbiprofen-treated PGHS-1
(0.072 s
1). The rate of peak 1 formation
appeared similar in flurbiprofen-treated PGHS-1 (0.25 s
1) and indomethacin-treated PGHS-1 (0.20 s
1). The rates for the decreases in intact
heme and the increases in chlorin were thus comparable to or somewhat
slower than the corresponding loss of peroxidase activity (Fig. 1).

View larger version (16K):
[in this window]
[in a new window]
|
Fig. 7.
Kinetics of derivatized heme formation in
inhibitor-treated PGHS-1. Integrated areas of
A400 for peak 1 (filled triangles),
peak 2 (filled circles), and peak 3 (open
triangles) are shown as a function of reaction time with EtOOH for
indomethacin-treated PGHS-1 (panel A) and
flurbiprofen-treated PGHS-1 (panel B). The values represent
averages of two data sets and are normalized to the area of the
corresponding A220 protein peak. Fitted curves
are shown as solid lines for peak 2, dashed lines
for peak 1, and dash-dot lines for peak 3.
|
|
Peroxidase Self-inactivation in PGHS-1 Reconstituted with
Mn-PPIX
Peroxidase Inactivation Kinetics--
EtOOH and
H2O2 are very poor substrates for the MnPGHS-1
peroxidase, so peroxidase inactivation in MnPGHS-1 was examined with either 15-HPETE or PPHP. Very little MnPGHS-1 peroxidase inactivation was observed in the first 10 s of incubation with PPHP; in
contrast, PGHS-1 lost 90% of its peroxidase activity over this period
(Fig. 8). The rate of MnPGHS-1 peroxidase
inactivation was 0.048 s
1, an order of
magnitude slower than the value of 0.60 s
1
determined for PGHS-1 under the same conditions (Fig. 8). Also, MnPGHS-1 retained ~40% residual activity even after prolonged reaction. The peroxidase inactivation kinetics for MnPGHS-1 using 15-HPETE were similar to those obtained with PPHP (data not shown). The
effects of peroxide concentration on MnPGHS-1 peroxidase inactivation kinetics were examined with 15-HPETE and PPHP (Fig.
9). MnPGHS-1 activity declined in an
exponential fashion in these reactions, with the rate constants ranging
between 0.033 and 0.050 s
1 for 15-HPETE
(average 0.043 ± 0.006 s
1) and 0.018 and 0.048 s
1 for PPHP (average 0.034 ± 0.011 s
1) (Fig. 9). This result indicates
that the peroxidase inactivation in MnPGHS-1, although much
slower than in PGHS-1, was still independent of peroxide concentration
and structure. There was no obvious dependence of the peroxidase
inactivation rate on the MnPGHS-1 enzyme concentration because similar
outcomes were obtained at two quite different enzyme concentrations
(Fig. 9).

View larger version (18K):
[in this window]
[in a new window]
|
Fig. 8.
Comparison of peroxidase self-inactivation
kinetics in MnPGHS-1 and PGHS-1. Decay of peroxidase activity of
1.0 µM PGHS-1 (open circles) and 1.8 µM MnPGHS-1 (filled circles) was measured
during reactions with 56 µM PPHP. Solid lines
represent single-exponential fits to the data; fitted rate constants
are given next to the curves.
|
|

View larger version (14K):
[in this window]
[in a new window]
|
Fig. 9.
Effects of peroxide concentration and
structure on MnPGHS-1 peroxidase self-inactivation kinetics.
Peroxidase inactivation kinetics were determined for two concentrations
of MnPGHS-1 at the indicated levels of 15-HPETE (panel A)
and PPHP (panel B). Enzyme levels are indicated at the
upper left of each panel.
|
|
Metalloporphyrin Spectral Changes during Peroxidase Inactivation in
MnPGHS-1--
Reconstructed spectra obtained during 0.1-10 s of
reaction between MnPGHS-1 and PPHP showed progressive increases in
absorbance at 420 nm and decreases in absorbance at ~378 and ~472
nm, with isosbestic points at 392 and 447 nm (Fig.
10A). These spectral changes
are similar to those observed previously during MnPGHS-1 peroxidase
catalysis (14-17). Kinetic spectral data covering the first 50 s
of reaction were optimally fitted by a three-species, two-step model (A
B
C), with rate constants of 0.5 s
1
and 0.05 s
1. The slower rate constant for the
second step was comparable to the average value of 0.034 s
1 observed for peroxidase inactivation in
PPHP (Fig. 9). Deconvoluted spectra for the three spectral species are
shown in Fig. 10B. The spectrum of the first species
(X) was very similar to that of resting enzyme (compare
spectra a and X in Fig. 10B). The
spectrum of the second species (Y) was similar to that of
MnPGHS-1 Intermediate II, which contains Mn4+=O and a
radical (14). The spectrum of the third species (Z) showed
the same general pattern as that of the second species, but with
decreased intensity (Fig. 10B). The spectrum obtained after
10 min of reaction was similar to that of the second species, indicating that the third species was not stable (Fig. 10B).
The very slow reversal of spectral characteristics after formation of
the third species in MnPGHS-1 was not accompanied by a recovery of
peroxidase activity (see above). Global fitting the data to a single
step model, A
B, gave a rate constant 10 times faster than the
observed decay rate. Global fitting to a four-species model, A
B
C
D, did not give a satisfactory fit either because the
resulting rate constants for the first two steps were very similar, and
the deconvoluted spectra for species B and C were essentially
indistinguishable (data not shown).

View larger version (22K):
[in this window]
[in a new window]
|
Fig. 10.
Spectral changes in MnPGHS-1 during
peroxidase inactivation. Panel A, kinetic scans
obtained from stopped-flow measurements at 5-nm increments during
reaction of 2.4 µM MnPGHS-1 with 50 µM PPHP
at 24 °C. Singular value decomposition-analyzed spectra for the
1-50 s reaction period are shown as dashed thin lines, with
the direction of intensity change indicated by arrows.
Spectra for resting enzyme (a) and the terminal complex
after 10 min (b) are presented in thick lines for
comparison. Panel B, deconvoluted spectra for intermediates
resolved by fitting kinetic scan data for 0.1-50 s to a two-step
mechanism (X Y Z).
Deconvoluted spectra are shown for intermediates X,
Y, and Z. Spectra for resting enzyme
(a) and the terminal species (b) are included for
comparison.
|
|
Spectral changes during reaction of MnPGHS-1 with 15-HPETE resembled
those seen during reaction with PPHP shown in Fig. 10 and yielded
intermediates with very similar deconvoluted spectra characteristics
(data not shown), indicating that the spectral changes during
self-inactivation were not sensitive to the peroxide structure.
Metalloporphyrin Structural Changes during Inactivation of MnPGHS-1
Peroxidase--
The chromatographic profile of control MnPGHS-1 (Fig.
11A) showed a major 400 nm
absorbing peak (peak 2), corresponding to free MnPPIX, and a major 220 nm absorbing peak (peak 3), corresponding to PGHS-1 protein. The
chromatographic profile for MnPGHS-1 reacted with PPHP for 20 s
(Fig. 11B) and had three conspicuous 400 nm peaks. The
dominant peak (peak 2) had a retention time (~ 24 min) corresponding
to free MnPPIX. The retention time of peak 3, ~28 min, coincided with
the A220 peak, indicating that peak 3 represents protein-bound porphyrin. Peak 1, with a retention time of ~22 min,
was composed of two partially overlapping peaks, 1A and 1B. The
intensity of peak 2 at the first time point was considerably lower than
the control, but the peak 3 intensity changed little with time (Fig.
11, C and D). The intensities of peaks 1A and 1B showed small, time-dependent increases, with first-order
rate constants of 0.1 - 0.2 s
1 (Fig.
11D), somewhat faster than the inactivation rate of MnPGHS-1 peroxidase (Fig. 9). The sizes of peaks 1A and 1B indicate that these
species are likely to represent only a small fraction of the
metalloporphyrin even at 60 s of reaction, when the majority of
the peroxidase is inactivated. Thus, although the peak 1A and 1B
species accumulated during the time self-inactivation occurs, their
formation is probably not the cause of self-inactivation.

View larger version (28K):
[in this window]
[in a new window]
|
Fig. 11.
HPLC analysis of MnPPIX structural changes
during self-inactivation of MnPGHS-1 peroxidase. MnPGHS-1 (6.3 µM in 0.1 M potassium phosphate, pH 7.2, containing 0.1% Tween 20 and 10% glycerol) was reacted with an equal
volume of buffer or 63 µM PPHP (in H2O) at
room temperature for various times and then quenched with a solution of
60% acetonitrile and 1.2% TFA. Final concentrations of protein,
peroxide, and quenching solvents were one-third those of the original.
Chromatographic profiles are shown for the control (panel A)
and for MnPGHS-1 reacted with peroxide for 20 s (panel
B), with heme-containing eluents monitored by
A400 and protein eluents by
A220. Panels C and D,
integrated areas for A400 chromatographic peaks
1A (filled triangles), 1B (open triangles), 2 (circles), and 3 (diamonds) are shown as a
function of time. The values are the averages of two experiments and
are normalized to the area of the unmodified protein
A220 peak. Details are described under
"Experimental Procedures." Curved lines in panel
D represent exponential fits.
|
|
 |
DISCUSSION |
Cyclooxygenase Inhibitors and Peroxidase
Inactivation--
Self-inactivation of PGHS has long been thought to
reflect collateral damage to the protein or the heme prosthetic group
in side reactions of oxidized intermediates generated during peroxidase catalysis (2, 24). Previous kinetic and spectroscopic evidence (7)
indicated that the process of peroxidase inactivation in PGHS-1
originates with Intermediate II, a species with two oxidizing equivalents: ferryl heme in the peroxidase site and a tyrosyl radical
at Tyr-385 in the cyclooxygenase site (Scheme
1) (8, 10, 25). As depicted in Fig.
12, cyclooxygenase inhibitors, such as
indomethacin and flurbiprofen, bind in the cyclooxygenase channel of
PGHS-1, putting the inhibitor about 9 Å from the heme in the
peroxidase site (26). Inhibitor binding does not significantly perturb
the heme electronic absorbance spectrum (10) or the kinetics of the
initial steps in peroxidase catalysis (14). Cyclooxygenase inhibitors
thus seem unlikely to influence directly any side reactions of the
ferryl heme oxidant during inactivation. However, bound flurbiprofen
does approach within 4 Å of Tyr-385 (26), the location of the tyrosyl
radical, which is a key oxidant in cyclooxygenase catalysis (8, 12).
Earlier studies using site-directed mutagenesis and EPR
characterization (12, 13, 25, 27) and our recent ENDOR analysis (28)
strongly indicate that the peroxide-induced tyrosyl radical of
inhibitor-treated PGHS-1 is probably not located at Tyr-385. This
"alternative" tyrosyl radical is not competent for cyclooxygenase
catalysis (12). Such repositioning of the second oxidizing equivalent in Intermediate II by inhibitor clearly has the potential to alter oxidant-driven reactions during inactivation. In other studies, cyclooxygenase inhibitors increased the resistance of PGHS-1 peroxidase activity to denaturation by heat or pH extremes (29) and blocked proteolytic attack at the distant Arg-277 residue (30-32). As depicted in Fig. 12, considerable distance separates the positions of bound inhibitor and Arg-277 in PGHS-1. This indicates that cyclooxygenase site ligands can have long range effects on the structural dynamics of
the protein. Such structural changes could conceivably impact the
self-inactivation process. It was thus not surprising that the presence
of indomethacin or flurbiprofen was observed to alter events during
peroxidase inactivation, although the alterations turned out to be
perhaps more subtle than might have been anticipated.

View larger version (72K):
[in this window]
[in a new window]
|
Fig. 12.
Ribbon diagram representation of PGHS-1
structure. Coordinates are taken from the PDB file for 1prh (26).
The heme group, Tyr-385, and the inhibitor (flurbiprofen;
flurb) are displayed as stick structures, and the
trypsin cleavage site at Arg-277 is indicated.
|
|
As with the native enzyme, the peroxidase inactivation rate in the
inhibitor-PGHS-1 complex was independent of peroxide concentration and
structure (Fig. 1). This is likely because of the much shorter lifetime
of Intermediate I (Compound I) than that of Intermediate II, as found
for native PGHS-1. This result also indicates a similar overall
inactivation process in PGHS-1 and inhibitor-treated PGHS-1, originating from an enzyme intermediate after Intermediate I, probably
Intermediate II. The rate constant for peroxidase inactivation was not
changed significantly when cyclooxygenase inhibitors were bound, going
from 0.32 ± 0.09 s
1 for PGHS-1 itself
(7) to 0.19 ± 0.06 s
1 for the
indomethacin complex, and 0.37 ± 0.14 s
1 for the flurbiprofen complex (Fig. 1). As
with native PGHS-1, the heme spectral changes during inactivation of
the inhibitor complexes were consistent with a two-step process
beginning with Intermediate II. The rate constant for the first, faster
step in the spectral changes, 0.2-1.0 s
1,
was similar to the average rates for loss of peroxidase activity, 0.19-0.37 s
1, consistent with a linkage
between conversion of Intermediate II to Intermediate III and loss of
peroxidase activity.
Even though cyclooxygenase inhibitor did not alter the general features
of peroxidase inactivation as indicated by the similar heme
spectroscopic changes during the self-inactivation process (Fig. 2),
some events at the peroxidase site during inactivation were clearly
perturbed when inhibitor was present. The most prominent changes are
the increased levels of chlorin (peak 1) and heme-protein adduct (peak
3) generated when inhibitor is present during self-inactivation of
PGHS-1 (compare Fig. 3 with Fig. 5 of Ref. 7). The results of LC-MS
analysis of peak 1 (Figs. 4 and 5) and the luminescence assay of peak 3 (Fig. 6) confirm that these species are derived from the PGHS-1 heme.
The increased formation of these heme derivatives indicates that the
redox reactions at the heme site during inactivation are quite
different when inhibitor is bound at the cyclooxygenase site. Our data
suggest that there are at least three different processes that occur
during peroxidase inactivation in inhibitor-treated PGHS-1: 1) chemical
alteration of the protein, with heme left intact (peak 2); 2) chemical
modification of heme without attachment to protein (peak 1); and 3)
covalent attachment of heme to the protein (peak 3). The first process
is predominant for the peroxide-inactivation in native PGHS-1 (7),
whereas the latter two processes are prominent with inhibitor bound.
This result implies that in inhibitor-treated PGHS-1 the oxidized heme
intermediate(s) have a higher propensity to convert to the chlorin. It
may be that the redox coupling between the porphyrin cation radical and
Tyr-385 in native PGHS-1 is tighter than that between the porphyrin
cation radical and the alternative tyrosine in inhibitor-treated
PGHS-1. Thus, the porphyrin
cation radical efficiently oxidizes
Tyr-385 to produce the tyrosyl radical in native PGHS-1, whereas in the
inhibitor-treated enzyme, the cation radical instead tends to convert
to a chlorin by intramolecular charge transfer (23).
Chromatographic analyses of reaction-inactivated PGHS samples were
monitored at 400 nm. Heme oxidation products are expected to have
significant (though perhaps reduced) absorbance at 400 nm unless the
pyrrole structure is degraded. Peroxide inactivation of heme proteins
usually does not lead to complete destruction of the pyrrole structure
(23, 33). It thus seems likely that most heme products from the PGHS
reactions were detected, provided they eluted from the HPLC column.
About half of the heme absorbance was lost during inactivation of
inhibitor-treated PGHS-1, and the kinetics of the heme loss (0.1-0.4
s
1; Fig. 7) were similar to those for loss of
peroxidase activity (0.19-0.37 s
1; Fig. 1
and Scheme 1). The increases in the chlorin and heme-protein adduct
peaks accounted for about half of the decrease in the heme peak (Fig.
7). However, the extinction coefficients of modified heme products are
known to be lower than that of intact heme (33). Thus, the sizes of
peaks 1 (chlorin) and 3 (heme-protein adduct) in Figs. 3 and 7 are
likely to underestimate the amounts of modified heme, and the two
modified heme species observed may well account for the majority of the
decrease in intact heme (peak 2). It is particularly interesting that
inhibitor binding both increases the formation of heme-protein adduct
(Figs. 3 and 7) and probably shifts the location of the tyrosyl radical
in Intermediate II from Tyr-385 to another tyrosine residue (10-13,
25, 27, 28). Characterization of the site of heme attachment in the
protein adduct formed in inhibitor-treated PGHS-1 may provide clues for identifying the alternative tyrosine and may help explain why a radical
in that position would be disposed to chemical disruption of the heme.
Metalloporphyrin Substitution and Peroxidase
Inactivation--
MnPGHS-1 peroxidase was dramatically more resistant
than PGHS-1 peroxidase to inactivation by substrate (Figs. 1, 8, and
9). This slower inactivation in MnPGHS-1 parallels the earlier
observation of a slower destruction of catalytic activity during
reconstitution of apoenzyme with MnPPIX instead of heme (34). The clear
parallels between the slower spectral changes and the slower peroxidase inactivation upon substitution of the metalloporphyrin suggest that the
spectral changes reflect, directly or indirectly, the damaging events
leading to loss of activity. MnPPIX, like heme, protects PGHS-1 from
proteolytic attack at Arg-277 (30, 35), a residue some distance from
the peroxidase site (Fig. 12), indicating that the general structural
effects caused by heme binding are mimicked by MnPPIX binding, so such
structural changes are unlikely to account for the slower inactivation
in MnPGHS-1. The slower formation of Intermediate I in MnPGHS-1 than
PGHS-1 also cannot be the reason for the slower inactivation because
the rate of inactivation was not dependent on peroxide concentration
(Fig. 9), whereas the rate of Intermediate I formation is obviously dependent on the peroxide level (14). MnPGHS-1 does have cyclooxygenase activity comparable to that of PGHS-1 (34), and the MnPGHS-1 radical
has been shown to be chemically competent for cyclooxygenase catalysis
(13), making it likely that formation of the catalytic radical at
Tyr-385 is not impaired by metalloporphyrin substitution. Differences
between PGHS-1 and MnPGHS-1 are present in both the metalloporphyrin
and the amino acid radical parts of Intermediate II (10, 13, 17). Both
the oxyferryl and the Mn4+=O moieties are strong oxidizing
agents, but the half-filled d orbitals of the latter may
reduce its chemical reactivity with neighboring oxidizable groups on
the protein. In addition, the EPR of the initial peroxide-induced
radical in MnPGHS-1 is quite distinct from the corresponding radical in
PGHS-1 (10, 13, 17), leaving no doubt that the structure of the radical
in MnPGHS-1 is greatly influenced by the character of the
metalloporphyrin. In addition, the maximal radical accumulation in
MnPGHS-1 is about one-fourth that in PGHS-1 with either 15-HPETE or
arachidonate as substrate (10, 17). With metalloporphyrin substitution thus affecting both the peroxidase site redox characteristics and the
cyclooxygenase site radical structure and radical intensity, it is
possible that the dramatically slower rate of peroxidase inactivation
in MnPGHS-1 (Figs. 1, 8 and 9) is caused by changes at the peroxidase
site or the cyclooxygenase site, or both.
MnPGHS-1 samples incubated for even a short time with peroxide lost
about half the content of intact MnPPIX, and accumulation of modified
MnPPIX species accounted for only about 10% of the A400 lost from the parent compound (Fig. 11),
indicating that MnPPIX was converted primarily to a species not
detected chromatographically. This loss of MnPPIX might be explained by
rapid bleaching by peroxide of MnPPIX dissociating from the MnPGHS-1
preparation. PGHS-1 apoenzyme exhibits lower affinity for MnPPIX than
for heme (36). Because of this, the MnPGHS-1 preparations were not
treated with DEAE-cellulose to remove loosely bound metalloporphyrin,
in contrast to the routine during reconstitution with heme. Some
loosely bound MnPPIX would be able to dissociate from the protein,
remaining chemically intact (and thus detected in peak 2) unless
destroyed by exposure to peroxide in the rapid quench apparatus. The
inactivation process in MnPGHS-1 thus appears similar to the process in
PGHS-1 in that it is accompanied by relatively little generation of
recognizable modified metalloporphyrin or metalloporphyrin-protein
adduct (Fig. 11 and Ref. 7).
Overall, there are many similarities among the processes of peroxidase
inactivation in PGHS-1, inhibitor-treated PGHS-1, and MnPGHS-1, as
summarized in Scheme 1. In each case, the loss of activity appears to
originate with the equivalent of Intermediate II, rather than with
Intermediate I. In each case, the spectral changes suggest a two-step
process, with the conversion of Intermediate II to the next spectral
species being kinetically associated with loss of activity.
Modification of the peroxidase site by metalloporphyrin substitution
leads to dramatic changes in the rates of the inactivation process and
the associated spectral changes but has little effect on production of
modified metalloporphyrin or metalloporphyrin-protein adduct. In
contrast, modification of the cyclooxygenase site with inhibitor has
little effect on the inactivation kinetics but increases the damage to
the heme and formation of heme-protein adduct. Further characterization
of the nature of the damage to the protein during peroxidase
inactivation in PGHS-1 and MnPGHS-1 and to the heme and protein in
inhibitor-treated PGHS-1 should be useful in aiding our understanding
the chemical mechanism of PGHS-1 self-inactivation.
 |
ACKNOWLEDGEMENTS |
We thank Drs. Ute M. Kent and Paul F. Hollenberg for help in LC-MS measurements.
 |
FOOTNOTES |
*
This work was supported in part by United States Public
Health Service Grants GM44911 (to A.-L. T.), GM52170 (to R. J. K.), and ES08365 (to Y. O.), Training Grant GM07767 (to J. L. V.), and
CA165954 (to Paul F. Hollenberg.), and by a Burrow Wellcome Fund
new investigator award in toxicology (to Y. O.).The costs of publication of this
article were defrayed in part by the
payment of page charges. The article
must therefore be hereby marked
"advertisement" in
accordance with 18 U.S.C. Section
1734 solely to indicate this fact.
¶
To whom correspondence should be addressed: Division of
Hematology, Dept. of Internal Medicine, University of Texas Health Science Center, P. O. Box 20708, Houston, TX 77225. Tel.:
713-500-6771; Fax: 713-500-6810; E-mail:
ah-lim.tsai@uth.tmc.edu.
Published, JBC Papers in Press, March 14, 2001, DOI 10.1074/jbc.M100628200
 |
ABBREVIATIONS |
The abbreviations used are:
PGHS, prostaglandin
H synthase;
MnPGHS-1, PGHS-1 holoenzyme containing
MnPPIX;
MnPPIX, mangano protoporphyrin IX;
EtOOH, ethyl hydrogen
peroxide;
15-HPETE, 15-hydroperoxyeicosatetraenoic acid;
PPHP, trans-5-phenyl-4-pentenyl-1-hydroperoxide;
HPLC, high
pressure liquid chromatography;
TFA, trifluoroacetic acid;
LC, liquid
chromatography;
MS, mass spectrometry.
 |
REFERENCES |
1.
|
Samuelsson, B.,
Goldyne, M.,
Granström, E.,
Hamberg, M.,
Hammarström, S.,
and Malmsten, C.
(1978)
Annu. Rev. Biochem.
47,
997-1029[CrossRef][Medline]
[Order article via Infotrieve]
|
2.
|
Smith, W. L.,
and Lands, W. E. M.
(1972)
Biochemistry
11,
3276-3285[Medline]
[Order article via Infotrieve]
|
3.
|
Ohki, S.,
Ogino, N.,
Yamamoto, S.,
and Hayaishi, O.
(1979)
J. Biol. Chem.
254,
829-836[Abstract]
|
4.
|
Hemler, M. E.,
and Lands, W. E. M.
(1980)
J. Biol. Chem.
255,
6253-6261[Abstract/Free Full Text]
|
5.
|
Marshall, P. J.,
Kulmacz, R. J.,
and Lands, W. E. M.
(1987)
J. Biol. Chem.
262,
3510-3517[Abstract/Free Full Text]
|
6.
|
Markey, C. M.,
Alward, A.,
Weller, P. E.,
and Marnett, L. J.
(1987)
J. Biol. Chem.
262,
6266-6279[Abstract/Free Full Text]
|
7.
|
Wu, G.,
Wei, C.,
Kulmacz, R. J.,
Osawa, Y.,
and Tsai, A.-L.
(1999)
J. Biol. Chem.
274,
9231-9237[Abstract/Free Full Text]
|
8.
|
Karthein, R.,
Dietz, R.,
Nastainczyk, W.,
and Ruf, H. H.
(1988)
Eur. J. Biochem.
171,
313-320[Abstract]
|
9.
|
Dietz, R.,
Nastainczyk, W.,
and Ruf, H. H.
(1988)
Eur. J. Biochem.
171,
321-328[Abstract]
|
10.
|
Kulmacz, R. J.,
Ren, Y.,
Tsai, A.-L.,
and Palmer, G.
(1990)
Biochemistry
29,
8760-8771[Medline]
[Order article via Infotrieve]
|
11.
|
Kulmacz, R. J.,
Palmer, G.,
and Tsai, A.-L.
(1991)
Mol. Pharmacol.
40,
833-837[Abstract]
|
12.
|
Tsai, A.-L.,
Kulmacz, R. J.,
and Palmer, G.
(1995)
J. Biol. Chem.
270,
10503-10508[Abstract/Free Full Text]
|
13.
|
Tsai, A.-L.,
Palmer, G.,
Xiao, G.,
Swinney, D. C.,
and Kulmacz, R. J.
(1998)
J. Biol. Chem.
273,
3888-3894[Abstract/Free Full Text]
|
14.
|
Tsai, A.-L.,
Wei, C.,
Baek, H. K.,
Kulmacz, R. J.,
and Van Wart, H. E.
(1997)
J. Biol. Chem.
272,
8885-8894[Abstract/Free Full Text]
|
15.
|
Odenwaller, R.,
Maddipati, K. R.,
and Marnett, L. J.
(1992)
J. Biol. Chem.
267,
13863-13869[Abstract/Free Full Text]
|
16.
|
Strieder, S.,
Schaible, K.,
Scherer, H.-J.,
Dietz, R.,
and Ruf, H. H.
(1992)
J. Biol. Chem.
267,
13870-13878[Abstract/Free Full Text]
|
17.
|
Kulmacz, R. J.,
Palmer, G.,
Wei, C.,
and Tsai, A.-L.
(1994)
Biochemistry
33,
5428-5439[Medline]
[Order article via Infotrieve]
|
18.
|
Graff, G.,
Anderson, L. A.,
and Jaques, L. W.
(1990)
Anal. Biochem.
188,
38-47[Medline]
[Order article via Infotrieve]
|
19.
|
Kulmacz, R. J.,
and Lands, W. E. M.
(1987)
in
Prostaglandins and Related Substances: A Practical Approach
(Benedetto, C.
, McDonald-Gibson, R. G.
, Nigam, S.
, and Slater, T. F., eds)
, pp. 209-227, IRL Press, Washington, D. C.
|
20.
|
Henry, E. R.,
and Hotrichter, J.
(1992)
Methods Enzymol.
210,
129-192
|
21.
|
Vuletich, J. L.,
and Osawa, Y.
(1998)
Anal. Biochem.
265,
375-380[CrossRef][Medline]
[Order article via Infotrieve]
|
22.
|
Laemmli, U. K.
(1970)
Nature
227,
680-685[Medline]
[Order article via Infotrieve]
|
23.
|
Sugiyama, K.,
Highet, R.,
Woods, A.,
Cotter, R. J.,
and Osawa, Y.
(1997)
Proc. Natl. Acad. Sci. U. S. A.
94,
796-801[Abstract/Free Full Text]
|
24.
|
Egan, R. W.,
Paxton, J.,
and Kuehl, F. A., Jr.
(1976)
J. Biol. Chem.
251,
7329-7335[Abstract]
|
25.
|
Tsai, A.-L.,
Hsi, L. C.,
Kulmacz, R. J.,
Palmer, G.,
and Smith, W. L.
(1994)
J. Biol. Chem.
269,
5085-5091[Abstract/Free Full Text]
|
26.
|
Picot, D.,
Loll, P. J.,
and Garavito, R. M.
(1994)
Nature
367,
243-249[CrossRef][Medline]
[Order article via Infotrieve]
|
27.
|
Tsai, A.-L.,
Palmer, G.,
and Kulmacz, R. J.
(1992)
J. Biol. Chem.
267,
17753-17759[Abstract/Free Full Text]
|
28.
|
Shi, W.,
Hoganson, C. W.,
Espe, M.,
Bender, C. J.,
Babcock, G. T.,
Palmer, G.,
Kulmacz, R. J.,
and Tsai, A.-L.
(2000)
Biochemistry
39,
4112-4121[CrossRef][Medline]
[Order article via Infotrieve]
|
29.
|
Mizuno, K.,
Yamamoto, S.,
and Lands, W. E. M.
(1982)
Prostaglandins
23,
743-757[CrossRef][Medline]
[Order article via Infotrieve]
|
30.
|
Kulmacz, R. J.,
and Lands, W. E. M.
(1982)
Biochem. Biophys. Res. Commun.
104,
758-764[Medline]
[Order article via Infotrieve]
|
31.
|
Chen, Y.-N. P.,
Bienkowski, M. J.,
and Marnett, L. J.
(1987)
J. Biol. Chem.
262,
16892-16899[Abstract/Free Full Text]
|
32.
|
Kulmacz, R. J.
(1989)
J. Biol. Chem.
264,
14136-14144[Abstract/Free Full Text]
|
33.
|
Osawa, Y.,
and Williams, M. S.
(1996)
Free Rad. Biol. Med.
21,
35-41[CrossRef][Medline]
[Order article via Infotrieve]
|
34.
|
Ogino, N.,
Ohki, S.,
Yamamoto, S.,
and Hayaishi, O.
(1978)
J. Biol. Chem.
253,
5061-5068[Abstract]
|
35.
|
Chen, Y.-N. P.,
and Marnett, L. J.
(1989)
FASEB. J.
3,
2294-2297[Abstract/Free Full Text]
|
36.
|
Kulmacz, R. J.,
and Lands, W. E. M.
(1984)
J. Biol. Chem.
259,
6358-6363[Abstract/Free Full Text]
|
Copyright © 2001 by The American Society for Biochemistry and Molecular Biology, Inc.