Biochem. J. (2008) 414, 103–109 (Printed in Great Britain)
103
doi:10.1042/BJ20080397
Targeting human glutathione transferase A3-3 attenuates progesterone
production in human steroidogenic cells
Françoise RAFFALLI-MATHIEU*1 , Carolina ORRE*, Mats STRIDSBERG†, Maryam HANSSON EDALAT* and Bengt MANNERVIK*
*Department of Biochemistry and Organic Chemistry, Uppsala University, Biomedical Center, Husargatan 3, Box 576, SE-75123 Uppsala, Sweden, and †Department of Medical
Sciences, Clinical Chemistry, Uppsala University Hospital, SE-75185 Uppsala, Sweden
hGSTA3-3 (human Alpha-class glutathione transferase 3-3)
efficiently catalyses steroid 5 –4 double-bond isomerization
in vitro, using glutathione as a cofactor. This chemical transformation is an obligatory reaction in the biosynthesis of steroid hormones and follows the oxidation of 3β-hydroxysteroids catalysed
by 3β-HSD (3β-hydroxysteroid dehydrogenase). The isomerization has commonly been ascribed to a supplementary function of
3β-HSD. The present study is the first to provide evidence that
hGSTA3-3 contributes to this step in steroid hormone biosynthesis
in complex cellular systems. First, we find glutathione-dependent
5 –4 isomerase activity in whole-cell extracts prepared
from human steroidogenic cells. Secondly, effective inhibitors
of hGSTA3-3 dramatically decrease the conversion of 5 androstene-3,17-dione into 4 -androstene-3,17-dione in cell
lysates. Thirdly, we show that RNAi (RNA interference) targeting
hGSTA3-3 expression decreases by 30 % the forskolin-stimulated
production of the steroid hormone progesterone in a human
placental cell line. This effect is achieved at low concentrations
of two small interfering RNAs directed against distinct regions
of hGSTA3-3 mRNA, and is weaker in unstimulated cells, in
which hGSTA3-3 expression is low. The results concordantly
show that hGSTA3-3 makes a significant contribution to the
double-bond isomerization necessary for steroid hormone biosynthesis and thereby complements the indispensable 3β-hydroxysteroid oxidoreductase activity of 3β-HSD. The results indicate
that the lower isomerase activity of 3β-HSD is insufficient for
maximal rate of cellular sex hormone production and identify
hGSTA3-3 as a possible target for pharmaceutical intervention in
steroid hormone-dependent diseases.
INTRODUCTION
hGSTAs (human Alpha-class GSTs) have revealed that hGSTA33 catalyses the 5 –4 isomerization of 5 -AD (5 -androstene-3,
17-dione; a precursor of testosterone) and of 5 -PD (5 pregnene-3,20-dione; a precursor of progesterone) 230 and
25 times more efficiently respectively than 3β-HSD [8]. The
high steroid isomerase activity of hGSTA3-3 is unparalleled by
any other known human enzyme. Among members of the Alpha
class, hGSTA1-1 displays only 10 % of the hGSTA3-3 isomerase
activity towards 5 -AD, whereas hGSTA2-2 and hGSTA4-4 are
significantly less active than hGSTA1-1.
Immunohistochemical studies have shown that GSTAs are
selectively localized in the steroidogenic cells of the human
ovary [9]. Remarkably, RT–PCR (reverse transcription–PCR)
studies have revealed that, among the GSTAs, only hGSTA3-3
exhibits selective expression in steroidogenic tissues [8,10]. The
characteristic tissue distribution and in vitro enzymatic properties
of hGSTA3-3 strongly suggest a role for this enzyme in the
biosynthesis of steroid hormones in humans. However, direct
evidence for such a role in intact human cells is lacking.
Here, we address the question of the involvement of hGSTA33 in steroidogenesis in vivo, by studying the production of
steroid hormones in cells in which hGSTA3-3 activity or expression is suppressed. We identify potent inhibitors of the steroid
isomerase activity of recombinant hGSTA3-3 and design
siRNAs (small interfering RNAs) targeting specifically hGSTA33 mRNA. We demonstrate that these agents inhibit the ability
of human steroidogenic cells expressing hGSTA3-3 to perform
isomerization of 5 -AD and to produce progesterone. The present
The biosynthesis of steroid hormones in humans is a multistep
process involving several enzymes present in the endoplasmic
reticulum and in the cytosol [1]. The first step in the production
of all steroid hormones is the transfer of cholesterol from the
outer to the inner mitochondrial membrane in a process mediated
by the StAR (steroidogenic acute regulatory protein). Subsequently, cytochromes P450 and steroid dehydrogenases catalyse
diverse oxidation and isomerization reactions leading to the production of the active hormones. The 5 –4 double-bond isomerization in the steroid nucleus is an obligatory chemical
transformation along the biosynthetic pathway of all steroid
hormones. This reaction is generally ascribed to the enzyme
3β-HSD (3β-hydroxysteroid dehydrogenase), which exhibits
two different enzymatic activities [1]. First, 3β-HSD catalyses
the dehydrogenation of 5 -hydroxysteroids to 5 -ketosteroids.
Secondly, the 5 -ketosteroids are isomerized to 4 -ketosteroids.
3β-HSD is not the sole enzyme capable of isomerizing steroids.
It has been known for a long time that some GSTs (glutathione
transferases) can catalyse the glutathione-dependent doublebond isomerization of ketosteroids [2]. GSTs are a superfamily
of enzymes prominently involved in detoxication reactions,
via conjugation of the tripeptide glutathione with potentially
harmful electrophilic substances [3]. However, recent studies have
uncovered roles for GSTs in other biological processes, such as
regulation of signalling protein kinases [4–6] and biosynthesis
of prostaglandins [7]. In vitro studies using recombinant
Key words: Alpha-class glutathione transferase A (GSTA),
forskolin, human Alpha-class glutathione transferase 3-3
(hGSTA3-3), progesterone, steroid isomerase, steroidogenic cell.
Abbreviations used: 5 -AD, 5 -androstene-3,17-dione; CDNB, 1-chloro-2,4-dinitrobenzene; EA, ethacrynic acid; GST, glutathione transferase; GSTA,
Alpha-class GST; hGSTA3-3, human GSTA3-3; 3β-HSD, 3β-hydroxysteroid dehydrogenase; MTT, 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyl-2H -tetrazolium
bromide; RNAi, RNA interference; RT–PCR, reverse transcription–PCR; SF-1, steroidogenic factor-1; siRNA, small interfering RNA; TBTA, tributyltin acetate.
1
To whom correspondence should be addressed (email [email protected]).
c The Authors Journal compilation c 2008 Biochemical Society
104
R. Raffalli-Mathieu and others
study is the first to present evidence for the contribution of
hGSTA3-3 in the actual production of sex hormones in living
human cells.
EXPERIMENTAL
Chemicals
GSH, CDNB (1-chloro-2,4-dinitrobenzene), EA (ethacrynic
acid) and MTT [3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyl-2Htetrazolium bromide] were obtained from Sigma–Aldrich. 5 -AD
was from Steraloids (Newport, RI, U.S.A.) and TBTA (tributyltin
acetate) and tributyltin bromide were from ABCR (Karlsruhe,
Germany).
Recombinant enzymes
hGSTA3-3 was heterologously expressed in Escherichia coli BL21 and purified as described by Johansson and Mannervik [11].
hGSTs A1-1, M1-1, M2-2, M5-5, P1-1 and T1-1 were available
in our laboratory.
Human steroidogenic cells
H295R cells were obtained from Dr William E. Rainey (Medical
College of Georgia, Augusta, GA, U.S.A.), and JEG-3 cells were
obtained from the A.T.C.C. (Manassas, VA, U.S.A.).
Enzymatic activities
Conjugation of GSH to CDNB by the recombinant enzymes was
monitored under standard conditions [12]. Isomerization of 5 AD to 4 -AD was performed as previously described [8] in the
presence of 1 mM GSH and 0.1 mM 5 -AD, at 30 ◦C. The reaction
was monitored spectrophotometrically at 248 nm. The molar
absorption coefficient () for the 4 -AD product is 16 300
M−1 · cm−1 . Inhibition by EA was examined at concentrations
ranging between 0.5 and 50 μM. Concentrations of TBTA varying
between 0.025 and 2 μM were used. The measurements were
conducted in duplicate and repeated on two different occasions.
For analysis of isomerase activity in cell lysates (see below
for preparation of lysates), 15 μl of lysate (from 0.8 × 106 to
1.4 × 106 cells) was used for each measurement (performed in a
final volume of 1 ml). This volume was chosen to allow the highest
measurable activity with an acceptable background absorbance at
248 nm. The isomerization of 5 -AD varied linearly between 20
and 100 μM, and a concentration of 60 μM of 5 -AD was chosen
for inhibition experiments. The concentration of TBTA was
varied between 0.025 and 2 μM and the concentration of EA
was varied between 0.5 and 50 μM. DMSO was used as the solvent for CDNB, EA and TBTA, and was present at a final concentration of 2.5 % in the assay system.
Cell culture and treatments
H295R cells were cultured in DMEM (Dulbecco’s modified
Eagle’s medium)/F-12 Ham’s medium (Sigma), supplemented
with 5 % Nu-Serum (BD Biosciences) and the antibiotics
penicillin and streptomycin (Sigma). JEG-3 cells were cultured
in MEM (minimal essential medium) supplemented with 10 %
(v/v) fetal bovine serum, antibiotics, fungizone and non-essential
amino acids (all from Invitrogen). After 24 or 48 h, the culture
medium was saved for progesterone analysis and a cytotoxicity
test was performed as described below.
Preparation of cell extracts
H295R cell lysates were prepared by lysing cells at 70 % confluency in ice-cold lysis buffer containing 20 mM Hepes/KOH
c The Authors Journal compilation c 2008 Biochemical Society
(pH 7.9), 25 % (v/v) glycerol, 1.5 mM MgCl2 , 420 mM NaCl,
0.2 mM EDTA, 0.4 % Nonidet P40, 0.5 mM PMSF and
0.5 mM DTT (dithiothreitol) [13] [500 μl of buffer/(25–40) ×
106 cells]. The cell extract was centrifuged at 13 000 g for
15 min and the supernatant was immediately used for activity
measurements.
The cytotoxicity test
The MTT test was performed as described by Carmichael et al.
[14] with minor modifications. Briefly, the cells were exposed to
0.5 mg of MTT per ml of culture medium and incubated at 37 ◦C
for 1.5 h. The formazan crystals formed were dissolved in DMSO
and A540 was measured.
Progesterone determinations
Progesterone concentrations in the cell medium were measured
with an automated immunoassay system (Modular E170, Roche
Diagnostics) at the Department of Medical Sciences, Clinical
Chemistry, at Uppsala University Hospital. The analytical range
extended from 0.1 to 190 nM. The total coefficient of variation
was less than 5.5 % at all concentration levels. The amount of
progesterone produced by the cells was then corrected for the
number of viable cells as determined by the MTT assay (see
above).
siRNAs
Two siRNAs (small interfering RNAs) (si1 and si2) targeting
the coding region of GSTA3 mRNA (GenBank® accession no.
NM_000847, nt 206–226 and nt 464–484 respectively) were
used. Si1 sequence: antisense strand, 5 -ACUCCCAUCAUUUCUUAACTT-3 ; sense strand, 5 -GUUAAGAAAUGAUGGGAGUTT-3 . Si2 sequence: antisense strand, 5 -GGCAGGGAAAUAGCGACUUTT-3 ; sense strand, 5 -AAGUCGCUAUUUCCCUGCCTT-3 . Annealed si1 and si2, and a non-specific
siRNA (SilencerTM Negative Control no. 1) were from Ambion.
Transfections
siRNAs were administered to JEG-3 cells by two successive
reverse transfections performed with a 48 h interval between
them. In the first transfection, the siRNAs complexed
with LipofectamineTM 2000 (Invitrogen) according to the
manufacturer’s instructions were dispensed into six-well plates.
Exponentially growing JEG-3 cells were trypsinized, treated with
10 μM forskolin, and added to the complexes at a density of
2 × 105 cells per well. After 48 h, the cells were harvested and
submitted to a second reverse transfection. Cell media were
analysed for progesterone content and cell toxicity was assayed
48 h after the second transfection.
Semi-quantitative RT–PCR
Total RNA was isolated from JEG-3 cells using a RNeasy Mini
kit (Qiagen). Reverse transcription was performed with 2 μg
of RNA, using Enhanced Avian Reverse Transcriptase (Sigma)
and OligodT(23) (Sigma). Amplification was performed with Taq
polymerase (Fermentas). hGSTA3-specific PCR primers and
PCR conditions were as described by Morel et al. [10]. 3βHSD-specific primers were as designed by Blouin et al. [15].
Optimization was carried out to ensure linearity of the amount
of hGSTA3, 3β-HSD and actin PCR products formed. A 0.75 μl
portion of the reverse transcription product was chosen as the
template for the PCR reactions, and β-actin control primers [16]
were added after 17 cycles of amplification.
Human GSTA3-3 and steroidogenesis
Table 1
105
Inhibition of purified hGST isoenzymes by two organotins and EA
The effect of the inhibitors on the isomerization of 5 -AD and on the conjugation of glutathione
to CDNB by various recombinant human GSTs was assessed as described in the Experimental
section. ‘–’, not determined.
IC50 (μM)
Substrate
GST
TBTA
Triethyltin bromide
EA
5 -AD
A3-3
A1-1
A3-3
A1-1
A2-2
M1-1
M2-2
M5-5
P1-1
T1-1
0.09
0.04
0.0023
0.018
0.41
–
0.4
0.01
6.9
> 100
–
–
0.69
5.7
0.19
–
3.68
0.14
3.5
> 100
2.5
5
–
10*
–
1*
–
–
15*
–
CDNB
*From Hansson et al. [37].
Statistics
Results shown are means +
− S.D. The statistical analyses were
performed with a one-tailed unpaired Student’s t test.
Figure 1
GSTA expression and steroid isomerization in H295R cells
(A) Expression of GSTAs: mRNAs coding for hGSTs A1-1 (A1), A2-2 (A2), A3-3 (A3) and
A4-4 (A4) were reverse-transcribed and amplified by RT–PCR as described in the Experimental
section. The DNA size markers are indicated. The PCR fragment corresponding to full-length
hGSTA3 cDNA is indicated by an arrowhead. The bands below correspond to hGSTA3 splice
variants [8,10]. (B) Steroid isomerase activity in cell lysate. Crude cell lysate was prepared
and the isomerization of 5 -AD was monitored as described in the Experimental section, in
the presence or absence of cell lysate and with or without glutathione. Here, 100 % activity
represents the isomerization rate obtained in the presence of 15 μl of cell lysate and 1 mM
glutathione. (C) Importance of glutathione in the lysate-dependent isomerization of 5 -AD. The
pie chart is based on the results in (B).
RESULTS
Inhibition of the steroid 5 –4 isomerization catalysed by
hGSTA3-3
Organometallic compounds and the diuretic substance EA are
potent inhibitors of GST-catalysed reactions [17–19]. In order
to gain a deeper insight into the selectivity of the inhibitory
effect of these compounds for hGSTA3-3, we tested the ability
of EA and of two organometals to inhibit reactions catalysed
by purified recombinant hGSTA3-3 and by several other GSTs.
The conjugation of GSH to CDNB by recombinant hGSTs A11, A2-2, A3-3, M2-2 and P1-1 was monitored in the presence
of the inhibitors and the IC50 was determined. In parallel, the
effect of the substances on the steroid isomerization reaction
catalysed by hGSTA3-3 and hGSTA1-1 was evaluated. The results
are presented in Table 1. The observed IC50 values indicate that
organotin compounds tend to be stronger inhibitors of GSTAs than
most GSTs from other classes. The strongest inhibitory effect was
in the low nanomolar range and was observed for TBTA on the
conjugation of CDNB by hGSTA3-3. TBTA was a potent inhibitor
of the hGSTA3-3- and hGSTA1-1-catalysed steroid isomerization
as well. EA also inhibited this reaction efficiently, although with
an IC50 in the micromolar range. EA and TBTA were chosen for
further inhibition studies using a more complex biological system,
represented by crude lysates prepared from human steroidogenic
cells.
GSTA expression in H295R cells and 5 -AD isomerase activity in
cell lysates
As hGSTA3-3 is selectively expressed in steroid hormoneproducing organs in humans [8,10], we chose the adrenal cell
line H295R, known to possess the full enzyme complement
required for normal adrenal steroidogenesis, and to produce
significant amounts of testosterone [20,21]. We verified that
the cells secreted progesterone and testosterone as expected
(results not shown). We found by RT–PCR that hGSTA3-3 is
expressed in this cell line, as well as hGSTs A1-1, A2-2 and
A4-4 (Figure 1A). Two previously described splice variants of
Figure 2
cells
Inhibition of steroid isomerase activity in crude lysates of H295R
Inhibition of the 5 –4 isomerization of 5 -AD (60 μM) was analysed in the presence of EA
(0.5–50 μM) and TBTA (0.025–2 μM), as described in the Experimental section.
hGSTA3-3 were detected [8,10]. Crude protein extracts were
prepared from exponentially growing H295R cells and enzymatic
reactions were set up as in the assays with recombinant proteins,
except that 15 μl of cell extract was used instead of the
recombinant enzymes. 5 -AD isomerase activity was readily
detected in crude cell lysates in the presence of 1 mM glutathione. When glutathione was removed only 26 % of the
activity remained, whereas 12 % was detected when the lysate
was omitted (Figure 1B). Thus, in the presence of glutathione,
12 % of the isomerization was non-enzymatic, while glutathionedependent isomerization accounted for 74 % of the total activity
measured (Figure 1C).
Inhibition of 5 -AD isomerase activity in cell lysates
As shown in Figure 2, EA and TBTA efficiently inhibited 5 -AD
isomerase activity in crude cell lysates. IC50 values were found to
be 5 and 0.4 μM respectively, showing that TBTA is a stronger
inhibitor than EA in cell lysates, in accordance with the results
obtained with purified enzymes (see Table 1).
GSTA expression and progesterone production in JEG-3 cells
The placental cell line JEG-3 was chosen for RNAi (RNA
interference) studies. Unstimulated cells were found to express
c The Authors Journal compilation c 2008 Biochemical Society
106
R. Raffalli-Mathieu and others
Figure 4
Targeting of hGSTA3-3 mRNA by siRNA1
Total RNA was prepared 5, 48 or 48+24 h post-transfection. Here, ‘48+24 h’ refers to a double
transfection, where cells were first transfected during 48 h, after which they were trypsinized
and transfected a second time for 24 h (for further details, see the Experimental section).
An siRNA-targeting hGSTA3-3 mRNA (si1) and a control siRNA (siC) were used. The fragment
amplified from full-length hGSTA3 mRNA is indicated (GSTA3 FL). Two shorter splice variants of
hGSTA3-3 are detected below GSTA3 FL. The level of 3β-HSD mRNA in siC- and si1-transfected
cells is presented in the lower panel. Actin was used as an internal control.
and transfection and lasted for at least 48 h after transfection
(Figure 4). As expected, the siRNA targeting the exon 3–exon 4
junction did not affect the expression of splice variants lacking
exon 3 [8,10]. Of note, the silencing procedure did not affect the
expression of 3β-HSD mRNA (Figure 4).
Figure 3 Effect of forskolin on GSTA expression and progesterone
production in JEG-3 cells
Forskolin was used at a final concentration of 10 μM throughout the experiments.
(A) Increase in hGSTA3-3 mRNA level on treatment with forskolin (Fo) for 24 h, as assessed by
semi-quantitative RT–PCR. Control cells (C) received the vehicle DMSO alone. GSTA3 FL, PCR
product corresponding to full-length hGSTA3-3 transcript. The splice variants are indicated.
The lowest band is actin, as indicated. (B) Time-dependent effect of forskolin on the production
of progesterone by JEG-3 cells. The culture medium was recovered at the indicated times after
treatment with forskolin and analysed as described in the Experimental section. (C) Expression
of other hGSTAs assessed by semi-quantitative RT–PCR, in cells treated with forskolin (Fo)
or DMSO (C). The amplification of hGSTA1 mRNA from H295R cells performed in parallel is
presented as a control for the PCR reaction.
only minute amounts of hGSTA3-3. However, treatment with
forskolin, known to increase cellular cAMP levels and to stimulate
steroidogenesis [1], resulted in an increase in the hGSTA3-3
mRNA level (Figure 3A). This increase was highly reproducible
throughout all treatments with forskolin, and it encompassed also
hGSTA3-3 splice variants, as shown in Figure 3(A). JEG-3 cells
produce and release progesterone, which can be measured in
the culture medium. Treatment with 10 μM forskolin increased
progesterone production as expected (Figure 3B). The other
human GSTAs were essentially undetectable by RT–PCR in this
cell line, even after treatment with forskolin. (Figure 3C). Only
a very weak band corresponding to hGSTA4-4 was detected in
control and forskolin-treated cells.
Down-regulation of hGSTA3-3 by RNAi
We designed two alternative siRNAs targeting hGSTA3-3
mRNA. Transfection of either of the hGSTA3-3-specific siRNAs
concurrent with forskolin treatment resulted in the lack of
production of full-length hGSTA3-3 mRNA. In forskolin-treated
cells that received an unrelated control siRNA, the amount of
hGSTA3-3 mRNA increased as observed previously (Figure 4).
Thus the designed siRNAs selectively prevented the forskolindependent induction of hGSTA3-3 mRNA. The effect was
clearly visible already after 5 h of the combined induction
c The Authors Journal compilation c 2008 Biochemical Society
Attenuation of progesterone production in JEG-3 cells by RNAi
We next investigated whether siRNA-mediated knock-down of
hGSTA3-3 in forskolin-treated JEG-3 cells affected their ability to
synthesize progesterone. As shown in Figure 5, the production of
progesterone by JEG-3 cells in which hGSTA3-3 is silenced was
decreased by 30 %. Concentrations of siRNA down to 4 nM still
afforded a reduction of progesterone production (Figure 5A). Two
dissimilar hGSTA3-3-specific siRNAs both caused the attenuation
of progesterone production (Figure 5B). In cells not treated with
forskolin, and in which hGSTA3-3 expression therefore was very
low, silencing of hGSTA3-3 caused a smaller relative decrease in
progesterone secretion (10–15 %; Figure 5C).
DISCUSSION
In the present study, we report the first direct evidence for
the involvement of hGSTA3-3 in the biosynthesis of steroid
hormones in human cells. We used two different approaches to
interfere with hGSTA3-3 function in steroidogenic cells. The first
approach was based on the use of EA and TBTA, two inhibitors
of the steroid isomerization reaction catalysed by hGSTA3-3.
These compounds exhibited a strong inhibitory effect on 5 –
4 steroid isomerization in H295R cell lysates. EA and TBTA
are also potent inhibitors of hGSTA1-1 (Table 1), another GST
shown to exhibit steroid isomerase activity [8], although with a
lower efficiency than hGSTA3-3. It is therefore possible that the
inhibition of steroid isomerase activity observed in cell lysates
resulted from the inhibition of not only hGSTA3-3 but also,
to some extent, hGSTA1-1. 3β-HSD, which produces the 3keto-5 steroid substrate, also catalyses the subsequent 5 –
4 steroid isomerization, yet with a much lower efficiency
than hGSTA3-3 [8]. However, the activity of this enzyme was
previously reported to be only marginally affected by TBTA [22].
In addition, NADH, the essential activator of 3β-HSD [23], was
not added to our extracts. Because NADH is required for 3βHSD steroid isomerase activity, this enzyme is highly unlikely to
Human GSTA3-3 and steroidogenesis
Figure 5 Effect of hGSTA3-3 knockdown on progesterone production by
JEG-3 cells
(A) Importance of the amount of siRNA transfected. si1 (hGSTA3-3-specific) and siC (control
siRNA) were transfected (double transfection, 2 × 48 h) in JEG-3 cells, and progesterone (PG)
production was measured as the amount of progesterone released in culture medium during
48 h after the second transfection (for details, see the Experimental section). Normalized PG
production refers to the amount of progesterone produced corrected for the amount of viable cells
(measured by the MTT assay; see the Experimental section). (B) Similar effect of two different
siRNAs (si1 and si2) targeting hGSTA3 mRNA. The control siRNA (siC) was the same as in (A).
All siRNAs were transfected at a concentration of 20 nM. (C) Lack of effect of si1 in unstimulated
JEG-3 cells. Progesterone production was measured in siRNA-transfected cells treated with
forskolin (Fo), as in the experiments presented in (A) and (B), or the vehicle alone (DMSO) or
was left untreated (–) was measured. , P < 0.15; *P < 0.05; **P < 0.002; ***P < 0.001.
make a significant contribution to the 5 -AD isomerase activity
observed in our assay. Altogether, these results show that the
GST-dependent steroid isomerization is not restricted to in vitro
conditions, but is prominent in a complex biological system such
as crude cell lysates, where it can be suppressed by inhibitors of
GSTAs.
We next studied the effect of EA and TBTA on the capacity
of steroidogenic cells to produce steroid hormones. However,
EA and TBTA were found to be highly toxic to cultured
H295R cells (results not shown). We therefore chose another
human steroidogenic cell line, JEG-3, widely used for studies
on steroid hormone-related processes [24]. However, non-toxic
concentrations of EA and TBTA failed to diminish progesterone
production by cultured JEG-3 cells (results not shown). Possibly,
107
effects of TBTA on other steroidogenic enzymes [25–27] may
mask the consequences of inhibiting hGSTA3-3 on progesterone
production. Not much is known about the effect of EA
on steroidogenesis, but this electrophilic compound may act
unspecifically through its propensity to react with thiol groups.
The lack of specificity of these toxic inhibitors renders them
unsuitable for in vivo studies of the role of hGSTA3-3 in
steroidogenesis. Studies involving specific inhibitors of 3β-HSD
(such as trilostane or epostane) would not clarify the relative
contribution of hGSTA3-3, since they would block the dehydrogenase activity of 3β-HSD [28], thereby preventing synthesis of
the substrate for the isomerization reaction. To our knowledge, no
compound selectively inhibiting the steroid isomerase activity of
3β-HSD has been described.
We therefore chose the highly specific approach of knocking
down the hGSTA3 gene expression with RNAi. The resulting
attenuation of cellular progesterone production showed that
hGSTA3-3 is a significant contributor to progesterone synthesis in
JEG-3 cells. In the siRNA-treated cells, expression of the mRNA
coding for 3β-HSD, the enzyme to which steroid isomerization is
classically attributed [1], was not affected, and the transfection
conditions in the present study minimized the risk of offtarget effects. Moreover, we used cells where the most similar
mRNAs (hGSTA1, hGSTA2 and hGSTA4), which might be
potential off-targets, are undetectable by RT–PCR or present in
minute amounts. The concentration of siRNA transfected was low
throughout all experiments, and an attenuation of progesterone
production was still observable with a concentration of 4 nM
of siRNAs, indicating high specificity. Furthermore, we showed
that two siRNAs targeting different regions of hGSTA3 mRNA
caused a similar diminution of progesterone production. Finally,
a highly significant attenuation of progesterone secretion was
observed only in cells where hGSTA3-3 expression was induced
by forskolin. The marginal effect in untreated cells, in which
3β-HSD, but not GSTA3-3, is expressed at relatively high
levels [29], also shows that silencing does not affect 3β-HSD
function significantly. In the present study, the effect of hGSTA33 knockdown on progesterone production in JEG-3 cells was
limited to 30 %, which may be due to a compensatory effect of
forskolin, which induces 3β-HSD in this cell line [29]. Obviously,
the isomerase activity displayed by 3β-HSD could not be expected
to be inhibited by eliminating the contribution by hGSTA3-3.
However, our results clearly show that the isomerase activity of
3β-HSD is not sufficient for maximal rate of cellular sex hormone
production and identify hGSTA3-3 as a significant contributor.
It should be borne in mind that the relative contributions of 3βHSD and hGSTA3-3 to steroid isomerization may vary among
different cell types and under distinct physiological conditions.
The present findings therefore warrant further investigation of
the possible interplay between 3β-HSD and hGSTA3-3. Previous
work indicates that GSTAs are associated with intracellular
membranes [30,31], suggesting that hGSTA3-3 and 3β-HSD
have similar subcellular localizations, which should facilitate
functional co-operation between the two enzymes.
In accord with a role for hGSTA3-3 in steroidogenesis, we
note that the hGSTA3 gene possesses several features of wellknown steroidogenic enzymes. We describe for the first time
the induction of hGSTA3-3 by forskolin, known to increase the
level of intracellular cAMP, an essential inducer of steroidogenic
enzymes [1]. Moreover, examination of the promoter region of the
hGSTA3 gene reveals three putative SF-1 (steroidogenic factor1)-binding sites, which are not present in the corresponding
promoter regions of the closely related hGSTA1 and hGSTA2
genes (Figure 6). The SF-1 transcription factor is involved in the
up-regulation of most steroidogenic enzymes [1]. Whether these
c The Authors Journal compilation c 2008 Biochemical Society
108
R. Raffalli-Mathieu and others
Figure 6 Presence of putative SF-1-binding sites in the promoter region of
the hGSTA3 gene
The promoters of hGSTA1, hGSTA2 and hGSTA3 genes (retrieved from a BLAT
search; http://genome.ucsc.edu/) were aligned using the ClustalW software (http://www.ebi.
ac.uk/clustalw/). The three putative SF-1-binding sites found in the hGSTA3 promoter are
presented in boldface. Nucleotide assignments are indicated in parentheses (‘0’ denotes the first
transcribed nucleotide). *, deletions; ‘-’, identical bases.
sites in the hGSTA3 promoter region are functional in different
human steroidogenic tissues remains to be established and is the
subject of our ongoing studies.
Several human endocrine disorders, such as the polycystic
ovary syndrome [32], congenital adrenal hyperplasia [33] and
Cushing’s syndrome [34], are characterized by excessive steroid
hormone production. In addition, sex steroid hormones stimulate
neoplastic cell growth in a number of human malignancies,
such as prostate and breast cancers [35]. Of note, increased
expression of GSTAs has been reported in benign adrenocortical
adenoma from patients with Cushing’s syndrome [36]. It would
be worthwhile to study whether hGSTA3-3 is increased in these
tissues and in other pathological conditions involving increased
steroid hormone production. Further studies are warranted in order
to fully comprehend the importance of hGSTA3-3 in human
steroidogenesis. However, the present study provides not only
evidence for a role of hGSTA3-3 in this important biological
process, but also proffers the enzyme as a new possible target to
reduce steroid hormone production in human cells.
This work was supported by the Swedish Research Council, the Swedish Cancer Society
and the Åke Wibergs Stiftelse. C.O. was a recipient of stipends from the Sven and Lilly
Lawski Fund. The expert technical assistance of Ms Birgit Olin (Department of Biochemistry and Organic Chemistry, Uppsala University) is acknowledged. We thank Ms
Ann Musungu and Ms Qing Ma for assistance with inhibition studies and Dr William
E. Rainey for kindly providing the H295R cells. We also thank Ms Natalia Fedulova, Ms
Emilia Larsson and Ms Fatima Burovic (all of the Department of Biochemistry and Organic
Chemistry, Uppsala University) for valuable comments on this paper.
REFERENCES
1 Payne, A. H. and Hales, D. B. (2004) Overview of steroidogenic enzymes in the pathway
from cholesterol to active steroid hormones. Endocr. Rev. 25, 947–970
2 Benson, A. M. and Talalay, P. (1976) Role of reduced glutathione in the 5 -3-ketosteroid
isomerase reaction of liver. Biochem. Biophys. Res. Commun. 69, 1073–1079
3 Josephy, P. D. and Mannervik, B. (2006) Molecular Toxicology, 2nd edn, pp. 333–401,
Oxford University Press, New York
4 Adler, V., Yin, Z., Fuchs, S. Y., Benezra, M., Rosario, L., Tew, K. D., Pincus, M. R.,
Sardana, M., Henderson, C. J., Wolf, C. R. et al. (1999) Regulation of JNK signaling by
GSTp. EMBO J. 18, 1321–1334
5 Romero, L., Andrews, K., Ng, L., O’Rourke, K., Maslen, A. and Kirby, G. (2006) Human
GST A1-1 reduces c-Jun N-terminal kinase signaling and apoptosis in Caco-2 cells.
Biochem. J. 400, 135–141
6 Zhao, X., Fan, Y., Shen, J., Wu, Y. and Yin, Z. (2006) Human glutathione S-transferase P1
suppresses MEKK1-mediated apoptosis by regulating MEKK1 kinase activity in HEK293
cells. Mol. Cell 21, 395–400
7 Hayes, J. D., Flanagan, J. U. and Jowsey, I. R. (2005) Glutathione transferases. Annu. Rev.
Pharmacol. Toxicol. 45, 51–88
c The Authors Journal compilation c 2008 Biochemical Society
8 Johansson, A.-S. and Mannervik, B. (2001) Human glutathione transferase A3-3, a highly
efficient catalyst of double-bond isomerization in the biosynthetic pathway of steroid
hormones. J. Biol. Chem. 276, 33061–33065
9 Tiltman, A. J. and Haffajee, Z. (1999) Distribution of glutathione S-transferases in the
human ovary: an immunohistochemical study. Gynecol. Obstet. Invest. 47, 247–250
10 Morel, F., Rauch, C., Coles, B., Le Ferrec, E. and Guillouzo, A. (2002) The human
glutathione transferase alpha locus: genomic organization of the gene cluster and
functional characterization of the genetic polymorphism in the hGSTA1 promoter.
Pharmacogenetics 12, 277–286
11 Johansson, A.-S. and Mannervik, B. (2002) Active-site residues governing high steroid
isomerase activity in human glutathione transferase A3-3. J. Biol. Chem. 277,
16648–16654
12 Mannervik, B. and Widersten, M. (1995) Human glutathione transferases: classification,
tissue distribution, structure and functional properties. In Advances in Drug Metabolism
in Man (Pacifici, G. M. and Fracchia, G. N., eds.), pp. 407–459, The European
Commission, Luxembourg
13 Dignam, J. D., Lebovitz, R. M. and Roeder, R. G. (1983) Accurate transcription initiation
by RNA polymerase II in a soluble extract from isolated mammalian nuclei. Nucleic Acids
Res. 11, 1475–1489
14 Carmichael, J., DeGraff, W. G., Gazdar, A. F., Minna, J. D. and Mitchell, J. B. (1987)
Evaluation of a tetrazolium based semiautomated colorimetric assay: assessment of
chemosensitivity testing. Cancer Res. 47, 936–942
15 Blouin, K., Richard, C., Brochu, G., Hould, F.-S., Lebel, S., Marceau, S., Biron, S.,
Luu-The, V. and Tchernof, A. (2006) Androgen inactivation and steroid-converting enzyme
expression in abdominal adipose tissue in men. J. Endocrinol. 191, 637–649
16 Strassburg, C. P., Oldhafer, K., Manns, M. P. and Tukey, R. H. (1997) Differential
expression of the UGT1A locus in human liver, biliary, and gastric tissue: identification of
UGT1A7 and UGT1A10 transcripts in extrahepatic tissue. Mol. Pharmacol. 52, 212–220
17 Henry, R. A. and Byington, K. H. (1976) Inhibition of glutathione-S-aryltransferase from
rat liver by organogermanium, lead and tin compounds. Biochem. Pharmacol. 25,
2291–2295
18 Yalçin, S., Jensson, H. and Mannervik, B. (1983) A set of inhibitors for discrimination
between the basic isozymes of glutathione transferase in rat liver. Biochem. Biophys.
Res. Commun. 114, 829–834
19 Mannervik, B. and Danielson, U. H. (1988) Glutathione transferases – structure and
catalytic activity. CRC Crit. Rev. Biochem. 23, 283–337
20 Gazdar, A. F., Oie, H. K., Shackleton, C. H., Chen, T. R., Triche, T. J., Myers, C. E.,
Chrousos, G. P., Brennan, M. F., Stein, C. A. and La Rocca, R. V. (1990) Establishment
and characterization of a human adrenocortical carcinoma cell line that expresses multiple
pathways of steroid biosynthesis. Cancer Res. 50, 5488–5496
21 Rainey, W. E., Saner, K. and Schimmer, B. P. (2004) Adrenocortical cell lines.
Mol. Cell. Endocrinol. 228, 23–38
22 Heidrich, D. D., Steckelbroeck, S. and Klingmuller, D. (2001) Inhibition of human
cytochrome P450 aromatase activity by butyltins. Steroids 66, 763–769
23 Thomas, J. L., Myers, R. P. and Strickler, R. C. (1989) Human placental 3
β-hydroxy-5-ene-steroid dehydrogenase and steroid 5–4-ene-isomerase: purification
from mitochondria and kinetic profiles, biophysical characterization of the purified
mitochondrial and microsomal enzymes. J. Steroid Biochem. 33, 209–217
24 Ringler, G. E. and Strauss, III, J. F. (1990) In vitro systems for the study of human
placental endocrine function. Endocr. Rev. 11, 105–123
25 Nakanishi, T., Kohroki, J., Suzuki, S., Ishizaki, J., Hiromori, Y., Takasuga, S., Itoh, N.,
Watanabe, Y., Utoguchi, N. and Tanaka, K. (2002) Trialkyltin compounds enhance human
CG secretion and aromatase activity in human placental choriocarcinoma cells. J. Clin.
Endocrinol. Metab. 87, 2830–2837
26 Nakanishi, T., Hiromori, Y., Yokoyama, H., Koyanagi, M., Itoh, N., Nishikawa, J. and
Tanaka, K. (2006) Organotin compounds enhance 17β-hydroxysteroid dehydrogenase
type I activity in human choriocarcinoma JAr cells: potential promotion of 17β-estradiol
biosynthesis in human placenta. Biochem. Pharmacol. 71, 1349–1357
27 Doering, D. D., Steckelbroeck, S., Doering, T. and Klingmüller, D. (2002) Effects of
butyltins on human 5α-reductase type 1 and type 2 activity. Steroids 67, 859–867
28 Thomas, J. L., Boswell, E. L., Scaccia, L. A., Pletnev, V. and Umland, T. C. (2005)
Identification of key amino acids responsible for the substantially higher affinities of
human type 1 3β-hydroxysteroid dehydrogenase/isomerase (3β-HSD1) for substrates,
coenzymes, and inhibitors relative to human 3β-HSD2. J. Biol. Chem. 280,
21321–21328
29 Tremblay, Y. and Beaudoin, C. (1993) Regulation of 3β-hydroxysteroid dehydrogenase
and 17β-hydroxysteroid dehydrogenase messenger ribonucleic acid levels by cyclic
adenosine 3 ,5 -monophosphate and phorbol myristate acetate in human
choriocarcinoma cells. Mol. Endocrinol. 7, 355–364
30 Prabhu, K. S., Reddy, P. V., Jones, E. C., Liken, A. D. and Reddy, C. C. (2004)
Characterization of a class alpha glutathione-S-transferase with glutathione peroxidase
activity in human liver microsomes. Arch. Biochem. Biophys. 424, 72–80
Human GSTA3-3 and steroidogenesis
31 Yang, Y., Cheng, J.-Z., Singhal, S. S., Saini, M., Pandya, U., Awasthi, S. and Awasthi,
Y. C. (2001) Role of glutathione S-transferases in protection against lipid peroxidation.
Overexpression of hGSTA2-2 in K562 cells protects against hydrogen peroxide-induced
apoptosis and inhibits JNK and caspase 3 activation. J. Biol. Chem. 276, 19220–19230
32 Chang, R. J. (2004) A practical approach to the diagnosis of polycystic ovary syndrome.
Am. J. Obstet. Gynecol. 191, 713–717
33 White, P. C. (2001) Congenital adrenal hyperplasias. Best Pract. Res. Clin. Endocrinol.
Metab. 15, 17–41
34 Newell-Price, J., Bertagna, X., Grossman, A. B. and Nieman, L. K. (2006) Cushing’s
syndrome. Lancet 367, 1605–1617
109
35 Henderson, B. E. and Feigelson, H. S. (2000) Hormonal carcinogenesis. Carcinogenesis
21, 427–433
36 Sarkar, D., Imai, T., Kambe, F., Shibata, A., Ohmori, S., Hayasaka, S., Funahashi, H. and
Seo, H. (2001) Overexpression of glutathione-S-transferase A1 in benign adrenocortical
adenomas from patients with Cushing’s syndrome. J. Clin. Endocrinol. Metab. 86,
1653–1659
37 Hansson, J., Berhane, K., Castro, V. M., Jungnelius, U., Mannervik, B. and Ringborg, U.
(1991) Sensitization of human melanoma cells to the cytotoxic effect of melphalan
by the glutathione transferase inhibitor ethacrynic acid. Cancer Res. 51,
94–98
Received 20 February 2008/10 April 2008; accepted 22 April 2008
Published as BJ Immediate Publication 22 April 2008, doi:10.1042/BJ20080397
c The Authors Journal compilation c 2008 Biochemical Society