![[Feedback]](/icons/banner/feedbackACT.gif){kind=icon}
![[For Subscribers]](/icons/banner/subscriberACT.gif){kind=icon}
![[Archive]](/icons/banner/archiveACT.gif){kind=icon}
![[Search]](/icons/banner/searchACT.gif){kind=icon}
![[Contents]](/icons/banner/tocACT.gif){kind=icon}
|
|
|
|
|
||
Vol. 284, Issue 1, 51-60, 1998
Division of Biochemical Pharmacology, Department of Pharmaceutical Biosciences, Uppsala Biomedical Center, Uppsala University, Uppsala, Sweden
 |
Abstract |
|---|
Top
Abstract
Introduction
Procedures
Results
Discussion
References
|
|---|
Bisallylic carbons of polyunsaturated fatty acids can be hydroxylated in NADPH-dependent reactions in liver microsomes. Human recombinant cytochromes P450 and human and rat liver microsomes were assayed for bisallylic hydroxylation activity. CYP1A2, CYP2C8, CYP2C9, CYP2C19 and CYP3A4 converted [14C]linoleic acid to 14C-labeled 11-hydroxyoctadecadienoic acid (11-HODE), whereas [14C]arachidonic acid was oxygenated by CYP1A2 and CYP3A4 to 14C-labeled 13-hydroxyeicosatrienoic acid (13-HETE), 10-HETE and 7-HETE as determined by HPLC. Both substrates were also converted to many other metabolites. CYP2C9 appeared to form 12R-HETE and 13-HETE, whereas CYP2C8 formed 13-HETE, 11-HETE and 15-HETE as main monohydroxy metabolites. Fetal human liver microsomes metabolized linoleic acid to 11-HODE as a major hydroxy metabolite, whereas arachidonic acid appeared to be hydroxylated at C13, C20 and, to some extent, at C10, C19 and C7. Fetal liver microsomes mainly formed 13R-HETE, whereas adult human liver microsomes and CYP1A2 mainly formed 13S-HETE. 7,8-Benzoflavone (5 µM) and furafylline (20 µM), two inhibitors of CYP1A2, reduced the bisallylic hydroxylation activity of adult human liver microsomes. Treatment of rats with erythromycin or dexamethasone induced bisallylic hydroxylation of linoleic acid to 11-HODE in liver microsomes by 2- and 10-fold, respectively. The biosynthesis of 11-HODE by microsomes of dexamethasone-treated rats was inhibited by troleandomycin (ED50 = 1 µM) and by polyclonal antibodies against CYP3A1, suggesting that CYP3A1 could catalyze bisallylic hydroxylations in the dexamethasone-treated rat. We conclude from steric analysis of 13-HETE and the effects of CYP inhibitors on adult human liver microsomes that CYP1A2 might contribute to its bisallylic hydroxylation activity.
| |
Introduction |
|---|
|
Top
Abstract
Introduction
Procedures
Results
Discussion
References
|
|---|
Liver
and renal cortical microsomes and NADPH oxygenate polyunsaturated fatty
acids to a large number of metabolites (Capdevila et al.,
1995
; Oliw, 1994; Oliw et al., 1996). Many of the oxidizing enzymes have been identified as members of the P450 superfamily of
heme-thiolate enzymes. Hydroxylations of carbons near or at the 
end
and epoxidations of the double bonds by CYPs have attracted the largest
attention due to the biological activities of these metabolites of
arachidonic acid (Capdevila et al., 1995; McGiff, 1991;
Oliw, 1994; Oliw et al., 1996).
The CYP enzymes are important for metabolism of other endogenous
compounds like vitamin D, steroids, bile acids, prostaglandins and
leukotrienes as well as xenobiotics (Denison and Whitlock, 1995; Nelson
et al., 1996). Some CYP enzymes catalyze more or less
regiospecific oxygenations with fatty acids as substrates. The CYP4A
subfamily catalyzes 1 and 2 hydroxylation of polyunsaturated fatty acids (Capdevila et al., 1995; Oliw, 1994), whereas
CYP2E1 and other isozymes have been reported to catalyze 2 and 3
hydroxylations of arachidonic acid (Laethem et al., 1993;
Rifkind et al., 1995). Some human and rodent epoxygenases
have also been identified in the liver and kidney. For example, CYP1A2
and isoforms of CYP2B and CYP2C catalyze epoxygenation of arachidonic
acid (Capdevila et al., 1995; Daikh et al., 1994;
Oliw, 1994; Zeldin et al., 1995), but it seems likely that
many other constitutive enzymes with epoxygenase activity remain to be
identified both in liver and in extrahepatic organs.
Liver microsomes can catalyze two additional NADPH-dependent
oxygenation reactions: bisallylic hydroxylation and hydroxylation with
double bond migration (Oliw et al., 1993). Bisallylic
hydroxylation in liver microsomes of phenobarbital-treated rats
converts linoleic acid to 11-HODE and arachidonic acid to 7-HETE,
10-HETE and 13-HETE (Brash et al., 1995; Hörnsten
et al., 1996; Oliw et al., 1993). These
metabolites are acid labile and decompose to cis-trans
conjugated dienols nonenzymatically (Hamberg et al., 1992).
Hydroxylation with double bond migration, which also has been referred
to as allylic hydroxylation (Capdevila et al., 1995), leads
to enzymatic biosynthesis of cis-trans conjugated hydroxy
fatty acids. The mechanism has been investigated in microsomes of
phenobarbital-treated rats with linoleic acids stereospecifically
deuterated at C11 as substrates (Oliw et al., 1993).
Linoleic acid was converted to 9-HODE and 13-HODE with 80% to 82%
R stereochemistry by initial hydrogen abstraction at C11
with subsequent double bond migration and oxidation at C9 or C13. The
same system oxygenated arachidonic acid to 12-HETE (80% R)
in moderate amounts. Human liver microsomes can also form 13-HETE and
12R-HETE (Oliw, 1993). The mechanism of biosynthesis of 12R-HETE by
liver microsomes has not been determined, but it seems likely that it
occurs by hydroxylation with double bond migration. An alternative
route may be present in skin, in which 12R-HETE might be formed by
other mechanisms (McGiff, 1991; Oliw, 1994). 12R-HETE is a partial
agonist of the leukotriene B4 receptor and it has
therefore attracted some biological attention (McGiff, 1991).
The CYP enzymes that catalyze bisallylic hydroxylation and
hydroxylation with double bond migration have not been identified. Phenobarbital treatment of rats increased the bisallylic hydroxylation activity of liver microsomes by only a little. Treatment of rats with
DEX, which mainly induces the CYP3A subfamily in the rat liver (Pereira
and Lechner, 1995), increased the bisallylic hydroxylation activity of
liver microsomes on linoleic, arachidonic and eicosapentaenoic acids
(Hörnsten et al., 1996). Bisallylic hydroxylation
activity could also be increased by treatment of rats with acetone plus starvation, with imidazole and with 
-naphthoflavone, indicating that
other enzymes than the CYP3A subfamily could have this enzyme activity
in the rat.
The CYP enzymes with bisallylic hydroxylation activity can be
conveniently identified with the help of recombinant enzymes. Bisallylic hydroxylation activity of human recombinant CYP has not been
investigated, but a recent report addresses this question indirectly.
Rifkind et al. (1995) found that CYP1A2, CYP2C8 and CYP2C9
converted arachidonic acid to relatively large amounts of
cis-trans conjugated HETEs. In this study, it is likely that bisallylic hydroxy metabolites may have decomposed to
cis-trans conjugated HETEs during the acidic extractive
isolation. These investigators made the interesting observation that
CYP2C9 metabolized arachidonic acid to 12-HETE, but the stereochemistry
of 12-HETE was not examined.
The first objective of the present study was to investigate whether a series of recombinant human CYP enzymes could form bisallylic hydroxy metabolites of linoleic and arachidonic acids. Linoleic acid is a convenient substrate for this purpose, whereas arachidonic acid can be oxygenated to a much more complex mixture of metabolites. Biosynthesis of 12-HETE by CYP2C9 and its allelic variant R144C was also investigated. The second objective was to use the results from the screening of the CYP enzymes to study their contribution to the bisallylic hydroxylation activity of adult human liver microsomes. This was done by the aid of enzyme inhibitors and by chiral analysis of the main product, 13-HETE. Thus, we examined the effects of some enzyme-specific inhibitors on bisallylic hydroxylation activity of adult human liver microsomes and compared the chirality of 13-HETE formed by CYP and by liver microsomes. We also found that human fetal liver microsomes catalyzed bisallylic hydroxylation of fatty acids. Finally, the effects of gender, starvation and treatment with ERY and DEX on rat liver bisallylic hydroxylation activity were examined.
| |
Experimental Procedures |
|---|
|
Top
Abstract
Introduction
Procedures
Results
Discussion
References
|
|---|
Materials.
18:2n-6 (99%), 20:4n-6 (99%), DEX,
phenylmethylsulfonyl fluoride, testosterone, 7,8-benzoflavone, ERY and
TAO were from Sigma Chemical (St. Louis, MO).
[1-14C]18:2n-6,
[1-14C]20:4n-6 (55 Ci/mol) and
[4-14C]testosterone (56 Ci/mol) were from
Amersham (Amersham, UK). The radiolabeled fatty acids were usually
diluted with unlabeled fatty acids to a specific activity of 2.3 Ci/mol. 6-Hydroxytestosterone was from Steraloids (Wilton, NH).
Cartridges for extraction (SepPak/C18) were from
Waters (Milford, MA). 13S-HODE and 15S-HETE were obtained through
biosynthesis (Oliw et al., 1993). 9R,S-HETE, 11R,S-HETE, 12S-HETE and 20-HETE were from Cayman Chemical (Reading, UK). 12R-HETE
methyl ester was a gift of Dr. S.-E. Dahlén (Karolinska Institutet, Stockholm, Sweden). 18R-HETE and 19R-HETE were obtained by
biosynthesis with mycelia of Gäumannomycis graminis (Oliw, 1989), whereas 17-HETE was a gift from Dr. J. R. Falck (University of Texas Southwestern Medical Center, Dallas, TX). Sulfaphenazole and
furafylline were purchased from Ultrafine Chemicals (Manchester, UK).
Microsomes of recombinant human P450 expressed in lymphoblastoid cells
(CYP1A2, CYP2A6, CYP2B6, CYP2D6 and CYP2E1) and insect cells [CYP3A4,
CYP2C8, CYP2C9 (with Arg144 or Cys144) and CYP2C19] were purchased
from Gentest (Woburn, MA). Three samples of human livers were obtained
from the liver bank of Huddinge Hospital (courtesy of Dr. J. Säwe
and Dr. M.-L. Dahl): HL37 was from a 59-year-old woman who died from
head trauma after treatment with mannitol, ekvacillin and atropine;
HL42 from a 31-year-old man who died after subarachnoidal bleeding
after treatment with theophylline and phenytoin; and HL46 was from an
adult male with an unknown case history. Two samples of human fetal
livers (before week 24 of gestation) were kindly provided by Dr. A. Rane (Akademiska Sjukhuset, Uppsala, Sweden). Polyclonal rabbit
antibodies against CYP3A1 were purchased from XenoTech (Kansas City,
KS) and were used for inhibition of CYP3A1 as recommended by the
supplier.
Incubation with microsomes of recombinant human P450 expressed in
insect or lymphoblastoid cells.
From 25 to 50 pmol of CYP3A4,
CYP2C8, CYP2C9 or CYP2C19 in 0.5 to 1 ml of incubation buffer (0.05 mM
Tris-HCl, pH 7.4, containing 1 mM EDTA, 1 mM EGTA, 0.1 mM
phenylmethylsulfonyl fluoride and 0.5 mM dithiothreitol) were incubated
with 14C-labeled fatty acids (4-5 µM; 56 Ci/mol) and NADPH (1 mM) for 30 min at 37°C. P450 expressed in
lymphoblastoid cells were incubated for 2 hr at 37°C (
30-65 pmol
of CYP1A2 and CYP2A6, 45-90 pmol of CYP2E1, 65-130 pmol of CYP2D6 and
50-100 pmol of CYP2B6) with 15 to 20 µM of substrate. Microsomes of
insect and lymphoblastoid cells without expressed P450 were incubated
in the same way. Incubations were terminated by four volumes of
ethanol. An internal standard (13-HODE or 15-HETE) was added to check
for recovery on HPLC and used for estimation of biosynthesis of
metabolites during the incubation time period and for comparison
between enzymes (tables 1 and
2). The standards also verified the
retention times on HPLC.
|
|
Incubation with human and rat liver microsomes.
Human liver
microsomes were prepared by differential centrifugation as described
and stored at 
80°C until use (Oliw, 1993). Rat liver microsomes
were prepared separately in the same way from each animal. Fischer 344 rats (170-190 g; Mollegaard, Skensved, Denmark) were treated with DEX
(150 mg/kg for 4 days; three males, one female), with ERY (734 mg/kg
p.o. for 5 days, n = 3) or with solvent (1% Tween-20;
three males, one female) as described previously (Daujat et
al., 1991). Three male rats were also starved for 48 hr.
1 mg/ml and incubated
with 1.0 mM NADPH and the 14C-labeled fatty acids
(usually 0.1 mM, 2.3 Ci/mmol) for 30 min at 37°C with constant
shaking. Incubations were terminated as above. The effects of
mechanism-based inhibitors (furafylline and TAO; both dissolved in
methanol) were assessed after preincubation with the drug and NADPH for
15 min at 37°C. [14C]18:2 or
[14C]20:4 was then added to the incubation,
which was terminated after 30 min as described above. Sulfaphenazole
and 7,8-benzoflavone were added directly to the incubations without
preincubations. Control incubations and incubations with drugs were
performed in duplicate.
Purification of metabolites.
After extractive isolation
without acidification on a cartridge with octadecasilane silica
(SepPak/C18), the arachidonic acid metabolites
were first separated by RP-HPLC into three groups of metabolites:
triols and diols, HETEs and epoxides. HETEs typically eluted between 20 and 30 min. These fractions with HETEs were then pooled and analyzed by
SP-HPLC. 13-HETE and 19-HETE have similar retention times on RP-HPLC
(Brash et al., 1995), and in some experiments, unlabeled
19R-HETE (25-50 µg; monitored by UV at 207 nm) was added to localize
the 13-HETE metabolite in the chromatogram. Unlabeled 12-HETE was added
to incubations with CYP2C9 to facilitate identification of radiolabeled
12R-HETE by RP-HPLC, SP-HPLC and chiral HPLC. The metabolites of
linoleic acid were extracted as above and separated by RP-HPLC. The
11-HODE peak was further analyzed by SP-HPLC and GC-MS in some
experiments.
HPLC.
The equipment for HPLC has been described
(Hörnsten et al., 1996). In short, the columns
contained octadecasilane silica (5-µm, 250 × 4.6 mm) for
RP-HPLC and silica (5-µm, 250 × 4.6 mm) for SP-HPLC. The
RP-HPLC column was eluted with methanol/water/acetic acid (75:25:0.01)
at 1 ml/min, and the SP-HPLC column was eluted with
hexane/isopropanol/acetic acid (98:2:0.1) at 0.5 to 1 ml/min. The
elution order of HETEs on SP-HPLC was 12-HETE, 15-HETE, 11-HETE, 13-HETE, 18-HETE, 10-HETE, 19-HETE, 7-HETE and 20-HETE.
)-N-3,5-dinitrobenzoyl-
-phenylglycine (ionically
bonded; 5 µm, 250 × 4.6 mm; Oliw et al., 1993) and
Chiralcel OB-H for analysis of 12-HETE and Chiralcel OD-H (5 µm,
250 × 4.6 mm) for analysis of 13-HETE (Brash et al.,
1995). These chiral columns were eluted with 0.5%, 2% and 2%
isopropanol in hexane, respectively, at 0.5 to 1.0 ml/min. The HPLC
columns were connected to a photo diode array detector (Waters 991) and
to a radioactivity detector for HPLC (Radiomatic Flo-one 150TR,
Packard) using Ultima Flo AP (Packard) as scintillator. The internal
standard (13-HODE, 15-HETE) was monitored by UV absorption at 235 nm
during HPLC. In addition, large amounts of the bisallylic hydroxy
metabolites could be monitored by their end absorption at 205 to 220 nm
(Brash et al., 1995).
Derivatizations and GC-MS analysis.
Trimethylsilyl ethers
were prepared by treatment with bis(trimethylsilyl)trifluoroacetamide
and pyridine and methyl esters with diazomethane (Oliw et
al., 1993). A capillary GC (Varian 3100) with a nonpolar column
(30-m; DB-5, J&W Scientific; film, 0.25 µm; diameter, 0.25 mm;
carrier gas He, 15 p.s.i.) and an ion trap mass spectrometer
(ITS40, Finnigan MAT) were used as described (Oliw et al.,
1993). C-values were determined by the retention times of fatty acid
methyl esters (18:0, 20:0, 22:0 and 24:0). Steric analysis of 11-HODE
was performed after partial hydrogenation and ozonolysis of the
()-menthoxycarbonyl derivative as described previously (Hamberg
et al., 1992).
Other analyses.
Protein was determined as described by
Bradford (1976) using 
-globulin as a standard. Testosterone
6-hydroxylase activity was assayed according to directions of the
supplier of [14C]testosterone (Amersham) using
TLC on silica plates (20 × 20 cm; Whatman LK6D, Maidsone, UK;
eluent CH2Cl2/acetone, 4:1)
for separation. Radioactivity of the TLC plates was determined by a TLC
scanner (Berthold Dünnshichtsscanner II; Kebo, Stockholm, Sweden). The Rf values were 0.5 and
0.25 for testosterone and 6-hydroxytestosterone, respectively.
O-Dealkylation of ethoxyresorufin and 4
-hydroxylation of
diclofenac sodium by human liver microsomes were assayed as described
previously (Burke et al., 1985; Leemann et al.,
1992).
| |
Results |
|---|
|
Top
Abstract
Introduction
Procedures
Results
Discussion
References
|
|---|
Recombinant human CYP and linoleic acid.
RP-HPLC of
metabolites of [14C]18:2n-6, which were formed
by NADPH and CYP1A2 (60 pmol, 2 hr) expressed in lymphoblastoid cells and by CYP2C8 (25 pmol, 0.5 hr), CYP2C9 (50 pmol, 0.5 hr) and CYP3A4
(25 pmol, 0.5 hr) expressed in insect cells, are shown in figure
1, A-D. All four enzymes formed a
prominent metabolite with an elution time of 17 to 18 min. This
metabolite had the same retention time as
[14C]11-HODE formed by liver microsomes of
DEX-treated rats (Hörnsten et al., 1996). CYP1A2 and
CYP3A4 formed this metabolite, 11-HODE, as a major product. A fifth
enzyme, CYP2C19, also formed some 11-HODE (table 1).
|
21 min), 18-HODE
(23 min) and monoepoxides of 18:2n-6 (32 min) in all
chromatograms (fig. 1) and in experiments with CYP2C19. CYP1A2 appeared
to form larger amounts of 17-HODE (fig. 1A) than the other enzymes.
Epoxides were partly converted to diols, which eluted after 13 min, by microsomes of insect cells (e.g., fig. 1, C and D) but not
by microsomes of lymphoblastoid cells (fig. 1A). Finally, CYP2C9 and
its allelic variant R144C metabolized 18:2n-6 to the same profile of
metabolites.
The total biosynthesis of metabolites of 18:2n-6 by the enzymes with
bisallylic hydroxylation activity and the relative formation of major
metabolites are summarized in table 1. The main products formed by all
the enzymes were epoxides/diols, 1 and 2 hydroxy metabolites,
which is in agreement with previous studies (Oliw, 1994). The
bisallylic metabolite 11-HODE is nevertheless a relatively prominent
metabolite, particularly of CYP1A2 and CYP3A4.
Control microsomes of insect and lymphoblastoid cells metabolized
[14C]18:2n-6 insignificantly and no
radioactivity eluted between 15 and 20 min (i.e., in the
part of the chromatogram in which [14C]11-HODE
eluted). CYP2A6, CYP2B6, CYP2D6 and CYP2E1 did not form significant
amount of [14C]11-HODE as judged from RP-HPLC.
Attempts to obtain 11-HODE in sufficient amounts for GC-MS analysis by
using larger amounts of recombinant enzyme and higher concentrations of
substrate were unsuccessful. Incubation with 0.1 mM 18:2n-6 (or
20:4n-6) appeared to lead to enzyme inactivation, in agreement with a
previous report (Imaoka et al., 1992).
Recombinant human CYP and arachidonic acid. CYP1A2
metabolized [14C]20:4n-6 to many polar products
as judged from RP-HPLC (fig. 2A). The
main metabolites were HETEs and epoxides with little hydrolysis to
diols (not shown). When the HETEs were analyzed by SP-HPLC, several
major metabolites were separated, as shown in figure 2B. The major
metabolites had the same retention times as 13-HETE, 19-HETE, 7-HETE
and 10-HETE, with only little radioactivity eluting at the retention
time of 20-HETE. We noted a backside shoulder on the 13-HETE peak. It
has been reported that 16-HETE and 17-HETE have a slightly longer
retention time than 13-HETE (see Brash et al., 1995), and
these metabolites might contribute (Falck et al., 1990).
18-HETE also elutes after 13-HETE.
|
1 and 2 hydroxy metabolites was insignificant. As discussed
above, we could not obtain sufficient material for GC-MS analysis by
recombinant enzymes. To correct for day-to-day variations in retention
times of SP-HPLC, we used UV monitoring of internal standards (13-HODE,
15-HETE). We also analyzed pooled monohydroxy metabolites of
[14C]20:4n-6 formed by rat liver microsomes of
DEX treated rats for comparison (fig. 2D). The major metabolites of the
peaks in the chromatogram (marked 7-, 10-, 13-, 19- and 20-HETE in fig.
2D) were previously identified by GC-MS (Hörnsten et
al., 1996).
CYP2C9 formed two major peaks on SP-HPLC (fig.
3A), which eluted after 7 and 10
min, respectively, as well as many minor metabolites. The first
metabolite had the same retention time as 12-HETE on SP-HPLC (on
coinjection; fig. 3B), and it eluted mainly with the 12R-HETE
stereoisomer on chiral HPLC (fig. 3C). The second
14C-labeled metabolite had the same retention
time as 13-HETE.2 The two
metabolites were therefore tentatively identified as 12R-HETE and
13-HETE. In addition, radiolabeled material with the same elution time
as 15-HETE (directly after 12-HETE), 11-HETE (directly before 13-HETE),
10-HETE (after 13-HETE) and 19-HETE eluted during SP-HPLC, but there
was negligible biosynthesis of 7-HETE and 20-HETE. The allelic variant
R144C of CYP2C9 yielded similar results. CYP2C8 converted
[14C]20:4n-6 to main hydroxy metabolites with
the same retention times as 13-HETE, 15-HETE, 11-HETE, 10-HETE and
19-HETE on SP-HPLC (data not shown).
|
|
; Ono et al., 1996;
Schwab et al., 1988; Yun et al., 1992). A low
concentration of 7,8-benzoflavone (5 µM) was found to reduced the
biosynthesis of 11-HODE (HL42 and HL46), as discussed below.
|
Effect of CYP inhibitors on adult human liver microsomes. The screening of CYP enzymes indicated that CYP3A4, the CYP2C subfamily and CYP1A2 might contribute to the bisallylic hydroxylation activity of human microsomes. Specific inhibitors of these enzymes were studied.
TAO (10-50 µM) did not affect the bisallylic hydroxylation of three different adult human liver microsomes. TAO is a mechanism-based inhibitor of the CYP3A subfamily, including CYP3A4 (Newton et al., 1995; Ono et al., 1996). To confirm that
concentration of TAO was sufficient, we confirmed that 10 µM TAO
reduced 6-hydroxylation of testosterone in these human microsomes by
>90% (see Newton et al., 1995). In addition, TAO (1-10
µM) inhibited the biosynthesis of 11-HODE from 18:2n-6 by liver
microsomes of DEX-treated rats (see below).
We also investigated the effects of sulfaphenazole (50 µM), an
inhibitor of the CYP2C subfamily (except CYP2C19), and two inhibitors
of CYP1A2, furafylline (20 µM), a mechanism-based inhibitor and
7,8-benzoflavone (Crespi et al., 1997; Newton et
al., 1995; Ono et al., 1996), on the bisallylic
hydroxylation of 18:2n-6 by three different samples of adult human
liver microsomes (HL37, HL42, HL46).
Sulfaphenazole did not inhibit the bisallylic hydroxylation of linoleic
and arachidonic acids, but there appeared to be some reduction in the
biosynthesis of epoxides/diols activity, in agreement with studies by
Rifkind et al. (1995). We confirmed that drug concentration
was sufficient. Sulfaphenazole reduced by >90% the 4-hydroxylation
of diclofenac (HL42, HL46).
Furafylline (20 µM; three livers) and 7,8-benzoflavone (5 µM: HL42,
HL46) reduced the bisallylic hydroxylation activity on 18:n-6 by 60%
and 40% to 50%, respectively. At this concentration both drugs
inhibited the O-dealkylation of ethoxyresorufin by >90% (HL42, HL46).
Effect of CYP inhibitors on fetal liver microsomes.
Fetal
liver microsomes contain CYP3A7. TAO is also an inhibitor of CYP3A7
(Hashimoto et al., 1995), but TAO (10 µM) did not reduce
the biosynthesis of 11-HODE by fetal liver microsomes. Furafylline (20 µM) did not inhibit the biosynthesis of 11-HODE by fetal liver
microsomes.
Steric analysis of 13-HETE.
The stereochemistry of 13-HETE
formed by human liver microsomes and some recombinant CYP was
determined by chiral HPLC (Chiralcel OD-H) as described previously
(Brash et al., 1995). As a standard, we used 13-HETE formed
by liver microsomes of phenobarbital-treated rats, which was found to
contain 60% 13R-HETE (see Brash et al., 1995).
|
4 min
before 13S-HETE methyl ester. Analysis of 13-HETE from incubations with
CYP3A4 and from microsomes of DEX-treated rats did not show significant
amounts of this material.
Liver microsomes of Fischer rats.
11-HODE was formed as a
minor metabolite of 18:2n-6 by liver microsomes of untreated rats.
Starvation for 48 hr did not appear to increase its biosynthesis
(n = 3), whereas treatment with starvation and acetone
has been found to do so previously (Hörnsten et al., 1996). ERY treatment appeared to induce the biosynthesis of 11-HODE 2-fold, and DEX treatment induced biosynthesis by 10-fold. We confirmed that both treatments strongly (5-20-fold) induced the 6-hydroxylation of testosterone in liver. We also found that DEX
treatment increased bisallylic hydroxylation activity in liver microsomes of both sexes. As expected, there appeared to be an increased bisallylic hydroxylation of 20:4n-6 after treatment with DEX:
2-fold for 13-HETE and 10-HETE in some experiments and 4-fold for
7-HETE (data not shown). 13-HETE prepared by liver microsomes of
DEX-treated rats was found to be racemic by chiral HPLC.
). CYP3A1 was therefore an obvious candidate with bisallylic hydroxylation activity. This possibility was assessed by
TAO, an inhibitor of the CYP3A subfamily (Newton et al.,
1995; Ono et al., 1996), and by inhibitory antibodies
against CYP3A1. TAO dose-dependently reduced the biosynthesis of
11-HODE with an ED50 of 1 µM (fig.
7). TAO caused 85% inhibition at 10 µM. TAO (10 µM) also appeared to reduce the bisallylic
hydroxylation of 20:4n-6 to 13-HETE and 7-HETE by >50%. Finally,
polyclonal antibodies against CYP3A1 (1 mg/ml) were found to inhibit
almost completely the bisallylic hydroxylation of
[14C]18:2n-6 by liver microsomes of DEX-treated
rats, whereas the control incubation, which contained IgG (1 mg/ml)
instead, formed [14C]11-HODE (data not shown).
|
GC-MS analysis and steric analysis of 11-HODE.
11-HODE was
identified from an incubation with fetal liver microsomes with 18:2n-6
and NADPH. The C-value and the mass spectrum were as reported
previously (Hamberg et al., 1992). The stereochemistry of
11-HODE formed by liver microsomes of DEX treated rats was found to be
nearly racemic (58% R). 11-HODE is also formed by liver
microsomes of phenobarbital-treated rats, and in this case the result
was almost identical, 59% 11R-HODE (Oliw et al., 1993).
| |
Discussion |
|---|
|
Top
Abstract
Introduction
Procedures
Results
Discussion
References
|
|---|
The present study focuses on bisallylic hydroxylation and
hydroxylation with double bond migration of polyunsaturated fatty acids. Polyunsaturated fatty acids are also metabolized to epoxides and
to 1 and 2 hydroxy metabolites as major products (Capdevila et al., 1995; Oliw, 1994; Oliw et al., 1996).
These reactions can be catalyzed by many different CYP enzymes and have
been studied extensively, whereas the present work is the first
systematic study of human recombinant CYP with bisallylic hydroxylase
activity.
Human recombinant enzymes. We report three new findings. Five of 10 human recombinant CYP, viz., CYP1A2, CYP3A4, CYP2C8, CYP2C19 and two isoforms of CYP2C9 (Arg144 and Cys144), were found to metabolize 18:2n-6 to 11-HODE and 20:4n-6 to 13-HETE. CYP1A2 and CYP3A4 also hydroxylated the two other bisallylic carbons of 20:4n-6, although 13-HETE was clearly the main bisallylic hydroxy metabolite. CYP2C9 and CYP2C8 yielded a different pattern. The former also oxidized 20:4n-6 to 12R-HETE as a main metabolite and to small amounts of 15-HETE, 11-HETE and 10-HETE. The latter formed 13-HETE along with 15-HETE, 11-HETE and small amounts of 10-HETE. CYP1A2 and CYP3A4 can thus catalyze bisallylic hydroxylations, whereas CYP2C8 and CYP2C9 also can form cis-trans conjugated HETEs, presumably by hydroxylation with double bond migration. Some of the main hydroxy metabolites formed by these five enzymes are summarized in figure 8.
|
60% of the CYP enzymes of
human liver, but there are large interindividual variations
(Guengerich, 1995). In view of the bisallylic hydroxylation activity of
the recombinant enzymes discussed above, we expected that specific
inhibitors of CYP1A2, the CYP2C subfamily and CYP3A4 would reduce the
bisallylic hydroxylation activity of human liver microsomes. Only
inhibitors of CYP1A2 (furafylline, 7,8-benzoflavone) appeared to do so,
whereas inhibitors of the CYP3A subfamily (TAO) and the CYP2C subfamily
(sulfaphenazole) were without effects.
Furafylline and low concentrations of 7,8-benzoflavone are known to
inhibit CYP1A2 with selectivity (Newton et al., 1995; Ono
et al., 1996). Both drugs reduced the bisallylic
hydroxylation of 18:2n-6 by 60% and 40%, respectively. In addition,
both CYP1A2 and adult human liver microsomes mainly formed the 13S
stereoisomer of 13-HETE. The selective inhibitors and the steric
analysis suggested that CYP1A2 could contribute to the bisallylic
hydroxylation activity of adult human liver microsomes in
vitro, but other CYP enzymes might also contribute.
Our third finding is that fetal liver, microsomes had a prominent
bisallylic hydroxylation activity on 18:2n-6 and 20:4n-6 in comparison
with the epoxygenation and hydroxylation of these fatty acids.
Interestingly, the chirality of 13-HETE formed by adult and fetal liver
microsomes differed. 13-HETE from fetal liver contained >70% of the
R stereoisomer, and that from adult liver contained >70%
of the S stereoisomer. This finding suggested that fetal and
adult liver microsomes formed 13-HETE by different enzymes. Neither TAO
nor furafylline reduced the bisallylic hydroxylation activity of fetal
liver microsomes.
The lack of effect of TAO on the bisallylic hydroxylation activity of
adult and fetal human microsomes was unexpected. CYP3A4 and CYP3A7 are
quantitatively major enzymes in adult and fetal human liver,
respectively (Guengerich, 1995). We also confirmed that our human adult
microsomes catalyzed a typical CYP3A4 reaction, the 6-hydroxylation
of testosterone. The difference between human liver microsomes and
recombinant CYP3A4 could be related to differences in coupling between
cytochrome P450 reductase and CYP3A4. The recombinant CYP3A4 was
coexpressed with NADPH-P450 reductase. The enzymatic activity of CYP3A4
in microsomes can be augmented by 0.1 mM 7,8-benzoflavone (Berthou
et al., 1994; Schwab et al., 1988; Yun et
al., 1992). The mechanism is unknown, but the drug may facilitate
electron flow from the reductase to the monooxygenase. We observed that
7,8-benzoflavone (0.1 mM) could increase the bisallylic hydroxylation
activity of adult human liver microsomes. 13-HETE, which was formed in
the presence of 7,8-benzoflavone, contained more 13R-HETE and was thus
formed with less stereospecificity. CYP3A4 biosynthesized racemic
13-HETE. It therefore seems likely that 7,8-benzoflavone stimulated the
bisallylic hydroxylation activity of CYP3A4 in the microsomes.
Rat liver enzymes.
We extended our studies on induction of
bisallylic hydroxylation activity of rat liver microsomes
(Hörnsten et al., 1996). Treatment with two different
inducers of CYP3A1, DEX and ERY, increased the bisallylic hydroxylation
activity in rats of both sexes. Furthermore, TAO and polyclonal
antibodies against CYP3A1 strongly inhibited this enzymatic reaction.
It therefore seems likely that the principal bisallylic hydroxylation
activity of liver microsomes of rats treated with DEX and ERY is
associated with CYP3A1 or any of its closely related isozymes (Kirita
and Matsubara, 1993; Komori and Oda, 1994; Pereira and Lechner, 1995). Phenobarbital treatment can also induce CYP3A1 (Larsen and Jefcoate, 1995). This may contribute to the fact that liver microsomes of DEX and
phenobarbital-treated rats formed 11-HODE with the same chirality
(58-59% 11R-HETE).
Hydroxylation mechanisms.
Bisallylic hydroxy metabolites are
likely formed, in analogy with other hydroxylations, by hydrogen
abstraction by the heme ferryl oxygen and "oxygen rebound" (Ortiz
de Montellano, 1995). Oxygen is inserted with retention of
configuration during biosynthesis of 11-HODE (Oliw et al.,
1993). The bisallylic carbon must thus be positioned near the active
center of the enzyme. CYP1A2 and CYP3A4 apparently allowed all three
bisallylic carbons of 20:4n-6 to interact with the heme iron. CYP2C9
and CYP2C8 differed; they formed 13-HETE as a main hydroxy metabolite,
small amounts of 10-HETE and relatively large amounts of certain
cis-trans conjugated HETEs (CYP2C9, mainly 12R-HETE; CYP2C8,
mainly 11-HETE and 15-HETE). Both enzymes can abstract a hydrogen from
C13 and C10 and insert oxygen at these carbons. It seems likely that
the double bond may also migrate. After hydrogen abstraction at C13 by
CYP2C8, oxygen could be inserted at C11 and C15, yielding 11-HETE and 15-HETE. After hydrogen abstraction at C10 by CYP2C9, oxygen could be
inserted at C12, yielding 12R-HETE. Linoleic acid can undergo similar
transformations to 11R,S-HODE, 9R-HODE and 13R-HODE by rat liver
microsomes (Oliw et al., 1993).
). Alternatively, the reactive ferryl oxygen may to oxidize the
double bonds (epoxidation) adjacent to the bisallylic carbon. CYP2C9,
for example, formed 14(15)epoxyeicosatrienoic acid as the main
metabolite of 20:4n-6. All five CYP enzymes, which catalyzed bisallylic
hydroxylations, were also found to epoxidize double bonds. It therefore
seems likely that epoxidation and bisallylic hydroxylations could be
connected. However, CYP102 of Bacillus megaterium has been
reported to have arachidonate 6 epoxygenase activity without
formation of bisallylic or cis-trans conjugated HETEs
(Capdevila et al., 1996). All epoxygenases therefore may not
catalyze bisallylic hydroxylations.
12R-HETE and 15R-HETE.
CYP2C9 synthesized relatively large
amounts of 12-HETE and mainly the 12R antipode. 12R-HETE is also a well
known metabolite formed by human and rat liver microsomes (McGiff,
1991). The enantiomeric purity is 80% when extracted at acidic pH
and >90% at neutral pH. The chirality of 12-HETE suggests that CYP2C9
may contribute to biosynthesis of 12R-HETE by adult human liver
microsomes. CYP2C9 could be useful as a model enzyme for studying this
reaction in detail. CYP2C9 and its allelic form R144C can hydroxylate a
large number of drugs (e.g., tolbutamide, phenytoin,
diclofenac sodium and many other nonsteroidal anti-inflammatory drugs;
Goldstein and de Morais, 1994), but none of these substrates seems to
be subject to hydroxylation with double bond migration.
). The present work
demonstrated that adult human liver microsomes mainly formed 13S-HETE.
This is an interesting analogy with oxygenation of 18:2n-6 by rat liver
microsomes. Biosynthesis of 11S-HODE and 13R-HODE from
stereospecifically deuterated 18:2n-6 occurred by abstraction of the
pro-S hydrogen at C11 of 18:2n-6 and suprafacial oxygen insertion (Oliw et al., 1993). From the biosynthesis of
12R-HETE and 10-HETE by CYP2C9, it clearly will be of interest to
determine whether this occurs with formation of the S
antipode of 10-HETE.
Biological significance of CYP metabolites.
A biological role
of epoxides of 20:4n-6, 20-HETE and 19-HETE has been implicated in
different organ systems, hormonal signaling and pathophysiological
processes (Capdevila et al., 1995; Oliw et al.,
1996; Rahman et al., 1997). Cell wall receptors have not yet
been identified for these metabolites, but ion channels could be
targets for their actions (Li and Campbell, 1997). As is the case with
leukotrienes, prostaglandins and other eicosanoids, stereochemical
features will determine potency and biological activity. The bisallylic
hydroxy metabolites of 20:4n-6 have not yet been investigated for
biological potency. Our results demonstrate that 13-HETE is the main
bisallylic hydroxy metabolite, which can be formed with different
stereoselectivity by human and fetal liver microsomes. The R
and S antipodes of this molecule could be worthy of future
biological studies.
Summary. Adult human liver microsomes, CYP1A2, CYP3A4, CYP2C8, CYP2C9, CYP2C19 and fetal liver microsomes catalyzed bisallylic hydroxylations of 18:2n-6 and 20:4n-6. Effects of specific CYP inhibitors and steric analysis of 13-HETE formed by human liver microsomes, CYP1A2 and CYP3A4 suggested that CYP1A2 could contribute to the bisallylic hydroxylation activity of adult human liver microsomes in vitro, whereas CYP3A4 seemed to be of minor importance. Human fetal liver microsomes possessed a relatively prominent bisallylic hydroxylation activity, which seemed to be unrelated to CYP1A2 and CYP3A7. The fetal CYP enzymes with this catalytic activity should be determined.
| |
Acknowledgments |
|---|
We thank Dr. M. Hamberg for steric analysis of 11-HODE and Dr. T. Luthman for providing a Chiralcel OD-H column.
| |
Footnotes |
|---|
Accepted for publication September 8, 1997.
Received for publication April 22, 1997.
1 This work was supported by the Swedish Medical Research Council (Grant 06523). K.V. is supported by the Swedish Institute.
2 This was identified by liquid chromatography-mass spectrometry (J. Ericsson, J. Bylund, C. Su and E. H. Oliw, unpublished observation).
Send reprint requests to: Dr. Ernst H Oliw, Division of Biochemical Pharmacology, Department of Pharmaceutical Biosciences, Uppsala Biomedical Center, Uppsala University, P.O. Box 591, S-751 24 Uppsala, Sweden.
| |
Abbreviations |
|---|
CYP, cytochrome P450; DEX, dexamethasone; ERY, erythromycin; GC-MS, gas chromatography-mass spectrometry; HETE, hydroxyeicosatetraenoic acid; HODE, hydroxyoctadecadienoic acid; HPLC, high-performance liquid chromatography; P450, cytochrome P450; RP, reverse phase; SP, straight phase; TAO, troleandomycin; TLC, thin-layer chromatography.
| |
References |
|---|
|
Top
Abstract
Introduction
Procedures
Results
Discussion
References
|
|---|
-hydroxylation in human liver.
Life Sci
52: 29-34. -naphthoflavone.
Mol Pharmacol
33: 493-499[Abstract]. This article has been cited by other articles:
|
|
 |
|
  M. Fer, L. Corcos, Y. Dreano, E. Plee-Gautier, J.-P. Salaun, F. Berthou, and Y. Amet Cytochromes P450 from family 4 are the main omega hydroxylating enzymes in humans: CYP4F3B is the prominent player in PUFA metabolism J. Lipid Res., November 1, 2008; 49(11): 2379 - 2389. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. Choudhary, I. Jansson, I. Stoilov, M. Sarfarazi, and J. B. Schenkman METABOLISM OF RETINOIDS AND ARACHIDONIC ACID BY HUMAN AND MOUSE CYTOCHROME P450 1B1 Drug Metab. Dispos., August 1, 2004; 32(8): 840 - 847. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 H. Wang, Y. Zhao, J. A. Bradbury, J. P. Graves, J. Foley, J. A. Blaisdell, J. A. Goldstein, and D. C. Zeldin Cloning, Expression, and Characterization of Three New Mouse Cytochrome P450 Enzymes and Partial Characterization of Their Fatty Acid Oxidation Activities Mol. Pharmacol., May 1, 2004; 65(5): 1148 - 1158. [Abstract] [Full Text]
|
|
|
|
 |
|
, Peak in the chromatograms had the same elution
time as [14C]11-HODE and was totally absent in control
insect cells and control lymphoblastoid cells. The other products were
not characterized, but diols, 17-HODE, 18-HODE and monoepoxides of
18:2n-6 eluted after 
