(Received for publication, October 22, 1996, and in revised form, December 23, 1996)
From the Department of Pharmacology and Toxicology, College of Pharmacy, University of Arizona, Tucson, Arizona 85721
Alanine-scanning mutagenesis was performed on
amino acid residues 210-216 of cytochrome P450 3A4, the major
drug-metabolizing enzyme of human liver. Mutagenesis of this region,
which has been proposed to align with the C-terminal ends of F-helices
from cytochromes P450BM-3, P450terp, and
P450cam, served as a test of the applicability of the
substrate recognition site model of Gotoh (Gotoh, O. (1992) J. Biol. Chem. 267, 83-90) to P450 3A4. The results,
using two steroid substrates, indicated that substitution of Ala for
Leu210 altered the responsiveness to the effector
-naphthoflavone and the regioselectivity of testosterone
hydroxylation. Replacement of Leu211 by Ala also decreased
the stimulation by
-naphthoflavone, whereas mutations at residues
212-216 had little effect. The diminished flavonoid responses of the
210 and 211 mutants were observed over a wide range of progesterone and
-naphthoflavone concentrations. Further characterization was
performed with the additional effectors
-naphthoflavone, flavone,
and 4-chromanone. The finding that P450 3A4 with one altered residue,
Leu210
Ala, can have both an altered testosterone
hydroxylation profile and response to flavonoid stimulation provides
evidence that the substrate binding and effector sites are at least
partially overlapping.
Human cytochrome P450 3A4 (P450 3A4)1
is capable of catalyzing the oxidative metabolism of a wide range of
chemical compounds of very different structures (1). Of particular
interest is the large number of drugs known to be metabolized by P450
3A4 (2). This enzyme may account for as much as 60% of the P450 in a
human liver sample (3), making it one of the single most important
human xenobiotic-metabolizing enzymes. Although 3A4 can accommodate a
wide variety of structurally diverse substrates, it still exhibits
remarkable regio- and stereoselectivity with many compounds. For
example, the enzyme catalyzes the 2-, 6
-, and 15
-hydroxylation
of testosterone, the 6
- and 16
-hydroxylation of progesterone (4),
the 1- and 4-hydroxylation of midazolam (5), and the M1-, M17-, and
M21-oxidation of cyclosporin A (6). In addition to these properties,
metabolic activities of enzymes of the P450 3A subfamily are modulated
by naturally occurring phenolic compounds known as flavonoids,
resulting in either the stimulation or inhibition of enzyme activity
(7-11). Several sites for effector binding have been postulated based on the proposed mechanisms of action. Hypotheses include (a)
-NF increases the affinity of P450 for the substrate (12),
(b)
-NF increases the affinity of P450 for reductase
(10), (c)
-NF binds in the same pocket as the substrate
and increases coupling efficiency (13), or (d)
-NF is an
allosteric effector, binding at a distinct site and causing a
conformational change in the substrate binding pocket (12-14).
Despite the wealth of information on the importance, regulation, and
substrate specificity of the cytochrome P450 3A subfamily, structure-function analysis of these enzymes has not been rigorously approached. Unlike cytochromes P450 of family 2, 3A enzymes within or
across species exhibit few dramatic substrate specificity differences that could provide obvious leads for site-directed mutagenesis of
particular residues. Based on comparative sequence alignments and
analogy with P450 101, the existence of six substrate recognition sites
(SRSs) within family 2 has been proposed (15). Extensive analysis of
P450 2B enzymes by this laboratory (16-23) and of enzymes in the P450
2A (24-26) and 2C (27-32) subfamilies have so far confirmed that
virtually all amino acid residues and chimeric fragments identified as
critical for the substrate specificity of P450 2 forms fall within or
near the putative SRSs. To assess the applicability of the SRS model to
the P450 subfamily 3A we have targeted P450 3A4 residues that are
predicted (33) by a structure-based alignment to correspond to active
site residues from P450BM-3, P450cam, and
P450terp with the goal of identifying residues that form or influence the enzyme active site and/or effector binding site. We have
chosen to start our analysis of P450 3A structure-function at a region
that has been aligned by Hasemann et al. (33) with SRS-2.
The SRS-2 residues 209 of P450s 2A4 and 2A5 (24-26, 34) and 206 of
P450 2B1 (17) have been identified as substrate contact residues. The
precise equivalents of these family 2 residues in P450BM-3,
P450cam, or P450terp are uncertain due to a
high degree of structural variability among the bacterial enzymes in
these regions. We reasoned that verification of an equivalent SRS-like region in P450 3A4 in this difficult to align region would be an
excellent indicator of the accuracy of the alignment and of whether
other SRSs predicted by the alignment are likely to exist. We report
here the use of alanine-scanning mutagenesis to examine the
structure-activity relationships of seven residues from the P450 3A4
region analogous to SRS-2. Based on our analysis, a Leu Ala
substitution at residue 210 or 211 diminishes responsiveness to
-NF.
This report details the first site-directed mutagenesis study of any
P450 3A enzyme and provides definitive evidence for the location of
P450 3A4 residues that influence flavonoid stimulation in a region
shown to constitute part of the active site of several bacterial and
mammalian cytochromes P450.
Primers for PCR amplification were obtained from
the University of Arizona Macromolecular Structure Facility (Tucson,
AZ) and National Biosciences, Inc. (Plymouth, MN). Restriction
endonucleases and bacterial growth media were purchased from Life
Technologies, Inc. CHAPS, progesterone, testosterone, -NF, NADPH,
and DOPC were purchased from Sigma.
[4-14C]Progesterone was obtained from DuPont NEN.
[4-14C]Testosterone was obtained from Amersham Life
Science, Inc. HEPES was purchased from Calbiochem. Thin layer
chromatography plates (silica gel, 250 mm, Si 250 PA (19C)) were
purchased from Baker (Phillipsburg, NJ). All other reagents and
supplies not listed were obtained from standard sources.
The P450 3A4 cDNA
was isolated from human liver by PCR amplification of a first strand
cDNA library created using the 3-rapid amplification of cDNA
ends kit (Life Technologies, Inc.) and total human liver RNA (Clontech;
Palo Alto, CA). The N-terminal (5
-GTAGTGATGGCTCTCAGCCCAGACTTGGCC-3
) and C-terminal (5
-GAAAATTCAGGCTCCACTTACGGTGCCATCCCTTGACTCAACCTT-3
) primers were designed to amplify the entire 1.5-kilobase pair coding
sequence (35). Reaction conditions were as follows: one cycle of
94 °C for 5 min, 45 °C for 5 min, and 72 °C for 5 min, followed by 29 cycles of 94 °C for 2 min, 45 °C for 2 min, and 72 °C for 2 min. To increase expression of the cDNA in
Escherichia coli, N-terminal modifications were made by PCR
using the primer 5
-GGGC
CTCTGTTATTAGCAGTTTTTCTGGTGCTCCT-3
as
described previously (36) except for the incorporation of an
NcoI site (underlined) to facilitate cloning rather than an
NdeI site. Following modification, the 3A4 cDNA was
cloned directly into the pCRII vector using the TA Cloning Kit
(Invitrogen, San Diego, CA) and subsequently subcloned as an
NcoI-KpnI fragment into the E. coli
expression plasmid pSE380 (Pharmacia Biotech Inc.), creating pSE3A4. A
BamHI site that originated from the pCRII multiple cloning
site was deleted from pSE3A4 by cutting at the closely spaced
SpeI sites and removing the short region containing the
extra BamHI site followed by religation of the
SpeI ends. The entire modified 3A4 insert was sequenced and
was determined to be most similar to the 3A4 variant hPCN1 (37),
differing only by a silent mutation (CTC
CTG) at the position that
encodes Leu449.
Primers
used to modify codons 210-216 were designed in the reverse orientation
and are shown in Fig. 1. The forward primer used for PCR
annealed to a position that corresponds to codons 146-152. PCR
conditions were as follows: 1 cycle of 94 °C for 5 min, followed by
29 cycles of 94 °C for 1 min, 45 °C for 1 min, and 72 °C for 2 min. The resulting PCR products were digested with AatII and
BamHI, creating a 128-base pair fragment that was used to
replace the wild type segment in pSE3A4. The sequence of the entire
128-base pair mutagenized fragment was confirmed by sequencing.
Preparation of Solubilized E. coli Membranes
CHAPS-solubilized membrane preparations were made
essentially as described previously (38). Briefly, pSE3A4-containing
E. coli DH5 cells were grown at 37 °C with 240-rpm
shaking in 250 ml of liquid TB media (12 g of Bacto-tryptone, 24 g
of Bacto-yeast extract, 4 ml/liter glycerol) to midlog phase before
adding 1.0 mM
isopropyl-1-thio-
-D-galactopyranoside and 80 mg/ml
-amino levulinic acid. Cells were harvested after an additional 48-h incubation at 30 °C with 190-rpm shaking. Membranes were solubilized in MOPS buffer (100 mM MOPS (pH 7.3), 10% glycerol, 0.2 mM dithiothreitol, 1 mM EDTA) containing 0.5%
CHAPS. Recovery yields of 120-160 nmol of P450 3A4/liter of culture
were routine as determined from reduced carbon monoxide difference
spectra. P420 contamination was <5%.
The
cDNA encoding rat NADPH-P450 reductase was kindly provided by Dr.
Todd Porter (College of Pharmacy, University of Kentucky, Lexington,
KY). Reductase was expressed from a T7 expression plasmid, which was
created by moving the XbaI-HindIII fragment from
pOR262 (39) containing the N-terminally modified rat reductase with a
fused ompA signal peptide and ribosome binding site into
pET29a(+) (Novagen, Madison, WI). The resulting plasmid, pETOR262, was
transformed into E. coli strain HMS174(DE3) (Novagen). Cells
were grown at 37 °C with 240 rpm in 1-liter cultures of TB medium
with 30 µg/ml kanamycin to midlog phase and then induced with 0.01 mM isopropyl-1-thio--D-galactopyranoside and
grown for an additional 20 h at 30 °C with 190-rpm shaking. Reductase was solubilized from E. coli membranes and
purified on a 2
,5
-ADP column as described previously (39).
CHAPS-solubilized E. coli membrane preparations were used directly by reconstituting 10 pmol of P450 with 40 pmol of E. coli-expressed rat
NADPH-P450 reductase, 10 pmol of rat cytochrome
b5, and 0.1 mg/ml DOPC in a minimal volume;
assays were performed for 5-15 min at 37 °C in 15 mM
MgCl2, 50 mM HEPES buffer (pH 7.6), 0.1 mg/ml
DOPC, 0.06% CHAPS, and 1 mM NADPH. Steroid and -NF
stock solutions were made in 100% methanol. Care was taken so that
methanol concentrations in all reactions were equivalent and did not
exceed 1% of the total reaction volume. Progesterone and
-NF
concentrations varied according to the experiment and are described in
the individual figure legends. Assays involving testosterone were
performed using 25 µM testosterone with or without 25 µM
-NF. Identification of metabolites of progesterone
(6
-OH- and 16
-OH-progesterone) and testosterone (2
-OH-,
6
-OH-, and 16
-OH-testosterone) was by co-chromatography using
authentic standards (Steraloids, Wilton, NH). A commercial source for
15
-OH-testosterone was not available; identification of this
metabolite was by relative mobility in TLC and comparison with the
published values for TLC analysis of testosterone metabolites (40).
The SRS model (15) for P450 family 2 enzymes has proven
to be useful for predicting residues that may influence the geometry of
the active site and thus affect enzymatic activity. In a search for
residues that determine substrate specificity and flavonoid stimulation
of human P450 3A4, a P450 family 3 enzyme, we made site-directed
mutations over a short region in P450 3A4 that is predicted to align
with the C-terminal end of helix F. This region contains the active
site residues Thr185, Phe188, and
Met185 from P450cam, P450terp, and
P450BM-3, respectively (Fig. 2), and
corresponds to SRS-2 in P450 family 2. Alanine-scanning mutagenesis (41) was used to individually change P450 3A4 amino acid residues 210-216 to alanine. All of the resulting alanine substitution mutants
were efficiently expressed in an E. coli heterologous expression system at levels comparable with the N-terminally modified wild type construct (data not shown).
Progesterone Hydroxylase Activities of P450 3A4 Wild Type and Alanine Substitution Mutants 210-216
-Solubilized membrane
fractions from E. coli expressing P450 3A4 wild type or
individual alanine substitution mutants were examined in reconstituted
systems containing HEPES buffer with MgCl2, cytochrome
b5, DOPC, and 0.06% CHAPS as described under "Experimental Procedures" for progesterone hydroxylase activity. Table I displays the results of these analyses. All
mutants retained progesterone hydroxylase activity, forming two
predominant metabolites, 6-OH and 16
-OH progesterone. In the
absence of
-NF, the ratios of 6
-hydroxylase to
16
-hydroxylase activities (
/
ratios) of most of the
mutants were similar to the wild type enzyme, the exception being
mutants L210A and L211A, which had slightly elevated
/
ratios.
|
Previous studies have shown that, depending on the substrate and -NF
concentrations employed, individual metabolites from a single substrate
can be differentially stimulated/inhibited by
-NF (13, 42). Table I
demonstrates that the 16
-hydroxylase activity of wild type P450 3A4
is stimulated to a slightly greater extent by
-NF than the
6
-hydroxylase activity of the enzyme. At a substrate concentration
of 25 µM progesterone, wild type P450 3A4 had a
/
ratio equal to 6.0 in the presence of the methanol control and 4.5 in
the presence of 25 µM
-NF. One of the mutants, L210A,
showed the opposite pattern of stimulation; i.e.
6
-hydroxylase activity was stimulated more by
-NF than the
16
-hydroxylase activity. Because production of the
16
-OH-progesterone metabolite was only stimulated 1.3-fold (compared
with 6.0-fold for wild type enzyme) in the presence of
-NF, the
/
ratio was 19.2 for L210A as compared with 4.5 for wild type
P450 3A4. Another interesting mutant is L211A, which showed less than
2-fold stimulation of both the 6
- and 16
-hydroxylase activities
by
-NF (1.2- and 1.8-fold, respectively). Alanine substitution
mutants 212-216 maintained
/
ratios more closely resembling the
wild type enzyme in either the absence or presence of
-NF; for these
mutants,
-NF always caused at least 2.2-fold stimulation and
increased the 16
- more than the 6
-hydroxylase activity.
The effect of -NF on the observed progesterone hydroxylase activity
was tested over a range of
-NF concentrations to determine whether
the low levels of stimulation of the activities of the L210A and L211A
mutants could be overcome by using higher concentrations of
-NF.
When the progesterone concentration was kept at 25 µM and
the
-NF concentration was increased to 200 µM, no
further increase in progesterone hydroxylation was seen for P450 3A4
wild type, L210A, or L211A as compared with the stimulation by 25 µM
-NF (data not shown).
It was previously noted (12) that
in the absence of -NF the rates of progesterone 6
-hydroxylation
by rabbit and human liver microsomes gave nonlinear kinetics in a
double reciprocal plot. Upon the addition of
-NF, however, the
double reciprocal plot became linear. To determine whether the kinetics
for L210A and L211A, the two mutants least responsive to
-NF
stimulation, were similar to P450 3A4 wild type, the dependence of
progesterone 6
-hydroxylation on substrate concentration in the
presence or absence of 25 µM
-NF was determined (Fig.
3). Double reciprocal plots showed that wild type P450
3A4, L210A, and L211A displayed nonlinear kinetics in the absence of
-NF. When 25 µM
-NF was included, the double
reciprocal plot became linear.
The kinetic analyses revealed that at high substrate concentrations the
progesterone hydroxylase velocities obtained in the absence of -NF
approached those obtained in its presence. To better illustrate this
phenomenon, the -fold stimulation by 25 µM
-NF was
plotted as a function of progesterone concentration (Fig.
4). At high substrate concentrations (320 µM progesterone), essentially no stimulation of
6
-hydroxylase activity by
-NF was seen for wild type P450 3A4,
L210A, or L211A (Fig. 4A). As the progesterone concentration
was lowered, wild type 3A4 became more responsive to stimulation by
-NF. P450 3A4 mutants L210A and L211A were less responsive than wild
type to
-NF stimulation, especially at the lower concentrations of
progesterone. For example, at 10 µM progesterone and 25 µM
-NF, the 6
-hydroxylase activity of L210A was
stimulated 4.3-fold and L211A was stimulated 2.6-fold as compared with
the 10-fold stimulation of the wild type enzyme. Fig. 4B
demonstrates a similar relationship between stimulation of progesterone
16
-hydroxylation and progesterone concentration. At 320 µM progesterone, no
-NF stimulation of
16
-hydroxylase activity was observed. As the progesterone
concentration was decreased, the responsiveness of the wild type enzyme
to
-NF stimulation increased. Thus, at 10 µM
progesterone, the 16
-hydroxylase activity was stimulated 30-fold by
-NF. P450 L211A was stimulated only 6-fold at 10 µM
progesterone, and L210A was stimulated just 2-fold.
Stereoselectivity of Testosterone Hydroxylation by Alanine Substitution Mutants
Table II summarizes the
hydroxylase activities for wild type P450 3A4 and all of the mutants.
As previously reported (4, 42), wild type P450 3A4 produces primarily
three metabolites, 2-OH-, 6
-OH-, and 15
-OH-testosterone, with
6
-hydroxylase accounting for 91% of the activity. All mutants
retained testosterone hydroxylase activity, remaining primarily
testosterone 6
-hydroxylases. In the absence of
-NF, the
metabolite percentages were approximately the same between the wild
type enzyme and alanine substitution mutants 212-216. L210A had lower
testosterone 2
- and 15
-hydroxylase activities than the wild type
enzyme both in the absence and presence of
-NF. L210A produced an
additional metabolite, 16
-OH-testosterone, at a rate (0.08 nmol/min/nmol) much higher than the wild type enzyme (<0.01
nmol/min/nmol). L211A, in the absence of
-NF, had a higher
percentage of 2
- and 15
-hydroxylase activities than the wild type
enzyme.
|
As with the unequal -NF stimulation of wild type P450 3A4
progesterone 16
- and 6
-hydroxylase activities, the testosterone 2
- and 15
-hydroxylase activities were preferentially stimulated 12.6- and 9.6-fold, respectively, as compared with the 2.9-fold stimulation of the testosterone 6
-hydroxylase activity. In the presence of
-NF the testosterone metabolite profiles of mutants at
positions 212-216 remained remarkably similar to the wild type enzyme.
However, altered
-NF responsiveness of P450 3A4 mutants L210A and
L211A was observed with testosterone, as illustrated in Fig.
5. L210A, because of decreased stimulation of 2
- and 15
-hydroxylase activities, maintained a high percentage of
6
-OH-testosterone activity (94% versus the wild type
74%). In contrast, the P450 3A4 mutant L211A, because of decreased
stimulation of 6
-OH testosterone activity, displayed a lowered
percentage of 6
-OH-testosterone formation (53% versus
the wild type 74%). F213A showed
-NF stimulation of 2
- and
15
-hydroxylase activities 2-fold greater than wild type enzyme,
although the metabolite percentages in the presence of
-NF closely
resembled the wild type profile. The remaining mutants maintained
testosterone metabolite ratios and responsiveness to
-NF very
similar to the wild type enzyme (Table II).
Response of P450 3A4 Wild Type, L210A, and L211A to Partial or Modified Analogs of
A previous study (13) tested a number
of partial or modified analogs of -NF for their ability to modulate
P450 3A4 activity. Using polycyclic aromatic hydrocarbons as test
substrates, it was found that
-NF, flavone, and 4-chromanone could
stimulate P450 3A4 activity and that
-NF was a potent inhibitor of
enzyme activity. In an attempt to ascertain whether the P450 3A4
mutants L210A and/or L211A displayed an altered response to
stimulators/inhibitors other than
-NF, the effects of flavone,
4-chromanone, and
-NF on progesterone hydroxylase activities were
tested (Fig. 6). Similar to the previous study (13),
flavone was an effective stimulator of P450 3A4 as measured by
progesterone 6
- and 16
-hydroxylase activities. However, rather
than being an inhibitor,
-NF was a mild stimulator of P450 3A4
progesterone hydroxylase activity (1.3- and 1.9-fold stimulation of the
6
-OH and 16
-OH progesterone activities, respectively). In
addition, in contrast to its stimulation of polycyclic aromatic
hydrocarbon metabolism, 4-chromanone had largely no effect on
progesterone hydroxylase activities. Interestingly, the P450 3A4 mutant
L210A displayed near wild type levels of stimulation by flavone and
-NF with respect to progesterone 6
-hydroxylase activity; however,
none of the compounds affected the 16
-hydroxylase activity of this
mutant. The progesterone 6
- and 16
-hydroxylase activities of
L211A were affected by less than 2-fold for each of the
stimulators/inhibitors tested.
The results presented here describe for the first time the
identification of certain key amino acid residues that play a role in
substrate specificity and flavonoid stimulation of P450 3A4. The region
selected for mutagenesis was predicted to be structurally similar to a
portion of the F-helices of several bacterial enzymes of known crystal
structure (33). However, even the alignment of proteins of known
structure is subject to interpretation in this and several other key
regions (for review, see Ref. 43). For example, a previous study of
SRS-1 of P450 2C2 used a degenerated cassette to introduce random amino
acid substitutions into a portion of the enzyme; this was helpful in
confirming an alignment with P450cam (32). Targeting
contiguous residues that are predicted to form SRSs has the potential
of identifying both the residues that comprise the active site and,
just as importantly, adjacent residues that do not contribute to the
active site. For mutagenesis of the proposed SRS-2 region of P450 3A4,
Ala was used as the substituting residue because of its small side
chain, low tendency to distort the C- backbone, and minimal
electrostatic effects. Of the seven P450 3A4 residues targeted by Ala
in this report, substitutions at positions 212 and 214-216 did not
affect activity, suggesting that these residues do not effectively
influence the substrate or effector binding sites. F213A displayed
greater
-NF stimulation of testosterone hydroxylase activities than
wild type P450 3A4. The remaining two mutants, L210A and L211A,
exhibited altered basal activity and/or flavonoid responsiveness,
suggesting that the substrate binding and effector sites are at least
partially overlapping.
The presence of highly conserved residues can be an indication of the
importance of a position to enzyme function. The two Leu residues at
positions 210 and 211 of P450 3A4 are highly conserved, being present
in P450s 3A1 and 3A2 from rat, 3A3 and 3A7 from human, 3A6 from rabbit,
3A8 from monkey, and 3A12 from dog. The exception is human P450 3A5,
which instead has Phe at position 210. Substitution of Ala for Leu at
residues 210 and 211 did not cause P450 3A4 to lose its preference for
steroid 6-hydroxylation. This observation is in contrast to
mutations at SRS-2 position 209 of 2A4 and 2A5 (24, 26, 44) and the
corresponding position 206 of 2B1 (17), which conferred a radical shift
in steroid hydroxylase specificities. Hydroxylation of steroids by P450
3A4, which is likely to have a very large active site, may be
controlled more by the inherent chemical reactivity of the allylic
6-position than by steric constraints exerted by the enzyme, similar to
P450 2B4-mediated hydroxylation of camphor. With P450 2B4, because of
considerable movement of the substrate molecule in the active site,
camphor hydroxylation occurred preferentially at sites predicted by
chemical reactivities (45). Subtle effects on testosterone hydroxylation could be seen for the P450 3A4 mutants L210A and L211A at
positions other than the 6-position of the substrate. For example, in
the absence of stimulator, L210A had a lower percentage of testosterone
2
- and 15
-hydroxylation and increased testosterone 16
-hydroxylation compared with the wild type enzyme. Interestingly, substitution of Ala at the adjacent position, Leu211, had
an opposite effect on the percentage of testosterone 2
- and
15
-hydroxylase activities, increasing them above the wild type
levels. In the proposed alignment (Fig. 2), Leu210 aligns
with active site residues Thr185 of P450cam
and Phe188 of P450terp. In the alignment of
Hasemann et al. (33) Leu210 of P450 3A4 is also
aligned with position 209 of P450 2A4 and position 206 of P450 2B1.
These results suggest that the alignment of P450 3A4 with
P450terp, P450cam, and SRS-2 of P450 2A4 and P450 2B1 may be fairly accurate for this region. Met185 of
P450BM-3 is not aligned with P450 3A4 position 210, although it is not aligned with the corresponding residues in
P450terp or P450cam either. Another published
alignment has the residues from positions 185 of P450cam,
188 of P450terp, and 185 of P450BM-3 aligned with position
206 of P450 2B1 (46).
It was previously noted with polycyclic aromatic hydrocarbon substrates
(13) that -NF differentially stimulates P450 3A4 hydroxylation of
the narrower portions of the substrate. Our results with
-NF
stimulation of P450 3A4 testosterone and progesterone hydroxylation
follow this general rule; namely, the progesterone 16
-hydroxylase
activity and the testosterone 2
- and 15
-hydroxylase activities
(positions that are located at the narrow ends of both steroids) are
stimulated more than hydroxylation at the 6
(middle of the molecule)
position. The two substitutions Leu210
Ala and
Leu211
Ala seem to have opposite effects on
differential stimulation of hydroxylation at the end versus
the middle of the steroid; L210A shows less
-NF stimulation of end
position hydroxylation, while L211A shows less stimulation of
6
-hydroxylation. Previously, the importance of the exact residue
identity at position 209 of P450 2A5 was studied by engineering nine
individual amino acid substitutions (24). A more detailed analysis of
positions 210 and 211 of P450 3A4 will require additional
substitutions, additional substrates, and computer modeling (46).
The mutagenesis results, which identify residues that are able to
influence stereospecificity of testosterone hydroxylation and effector
activity, do not yet point to an obvious mechanism for -NF
stimulation. An early study with rabbit and human liver microsomes
demonstrated that
-NF stimulation of progesterone 16
-hydroxylation and 17
-estradiol 2-hydroxylation was largely due
to a decreased Km with little change in
Vmax (12). From their kinetic data the authors
were not able to distinguish whether the interaction between enzyme and
effector occurred at the active site or at a separate effector binding
site. A more recent study (13) using P450 3A4 expressed in HepG2 cells
presented evidence that phenanthrene and
-NF are present
simultaneously in the active site and that the stimulation of
phenanthrene metabolism by
-NF is a result of increased
Vmax and not decreased Km. Again, the authors were unable to pinpoint an exact mechanism for
-NF action. Our kinetic results (Fig. 3), using the data points from
the highest progesterone concentrations, are most consistent with
-NF stimulating P450 3A4 progesterone hydroxylation rates by
lowering the Km and not by increasing the
Vmax. The mechanism of
-NF stimulation of
P450 3A4 remains to be elucidated, but we do now have evidence that
single residue substitutions in an active site region of P450 3A4 can
influence the ability to respond to effector, suggesting that the
active site and the effector site are at least partially
overlapping.
The finding that the structure-based alignment of Hasemann et al. (33) was accurate enough to predict the existence of key P450 3A4 residues in one of the most difficult to align regions suggests that the SRS model may be highly appropriate for the 3A subfamily. We anticipate that mutagenesis of the other more conserved SRS regions will be useful for elucidating the key residues that determine structure-activity relationships of the important P450 3A subfamily. Additional mutants with altered substrate specificities and/or altered responsiveness to flavonoids should help to determine whether substrate binding and effector sites are partially or completely overlapping.
The use of mutant enzymes that exhibit enhanced as well as diminished
stimulation by -NF may prove to be valuable in elucidating the
mechanism of action of flavonoids on P450 3A enzymes. In order to
understand the mechanistic basis for flavonoid action, wild type and
mutant enzymes will need to be studied in terms of some of the key
individual steps in the P450 cycle. For example, the stoichiometry of
hydrogen peroxide or hydroxylated product formed/NADPH or oxygen
consumed can give a measure of the degree of uncoupling observed in the
presence or absence of effector. Recent studies with site-directed
mutants of bacterial cytochromes P450cam and P450BM-3 provide an excellent illustration of the
usefulness of these types of experiments in understanding the
mechanistic basis of altered substrate metabolism by mutant enzymes
(47, 48). The effects of cytochrome b5 will also
be examined in light of the previous finding that some mutations at
position 209 in SRS-2 of the P450 2A subfamily altered the coupling
efficiency of electron transport to substrate oxidation (49).
We thank Dr. Gilbert H. John for isolating the P450 3A4 cDNA and David Stepp for technical assistance.