| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
ARTICLES |
Department of Population Health and Reproduction, School of Veterinary Medicine, University of California (C.J.C., K.W.W., A.J.C.), Davis, California 95616; and the Center of Marine Biotechnology, University of Maryland Biotechnology Institute (J.M.T.), Baltimore, Maryland 21202
Address all correspondence and requests for reprints to: Dr. A. J. Conley, VM-PHR, School of Veterinary Medicine, University of California, Davis, California 95616. E-mail: ajconley{at}ucdavis.edu
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Aromatase cytochrome P450 is encoded by a single gene in humans, even though it is expressed in a broad array of tissues, including many that are steroidogenic in the classical sense as well as others generally considered nonsteroidogenic (15). For instance, in addition to the placenta and the gonads of both men (17) and women (18), P450arom is expressed in brain, liver (19), adipose tissue (20), and a number of tissues in the fetus (21). Additionally, tumors from numerous sites, such as breast (22, 23), gonads (24), prostate (19), and even myeloid leukemia cells (25), have been shown to express P450arom. An elaborate mechanism involving alternative splicing of untranslated first exons and associated promoter elements differentially regulates tissue-specific expression in these sites (15). The ovary is notable in this regard because it appears to be the only tissue expressing P450arom in which splicing is not invoked. Instead, the first untranslated exon is contiguous with exon II (26). Hinshelwood et al. (27) showed that this characteristic of human ovarian P450arom expression is apparently shared in equine and porcine species, even at the nucleotide level. Moreover, this contrasts the lack of sequence conservation in the 5'-untranslated region among placental P450arom transcripts from these species (27) and even P450arom transcripts expressed in the porcine testis (28). Thus, this feature of the basic structure of the CYP19 gene regulating P450arom expression appears conserved only in the female gonad. These observations are consistent with the suggestion that P450arom may have first evolved in the vertebrate ovary (15) and that expression in the placenta is a more recent event among eutherian mammals. How the functional properties of P450arom might have changed to accommodate the physiological needs of the developing fetus in utero for both androgen metabolism and/or estrogen synthesis is an interesting, but unanswered, question.
As noted above, and in contrast to the wealth of information on human P450arom, many fewer studies have been conducted on the aromatase enzyme system of other mammals. Recent investigations in this laboratory and others have determined that there are essential differences in the biology of P450arom in the pig. Unlike any other mammal investigated to date, the pig expresses functionally distinct isozymes in a tissue-specific manner (12). Evidence continues to accumulate that these are encoded by completely duplicated genes clustered on chromosome 1 (29, 30, 31). Our studies have shown that, unlike other mammals, a gonadal isozyme is expressed in the theca interna as well as the stratum granulosum of the ovarian follicle (32), the Leydig cell of the testes, and, interestingly, the zona reticularis of the adrenal gland of the pig (28). The porcine placenta expresses a second isozyme (12, 29), and a third is expressed in the early, preattachment porcine blastocyst (13). However, the roles served by multiple isozymes of P450arom in the pig remain unclear and are difficult to elucidate given such a broad range of cellular environments possibly influencing catalytic function, in addition to other factors, including levels of expression. In fact, few studies conducted to date have compared the function of isozymes of P450arom; none has done so in the same cellular background or examined isozymes that have evolved in the same species. Kinetic experiments conducted on recombinant enzyme generated by transient transfection often suffer from low, unpredictable levels of expression. Therefore, the catalytic characteristics of the placental and gonadal isozymes of porcine P450arom were compared under conditions of stable recombinant expression in the hope of gaining insight into the evolution of function of this important enzyme system. The existence of gonadal and placental isozymes of porcine P450arom provides a unique opportunity to explore the functional evolution of a highly conserved enzyme critical for reproductive success in mammals.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
,2-3H]androstenedione, 52.1 Ci/mmol;
[7-3H]androstenedione, 24.5 Ci/mmol;
[1ß,2ß-3H]testosterone, 40.3 Ci/mmol;
[1,2-3H]testosterone, 60.0 Ci/mmol; and
[7-3H]testosterone, 30.0 Ci/mmol. Authentic steroids were
purchased from Steraloids, Inc. (Wilton, NH) and Sigma Chemical Co. (St. Louis, MO). RIA reagents were purchased from
Diagnostics Systems Laboratories, Inc. (Webster, TX). HPLC
solvents were of ultrapure quality from Burdick and Jackson
(Muskegon, MI). Cell protein concentrations were estimated using
bicinchoninic acid protein assay reagent kits (Pierce Chemical Co., Rockford, IL). Immunoblots were conducted on polyvinyl
membranes (Immobilon-P, Millipore Corp., Bedford, MA), and
proteins were detected by chemiluminescence (ECL, Amersham Pharmacia Biotech, Arlington Heights, IL). The imidazole enzyme
inhibitors fadrazole (CGS16949A) and etomidate were gifts from
Ciba-Geigy (Summit, NJ) and Abbott Laboratories (Abbott Park, IL), respectively. All other
chemicals were of the highest quality available.
Transient transfection
Experiments were conducted in which the parent 293 cell line was
used for recombinant P450arom expression by transient transfection.
Full-length complementary DNAs (cDNAs) were constructed in the pCMV
expression plasmid. Conditions including levels of plasmid DNA (1.5
µg), Lipofectamine (5 µl), and time (6 h) were carefully optimized
for 5 x 105 cells cultured in six-well plates.
Kinetic analyses of aromatase activity using androstenedione and
testosterone were conducted 48 h posttransfection.
Establishment of cell lines exhibiting stable expression of porcine
aromatases
Cultures of the human 293 fetal kidney cell line were
maintained in DMEM high glucose with 10% FCS, 10
mM HEPES, and increasing concentrations (100–800 µg/ml)
of geneticin (G418). All cells were killed at 600 µg/ml during 6 days
of culture. Both placental and gonadal porcine P450arom cDNAs were
subcloned into the pcDNA3 vector (Invitrogen, Carlsbad,
CA) containing the NeoR ORF gene. Transfection of 293 cells
was carried out on 1 x 106 cells using 30 µl
Lipofectamine and 5 µg linearized plasmid DNA for 5 h. Cells
were placed into medium without G418 for 48 h, after which time
they were split 1:15 and transferred to selective medium containing 600
µg/ml G418 thereafter. On day 14, individual colonies were lifted (24
for each transfection) using sterilized circles of Whatman
filter paper (Clifton, NJ) soaked in trypsin solution and transferred
to 24-well plates. Cells were grown and subcultured under the same,
continuous antibiotic selection pressure until enough cells were
obtained to cryopreserve aliquots of 2.5 x 106
cells/ml, and separate lines expressing the gonadal and placental
isozymes were stored. Each line was recovered from liquid nitrogen and
cultured in G418 (600 µg/ml) to verify viability. P450arom activities
were determined at different stages of confluence and passage number by
the tritiated water assay using [1ß-3H]androstenedione,
as described below. Protein expression was further investigated by
Western immunoblot analysis, also described below. All subsequent
studies were conducted on lines designated G2 (gonadal P450arom
isozyme) and P11 (placental P450arom isozyme).
Aromatase activity by tritiated water assay
Screening cells in monolayer culture for P450arom activity
required measuring the incorporation of tritium into
3H2O from
[1ß-3H]androstenedione. Incubations were carried out at
37 C in 5% CO2 with 150 nM substrate (20%
labeled, 80% radioinert) in duplicate wells. Incubation times varied
from 30 min for the placental line to 2 h for the gonadal line.
Inhibition of P450arom activity was achieved with the addition of
fadrazole or etomidate (each at 1 µM) to the culture
medium along with the substrate. Reactions were stopped by the addition
of 0.5 vol cold 30% trichloroacetic acid and were extracted with 2 vol
chloroform. A portion of the aqueous phase was mixed with an equal
volume of 5% charcoal-0.5% dextran suspension, centrifuged for 30 min
at 2000 x g, and quantified by liquid scintillation
counting. Background values for 3H2O released
in the absence of cells were subtracted from each point. Substrate
saturation and kinetic studies were carried out as described above from
duplicate or triplicate wells using either
[1ß-3H]androstenedione or
[1ß,2ß-3H]testosterone as substrate.
Stereo-specific hydrogen loss
These measurements employed the use of [1ß-3H]-,
[1ß,2ß-3H]-, and
[1,2-3H]androstenedione and
[1ß,2ß-3H]- and
[1,2-3H]testosterone as substrates. Culture medium
(2 ml) containing substrate was added to 1 x 106
cells/35-mm dish. Aliquots were removed at numerous time points from
triplicate wells (100 µl for P450arom activity measurement by
tritiated water assay as described above, and 50 µl for estrone or
estradiol determination by RIA). The amount of
3H2O and the amount of estrogen formed with
each labeled substrate were used to calculate the distribution label as
the percentage retained in the estrogen product or released as
3H2O. This percentage, estimated at each time
point in replicated experiments, was used in the calculation of a
mean for each labeled substrate for each isozyme.
Estrone and 17ß-estradiol RIA
Estrogens were measured in culture medium without extraction
using a double antibody system with 125I-labeled estrone or
estradiol tracers, enabling assays to be performed on samples
containing 3H-labeled estrogens. Standard curves (0.2–500
pg/tube) for 17ß-estradiol and estrone were prepared in assay buffer
and culture medium equivalent to the diluted sample medium, and
antibody addition was adjusted to approximately 40% binding. Assays
were sensitive to 1.0 and 0.2 pg for estrone and estradiol,
respectively, and the intra- and interassay coefficients of variation
were less than 10% and 15%, respectively, in each case.
Western immunoblot analysis
Cells were homogenized in PBS containing 1% sodium cholate and
0.1% SDS and sonicated for 3 sec. Equal amounts of protein (50 µg)
were subjected to SDS-PAGE (8% gel) in buffer containing 50
mM Tris, 383 mM glycine, 0.1% SDS, and 0.4
mM EDTA. Separated proteins were transferred by electroblot
onto membranes and immunoblotted with antisera (1:1000) raised against
recombinant human P450arom protein (courtesy of Dr. N. Harada, Fujita
Health University, Aichi, Japan).
Steroid product analysis by HPLC
Medium from cell lines incubated with 7-3H-labeled
tracers were removed after 30 min to 2 h of culture. A mixture of
authentic steroids (2.5 nmol each of testosterone,
19-hydroxytestosterone, 19-nortestosterone, estradiol, androstenedione,
19-hydroxyandrostenedione, 19-norandrostenedione,
19-oxoandrostenedione, and estrone) was added and equilibrated with
each sample to facilitate extraction and to serve as standards during
analysis. The steroids were extracted from the medium with a 10-fold
volume of methylene chloride, and the organic phase was transferred and
evaporated under reduced pressure. The steroidal residue was subjected
to HPLC analysis (computer-controlled model HP1100 with photo diode
array detector, Hewlett-Packard Co., Palo Alto,
CA), and the radioactivity of the eluant was monitored by an
in-line radioisotope detector (Ramona 2000, RayTest, New Castle, DE).
The quarternary mobile phase was made up of water, methanol,
acetonitrile, and tetrahydrofuran (solvents A–D, respectively) flowing
at 1 ml/min. The column (4.6 x 25 mm, 5µ-Ultrasphere ODS,
Beckman Coulter, Inc., Fullerton, CA) was maintained at 30
C and initially equilibrated to 80% A, 10% B, 6% C, and 4% D. These
conditions were maintained for the first 3 min of the gradient, after
which the solvent composition changed linearly over 15 min to 63% A,
20% B, 11% C, and 6% D. Over the next 14 min, the solvent
composition linearly changed to 45.5% A, 25.6% B, 15.9% C, and 13%
D. The solvent system quickly (1 min) changed to 10% B, 83% C, and
4% D, and these conditions were held for an additional 2 min. The
column was reequilibrated to the initial solvent conditions for 15 min
before the next injection of sample.
Statistical analysis
Product formation was analyzed by linear regression. Kinetic
constants were evaluated by ANOVA using the general linear models and
Kruskal-Wallis nonparametric test procedures (SAS Institute, Inc., Cary, NC).
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Cell lines exhibiting stable expression of the porcine placental and
gonadal isozymes of P450arom were developed to overcome the problems
encountered with transient transfection, especially with the gonadal
P450arom. Multiple independent cell lines expressing the gonadal
(n = 6) and placental (n = 11) isozymes of porcine P450arom
were isolated from 293 colonies under constant antibiotic selection
pressure. Consistently, cells lines expressing the gonadal P450arom
isozyme plated more efficiently, formed simple monolayers, and reached
confluence earlier than lines expressing the placental isozyme. Cell
lines expressing the placental P450arom isozyme not only grew more
slowly, but also tended to grow in distinct clusters around the dish
rather than as a more uniform monolayer. These growth characteristics
were not seen to change in the presence of the P450arom inhibitor
fadrazole (1 µM). Lines were also screened for expression
of P450arom by immunoblot analysis. Similar levels of immunodetectable
P450arom were demonstrated in cell lines P11 and G2, expressing the
placental and gonadal isozymes, respectively. The apparent molecular
size was approximately 48 kDa, and no expression was observed in the
nontransfected parent 293 cells (Fig. 1
).
|
). Consistently higher activity was
exhibited by the cell line expressing the placental isozyme than by
that expressing the gonadal isozyme of porcine P450arom. Activity
varied from 25–41 pmol/mg·2 h in the gonadal cell line and from
86–239 pmol/mg·2 h in the placental cell line as confluence
progressed from 40–100% before and after passage (Fig. 2). The lowest
catalytic rates were observed when cells had reached 100% confluence,
and all subsequent experiments were conducted on cells at a density of
around 80% to maximize the yield of the recombinant P450arom. Greater
than 97% of the aromatase activity was inhibited by fadrazole (1
µM) addition in both cell lines. Gonadal P450arom
activity was also inhibited by 92% in the presence of the other
imidazole inhibitor, etomidate (1 µM), which marginally
inhibited placental P450arom activity by less than 10%.
|
. The rate of estrogen formation was
generally greater from androstenedione than from testosterone and
considerably faster for the placental P450arom than the gonadal
isozyme, consistent with the results of activity analyses. Product
formation was confirmed in parallel cultures by HPLC analysis, which
demonstrated the presence of estrone, 19-oxoandrostenedione, and
19-hydroxyandrostenedione (Fig. 4).
Similarly, estradiol, 19-oxotestosterone, and 19-hydroxytestosterone
were observed in cultures incubated with testosterone (data not shown).
No evidence of other steroid products was noted either early (<10%)
or late (>50%) in the metabolism of either of the substrates. These
results also further confirmed the lack of interconversion of either
androgen substrates or estrogen products by endogenous oxidoreductases
during the course of these experiments.
|
|
,2-3H-labeled steroids and expressed as a percentage
of estrone and estradiol synthesis determined in medium by RIA after
various times of incubation. An example is illustrated in Fig. 5, which presents the results of
experiments examining [1ß-3H]- and
[1ß,2ß-3H]androstenedione metabolism by the placental
P450arom. The results of metabolism of all tracers by both isozymes are
summarized in Table 1. These data
demonstrated a preferential loss of label from
[1ß-3H]androstenedione to form
3H2O (Table 1) which represented about 90% of
the estrone formed. However, the rate of 3H2O
formation from tracers labeled in both the 1ß- and 2ß-positions was
lower for the placental isozyme than for the gonadal P450arom, and this
difference was particularly marked for
[1ß,2ß-3H]testosterone (56% vs. 84%,
respectively). Similar studies conducted with transiently expressed,
human P450arom demonstrated rates of 3H2O
release very similar to those obtained with the porcine placental
isozyme (Table 1). Only low rates of 3H2O
formation were observed from the 1,2-3H-labeled
tracers.
|
|
) from which apparent
Km values were calculated (Table 2). The Km for both
substrates tended to be lower for the placental isozyme than for the
gonadal isozyme and significantly lower (P < 0.05) for
testosterone than for androstenedione metabolism by the porcine
placental P450arom isozyme. There was little difference in
Vmax for androstenedione and testosterone aromatization by
the placental isozyme (244 and 198 pmol/mg·h), but a 3-fold higher
Vmax was observed for the metabolism of androstenedione
vs. testosterone for the gonadal isozyme (209 and 69
pmol/mg·h, respectively).
|
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The possible existence of isozymes of P450arom in any species was first recognized in the pig by Corbin et al. (12). Subsequent studies in this laboratory and others have confirmed the existence of multiple genes encoding the porcine P450arom isozymes (29, 30, 31). Although still a unique finding among mammals, at least two genes are also known to encode different isozymes of P450arom in the goldfish (14, 35), a polyploid species. The results of the present study, consistent with the preliminary observations (12), indicate that there are distinct functional differences between the placental and gonadal isozymes of porcine P450arom. Cell lines exhibiting stable expression of each isozyme, at similar levels based on immunoblot analysis, demonstrated slower aromatization of testosterone, particularly so for the gonadal isozyme, which, unlike the placental P450arom, was also susceptible to inhibition by etomidate (12). The differential sensitivity to inhibition by etomidate, generally considered an inhibitor of 11ß-hydroxylase cytochrome P450, is consistent with functionally important, structural variations in the substrate binding pockets of the placental and gonadal isozymes. Product analysis verified the formation of intermediates (19-hydroxy- and 19-oxoandrogens), apparently released from the catalytic site during oxidation before estrogen formation. Earlier studies of aromatase with porcine ovarian microsomes (36, 37) as well as others studying human, rat, and equine aromatases (38, 39, 40) demonstrated a significant release of intermediates. No evidence was found for the formation of 19-norandrostenedione or 19-nortestosterone, even though both steroids are resolved by the HPLC system that was developed and used in the current studies. This contrasts the results of Khalil et al. (41, 42), but is consistent with the report by Garrett et al. (43); both investigators used porcine granulosa cells as the source of P450arom. Garrett et al. (43) suggested that nonenzymatic conversion during sample processing might explain the appearance of 19-norandrogens in medium from cultured porcine granulosa cells. If enzymatic synthesis does occur, the conditions for formation of 19-nor products were not reproduced in that or the present study.
The establishment of cell lines expressing the placental and gonadal
isozymes of porcine P450arom also facilitated an examination of
stereo-specific loss of hydrogen from the ß orientation of positions
C1 and C2 on the steroid A ring of both
androgen substrates. This forms the basis of the widely used
radiometric procedure known as the tritiated water assay for aromatase
activity (33). The exact nature of the stereo-selective hydrogen loss,
chemical or biological, remains unclear and largely unexplored in other
species, except the rat (40). The data presented here suggest that the
tendency for selective loss of the ß-oriented hydrogen from the
C1 and C2 positions of androgen substrates
differed for the placental and gonadal isozymes, particularly in the
case of testosterone. Previous studies by Swinney et al.
(40) suggested that the rate of release of ß-hydrogen differed
between rat ovarian and human placental P450arom enzymes. Specifically,
less 3H2O was recovered during estrogen
synthesis from either [1ß-3H]- or
[1ß,2ß-3H]androstenedione by rat ovarian P450arom
than from human placental P450arom. Isotope effects were considered
unlikely, but the researchers were unable to determine whether the
differences reflected tissue (ovarian vs. placenta) or
isozyme (rat vs. human) effects. Our data on the percentage
of label appearing as 3H2O in cells transiently
expressing human P450arom correspond well with those reported by
Swinney et al. (40) for the four labels examined (90%
vs. 90% for [1ß-3H]androstenedione; 82%
vs. 85% for [1ß,2ß-3H]androstenedione;
61% vs. 49% for [1ß,2ß-3H]testosterone;
and 17% vs. 20% for
[1,2-3H]androstenedione). However, based on
P450arom isozymes expressed in the same cell background, the results of
the present study suggest that the P450 itself, whether species or
tissue derived, contributes to the observed differences in the rates of
ß-hydrogen loss. Thus, quantitative comparisons of aromatase activity
among species at least, if determined by tritiated water release,
should consider differential loss of label particularly if dual
1ß,2ß-3H-labeled tracers are used.
Our data suggest that the porcine placental isozyme is catalytically
very similar to human P450arom, but that the gonadal isozyme is more
distinct. These differences must involve important conformational
changes in the substrate binding pocket of the gonadal isozyme that
affect substrate orientation causing a tighter fit of the steroid
in the active site. Graham-Lorence et al. (44) hypothesized
that hydrogen abstraction and enolization occurred via a proton shuttle
mechanism involving aspartate 309, a highly conserved, acidic residue
that in structurally determined P450s projects into the pocket. This
residue was predicted to lie within 2A of the substrate C1
and C2. The stereo-specific loss of tritium from
[1ß-3H]androstenedione seen for both porcine isozymes
and human and rat P450arom (40) suggest that this substrate is probably
oriented with the ß face more exposed to the proton acceptor. The
less selective loss of tritium during aromatization of
[1ß,2ß-3H]testosterone suggests that this substrate
must be slightly rotated relative to the proton acceptor, exposing the
and ß labels more equally. The stereo-selectivity shown with the
-labeled substrates is consistent with the preferential removal of
hydrogen, which has a higher zero point energy than tritium. Whether
the characteristics of the stereo-selective loss of hydrogen during
aromatization correlates with other aspects of catalytic activity of
different isozymes of P450arom remains an interesting question.
However, it seems from these data and those reported by Swinney
et al. (40) that hydrogen abstraction and enolization of
androstenedione in both isozymes and of testosterone in the gonadal
isozyme are protein-facilitated reactions. Moreover, these reactions
might not necessarily proceed in the same way for androstenedione and
testosterone because of differences in orientation in the substrate
binding pocket that depend on the species- or tissue-specific form of
the P450arom in question.
Catalytic differences between the porcine placental and gonadal P450arom isozymes were also examined in kinetic experiments to estimate affinities for the C19 steroid substrates androstenedione and testosterone. Similar Km estimates were obtained by transient transfection and in stable expressing cell lines for androstenedione and testosterone in the case of the placental isozyme. Stable expression was necessary to obtain reliable estimates of Km for testosterone metabolism by the gonadal P450arom. Collectively, these data indicate that the two isozymes have similar affinities for androstenedione, but that the gonadal isozyme has a significantly lower affinity for testosterone than the placental P450arom. The apparent Km for androstenedione (77 and 104 nM for placental and gonadal isozymes, respectively) is in the range reported for human P450arom by numerous researchers (3, 40, 45, 46, 47). Apparent Km values reported for testosterone are more variable, sometimes similar to those for androstenedione (3, 47), slightly higher (46), or markedly higher (45), which may reflect different sources of enzymes or methods of purification. Our results suggest that there are functional differences in affinity for testosterone between the porcine placental and gonadal isozymes (33 vs. 116 nM, respectively) stably expressed in 293 cells, which, in the case of the placental isozyme, is also less than half the Km for androstenedione (77 nM). McPhaul et al. (4) reported slightly higher Km values derived from testosterone metabolism by chicken P450arom (150 nM) expressed in ovarian microsomes and for the recombinant enzyme expressed in vitro. Although the microsomal or in vitro environment may influence P450 activity (46), the differences noted between the porcine isozymes examined here are unlikely to suffer from such potential confounding factors.
The results of the present study also indicated that testosterone was aromatized more slowly than androstenedione, as suggested by the slower accumulation of estradiol and by estimates of the apparent Vmax from kinetic experiments. However, a valid comparison of catalytic rates depends directly on equal levels of P450arom expression in each cell line, which our data suggest may change with the stage of confluence. Although immunoblot analysis suggested that levels were generally similar in each cell line, this also assumes that the antiserum used recognizes the isozymes with equal affinity. In fact, Hinshelwood et al. (9) considered that a single amino acid change rendered bovine P450arom essentially undetectable by immunoblot analysis, consistent with the possibility that small changes in structure might have marked effects on epitope recognition. For this reason, we recently overexpressed each porcine isozyme in insect cells using a baculovirus vector. This has enabled estimates to be made of P450arom content in preparations of both isozymes by difference spectroscopy and analysis of androgen metabolism on an equal molar basis. Complete kinetic experiments using these preparations are in progress. However, preliminary results confirm that the placental isozyme of porcine P450arom is 3 times as active in aromatizing androgen as the gonadal isozyme and is more efficient in doing so at lower testosterone concentrations (Corbin, C. J., and A. J. Conley, unpublished observations). These data compare well with the 3- to 5-fold differences in catalytic rates observed in the stable cell lines, suggesting that P450arom expressions in cell lines are similar, but that differences in activity arise from the catalytic characteristics inherent in the P450 isozymes they express. It also suggests that there are no major differences in immunorecognition by the antisera used for immunoblot analysis. Collectively, the data suggest that the greatest differences in catalytic function between the porcine gonadal and placental P450arom isozymes relate to their metabolism of testosterone. Moreover, despite the potential for differences arising from expression levels, variations with confluence or peculiarities arising during the development of the cell lines specifically investigated, the evidence points strongly to P450arom isozymes themselves as the source of the functional divergence reported here.
Based on the data presented above, we conclude that the evolution of porcine P450arom expression in the placenta was associated with an increase in the efficiency of androgen metabolism, potentially involving both the affinity and rate of aromatization. These changes are consistent with the lower androgen concentrations likely to be presented to the enzyme expressed in utero compared with P450arom in the ovarian follicle (32, 48, 49). The apparent Km values typical of mammalian aromatases are considerably lower than those commonly reported for other steroid hydroxylases such as P450c17 (50), suggesting that P450arom may be adapted for expression in cells or tissues not also synthesizing androgen substrates (51). Equally, the greater affinity for testosterone may have important adaptive significance in terms of the physiological protection of female fetuses developing adjacent to males during sexual differentiation. It is perhaps noteworthy that there was no significant effect of the proximity of male fetuses on the development of females in pigs (52), whereas an effect is obvious in rodents (53), a species that lacks P450arom in the placenta (54). In other words, the lack of placental P450arom may leave the female fetuses of some mammalian species susceptible to androgenization in utero, a concept consistent with the pseudohermaphroditism observed in women suffering congenital P450arom deficiency (55, 56). The similarity between the porcine placental and human P450arom isozymes suggests that similar functional shifts to more efficient testosterone metabolism may have occurred in other species exhibiting placenta P450arom expression. In fact, the reported bovine CYP19 pseudogene (57) might represent the disappearance of an ancestral CYP19 gene encoding P450arom after evolution of expression in the bovine placenta (9). Further studies comparing P450arom function across species will add to our understanding of the roles of androgens and estrogens in vertebrate development and reproduction. To what degree this involves essential changes in regions of the P450 mediating substrate binding, membrane anchoring, redox partner association, or other physical features remains to be determined.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
Received June 8, 1999.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
-hydroxylase/C17–20 lyase and aromatase cytochrome P-450 in the
human ovary during the menstrual cycle. J Endocrinol 135:589–595-hydroxylase and 3ß-hydroxysteroid dehydrogenase in the
integration of gonadal and adrenal steroidogenesis via the 
4 and
5 pathways of steroidogenesis. Biol Reprod 56:789–799[CrossRef][Medline]
-3H, progesterone-4-14C and
androstenedione 4-14C by rat placental tissue.
Endocrinology 87:151–155This article has been cited by other articles:
 |
|
 |
|
  C.J. Corbin, T. Berger, J.J. Ford, C.E. Roselli, W. Sienkiewicz, B.C. Trainor, J.F. Roser, J.D. Vidal, N. Harada, and A.J. Conley Porcine Hypothalamic Aromatase Cytochrome P450: Isoform Characterization, Sex-Dependent Activity, Regional Expression, and Regulation by Enzyme Inhibition in Neonatal Boars Biol Reprod, August 1, 2009; 81(2): 388 - 395. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 I. Choi, J.-Y. Kim, E. J. Lee, Y. Y. Kim, C. S. Chung, J. Chang, N.-J. Choi, H.-J. Chung, and K.-H. Lee Ontogeny of Expression and Localization of Steroidogenic Enzymes in the Neonatal and Prepubertal Pig Testes J Androl, January 1, 2009; 30(1): 57 - 74. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
L. M. Wallock-Montelius, J. A. Villanueva, R. E. Chapin, A.J. Conley, H. P. Nguyen, B. N. Ames, and C. H. Halsted Chronic Ethanol Perturbs Testicular Folate Metabolism and Dietary Folate Deficiency Reduces Sex Hormone Levels in the Yucatan Micropig Biol Reprod, March 1, 2007; 76(3): 455 - 465. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
E.E. At-Taras, A.J. Conley, T. Berger, and J.F. Roser Reducing Estrogen Synthesis Does Not Affect Gonadotropin Secretion in the Developing Boar Biol Reprod, January 1, 2006; 74(1): 58 - 66. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. F. Skibola, J. D. Curry, C. VandeVoort, A. Conley, and M. T. Smith Brown Kelp Modulates Endocrine Hormones in Female Sprague-Dawley Rats and in Human Luteinized Granulosa Cells J. Nutr., February 1, 2005; 135(2): 296 - 300. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. J. Corbin, S. M. Mapes, J. Marcos, C. H. Shackleton, D. Morrow, S. Safe, T. Wise, J. J. Ford, and A. J. Conley Paralogues of Porcine Aromatase Cytochrome P450: A Novel Hydroxylase Activity Is Associated with the Survival of a Duplicated Gene Endocrinology, May 1, 2004; 145(5): 2157 - 2164. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
V. A. Lance, A. J. Conley, S. Mapes, C. Steven, and A. R. Place Does Alligator Testis Produce Estradiol? A Comparison of Ovarian and Testicular Aromatase Biol Reprod, October 1, 2003; 69(4): 1201 - 1207. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C.J. Corbin, F.M. Moran, J.D. Vidal, J.J. Ford, T. Wise, S.M. Mapes, V.C. Njar, A.M. Brodie, and A.J. Conley Biochemical Assessment of Limits to Estrogen Synthesis in Porcine Follicles Biol Reprod, August 1, 2003; 69(2): 390 - 397. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
F. M. Moran, J. J. Ford, C. J. Corbin, S. M. Mapes, V. C. Njar, A. M. Brodie, and A. J. Conley Regulation of Microsomal P450, Redox Partner Proteins, and Steroidogenesis in the Developing Testes of the Neonatal Pig Endocrinology, September 1, 2002; 143(9): 3361 - 3369. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. Conley, S. Mapes, C. J. Corbin, D. Greger, and S. Graham Structural Determinants of Aromatase Cytochrome P450 Inhibition in Substrate Recognition Site-1 Mol. Endocrinol., July 1, 2002; 16(7): 1456 - 1468. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. Kishida and G. V. Callard Distinct Cytochrome P450 Aromatase Isoforms in Zebrafish (Danio rerio) Brain and Ovary Are Differentially Programmed and Estrogen Regulated during Early Development Endocrinology, February 1, 2001; 142(2): 740 - 750. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
F.M. Morán, A.J. Conley, C.J. Corbin, E. Enan, C. VandeVoort, J.W. Overstreet, and B.L. Lasley 2,3,7,8-Tetrachlorodibenzo-p-Dioxin Decreases Estradiol Production Without Altering the Enzyme Activity of Cytochrome P450 Aromatase of Human Luteinized Granulosa Cells In Vitro Biol Reprod, April 1, 2000; 62(4): 1102 - 1108. [Abstract] [Full Text]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Endocrinology | Endocrine Reviews | J. Clin. End. & Metab. |
| Molecular Endocrinology | Recent Prog. Horm. Res. | All Endocrine Journals |
