0021-972X/98/$03.00/0
Journal of Clinical Endocrinology and Metabolism
Copyright © 1998 by The Endocrine Society
Vol. 83, No. 10
Printed in U.S.A.
Expression of 5a-Reductase in the Human Temporal
Lobe of Children and Adults*
BIRGIT STOFFEL-WAGNER, MATTHIAS WATZKA, STEPHAN STECKELBROECK,
LUCIA WICKERT, JOHANNES SCHRAMM, GABRIELA ROMALO,
DIETRICH KLINGMÜLLER, AND HANS-UDO SCHWEIKERT
Departments of Clinical Biochemistry (B.S., M.W., S.S., L.W., D.K.), Neurosurgery (J.S.), and Internal
Medicine (G.R., H.-U.S.), University of Bonn, Bonn, Germany
ABSTRACT
Androgens exert important biological effects on the brain, and
5a-reductase plays a crucial role in androgen metabolism. Therefore,
we investigated the expression of the two isozymes of 5a-reductase in
the human temporal lobe to determine the predominant isoform and
to elucidate the existence of possible sex differences and differences
between children and adults. We studied biopsy materials from the
temporal lobe of 34 women, 32 men, and 12 children. Quantification
of 5a-reductase 1 and 2 messenger ribonucleic acid (mRNA) was
achieved by competitive RT-PCR. 5a-Reductase activity was determined in tissue homogenates using [1,2-3H]androstenedione as the
substrate. Only 5a-reductase 1 mRNA was expressed in human temporal lobe tissue; 5a-reductase 2 mRNA was not expressed. 5a-
A
Reductase 1 mRNA concentrations did not differ significantly in the
cerebral cortex of women [25.9 6 7.9 arbitrary units (aU); mean 6
SEM] and men (20.4 6 2.8 aU) or in the cerebral cortex (23.3 6 4.4 aU)
and the subcortical white matter of adults (32.6 6 5.6 aU), but they
were significantly higher in the cerebral cortex of adults than in that
of children (6.4 6 2.3 aU; P , 0.005). The apparent Km of 5a-reduction
did not show significant differences between the two sexes. In conclusion, 5a-reductase 1 mRNA is expressed in the temporal lobe of
children and adults, but 5a-reductase 2 mRNA is not. 5a-Reductase
1 mRNA concentrations did not differ significantly in the sexes, but
they were significantly higher in specimens of adults than in those of
children. (J Clin Endocrinol Metab 83: 3636 –3642, 1998)
NDROGENS exert important biological effects on the
brain either directly or after 5a-reduction or aromatization (1– 4). Specific receptors for androgens have been
identified in several regions of the brain, through which
androgens could effect a genomic response (5).
5a-Reduction represents a major route of D4-androgen metabolism. 5a-Reductase (EC 1.3.99.5) uses NADPH to reduce
the double bound of a variety of steroid substrates with
generalized 3-oxo-D-4,5 structures (6). Recent cloning and
expression studies reported the isolation of complementary
DNAs (cDNAs) for two different isozymes (types 1 and 2) of
5a-reductase in rat as well as human tissues (7, 8). In addition
to biochemical and pharmacological differences, the type 1
and type 2 messenger ribonucleic acids (mRNAs) are differentially expressed in human tissues. 5a-Reductase 2 is the
predominant isoform found in male accessory sex organs,
whereas 5a-reductase 1 is present in tissues such as liver and
nongenital skin (9).
5a-Reductase activity has been demonstrated in neural
tissue from various animal species and human fetuses (1, 3,
10 –12). To date, there is little information on the androgen
metabolism in the human brain at different ages. Systematic
studies in human brain tissue are lacking. Although 5areductase enzymatic activity has been studied in only a few
frontal lobe and temporal lobe specimens of adults (13, 14),
5a-reductase has not yet been studied at the molecular level
in cortical tissue from children and adults. Only one study
reported 5a-reductase 1 expression in a few human cerebellum, hypothalamus, and pons tissue specimens that were
collected postmortem (9).
The cloning of 5a-reductase 1 and 2 cDNAs has enabled
this investigation of the isozyme expression of 5a-reductase.
It was designed to investigate the expression of 5a-reductase
isozymes in the human temporal lobe in a large number of
specimens from children and adults to determine the predominant isoform and to elucidate the existence of possible
sex differences and differences between children and adults.
To extend and confirm the results obtained in the mRNA
quantification experiments, 5a-reductase enzyme activity
was also determined.
Received February 4, 1998. Revision received June 12, 1998. Accepted
June 18, 1998.
Address all correspondence and requests for reprints to: Prof. D.
Klingmüller, Institut für Klinische Biochemie, Universität Bonn, Sigmund-Freud-Strasse 25, D-53105 Bonn, Germany.
* This work was supported by a grant from the Deutsche Forschungsgemeinschaft (Kl 524/4 –1). Presented in part at the 10th International
Congress of Endocrinology, San Francisco, CA, 1996 (Abstract P3-474).
[1,2-3H]Androstenedione (42 Ci/mmol) was purchased from New
England Nuclear Corp. (Dreieich, Germany). It was purified by thin
layer chromatography to assure a purity greater than 98%. Nonradioactive steroids were purchased from Steraloids, Inc. (Wilton, NH) or
Sigma Chemical Co. (Deisenhofen, Germany). NADPH, Taq polymerase,
and ribonuclease (RNase)-free deoxyribonuclease I (DNase I) were purchased from Boehringer Mannheim (Mannheim, Germany). Trizol reagent and the Superscript II preamplification system were obtained from
Subjects and Methods
Subjects
Biopsy materials removed at neurosurgery from 34 women (32.5 6
1.3 yr; mean 6 sem), 32 men (34.8 6 1.6 yr), and 12 children (8 6 1.4 yr)
with temporal lobe epilepsy undergoing partial temporal lobe resection
were used.
Steroids and reagents
3636
5a-REDUCTASE IN THE HUMAN TEMPORAL LOBE
Life Technologies (Paisley, UK). The pCR-script cloning kit and the RNA
in vitro transcription kit were purchased from Stratagene (La Jolla, CA).
The QIAquick PCR purification kit and the RNeasy total RNA kit were
obtained from Qiagen (Hilden, Germany). Primers were obtained from
Genosys (Cambridge, UK) or PE Applied Biosystems (Weiterstadt, Germany; Table 1).
Tissues
Temporal lobe biopsy materials were separated into cortex and subcortical white matter by inspection, transferred into liquid nitrogen
immediately after removal, and stored at 280 C. Cortex tissue specimens
were obtained from 19 women, 16 men, and 9 children; white matter
tissue specimens were obtained from 6 women, 7 men, and 1 child; and
both cerebral cortex and white matter tissue specimens were obtained
from 9 women, 9 men, and 1 child, respectively.
Liver tissues were obtained in a transplantation program from biopsies to exclude liver diseases from the Department of Surgery, University of Bonn (Bonn, Germany), and prostate tissues were obtained
from the Department of Urology, Waldkrankenhaus Bonn (Bonn, Germany). Tissues were transferred to liquid nitrogen immediately after
removal and stored at 280 C.
The study was approved by the local ethics committee, and informed
consent was obtained from all tissue donors.
mRNA quantification
mRNAs of 5a-reductase 1 and 2 were quantified with only a few
modifications according to a nested competitive RT-PCR protocol previously described (15).
Total RNA was extracted from 25–50 mg tissue using the Trizol
reagent. Traces of DNA were removed by treatment with RNase-free
DNase I, followed by a second RNA extraction. RNA was taken up in
RNase-free H2O and quantified by its spectrophotometric absorption at
260 nm.
Competitive RNA standards were prepared by overlap extension
mutagenesis of 5a-reductase 1 and glyceraldehyde-3-phosphate dehydrogenase (GAPDH) or by single step mutagenesis of 5a-reductase 2,
resulting in the loss of 7, 10, and 11 bp for 5a-reductase 1, GAPDH, and
5a-reductase 2, respectively, as previously described [5a-reductase
1/GAPDH (15, 16), 5a-reductase 2 (17)]. The mutant cDNAs (5a-reductase 1/GAPDH) were cloned with the pCR-script cloning kit. From these
plasmids, cDNA templates were amplified using primer pairs spanning
the mutagenized cDNA fragment and the T7 promoter region (T7 primer
and GAPDH reverse primer/T7 primer and 5a-reductase 1 reverse
primer; Table 1). Mutant 5a-reductase 2 PCR products were produced
by a PCR with a 59-primer containing the T7 promotor and 5a-reductase
3637
2 59-sequence and a 39-primer containing the 5a-reductase 2 39-sequence
with a deletion of 11 bp. These templates were cleaned using the QIAquick PCR purification kit and used to generate standard RNA by in vitro
transcription. Successful mutagenesis was confirmed by sequencing on
a semiautomated sequencer (373A, PE Applied Biosystems, Foster City,
CA). RNA in vitro transcription was performed using an RNA in vitro
transcription kit with T7 polymerase; cDNA templates were removed by
treatment with RNase-free DNase I (1 U/mg template). Standard RNA
was extracted with the RNeasy total RNA kit, and its concentration was
measured spectrophotometrically.
To estimate the amount of standard RNA required for quantification
of individual RNA samples, 4 –10 RNA samples of the respective tissue
groups were pooled. To aliquots of these mixtures containing 250 ng
RNA each, defined amounts of standard RNAs were added. Serial
dilutions ranged from 500 pg to 5 attograms (ag) for GAPDH and from
100 pg to 1 ag for 5a-reductase 1 and 2. Each mixture containing the
respective amount of RNA standard, and patient RNA was reverse
transcribed followed by PCR amplification. The optimal titration point
was defined as the concentration of standard RNA at which PCR products yielded signals of comparable intensity for standard and native
RNA (Fig. 1). A stock solution was prepared containing standard RNAs
for 5a-reductase 1, 5a-reductase 2, and GAPDH at the optimal titration
point. The concentration of this stock solution was selected in a way that
1 mL stock was sufficient for the RT of 250 ng total RNA. RT was
performed at 42 C for 60 min using 100 U Superscript II (Superscript
preamplification system). The resulting cDNA was diluted 20-fold with
water, and PCR was performed in a final volume of 20 mL containing
2 mL diluted cDNA, 10 mmol/L Tris-HCl (pH 8.3), 40 mmol/L KCl, 1.5
mmol/L MgCl2, 200 mmol/L of each deoxy-NTP, 0.5 U Taq polymerase,
and 4 pmol of each primer (Table 1). One primer of the primer pairs used
for GAPDH PCR or nested PCR (5a-reductase 1 and 2) was labeled with
fluorescent dyes. PCR amplification was carried out in microtiter plates
in a Unoblock (Biometra, Gottingen, Germany). Initial denaturation at
94 C for 4 min was followed by 32 (GAPDH) or 35 (5a-reductase 1 and
2) PCR cycles. Cycling conditions were 94 C for 35 s, 55 C for 50 s, and
72 C for 90 s. A final extension step of 5 min at 72 C was used. Nested
PCR of 5a-reductase 1 and 2 was performed under the same conditions.
Fluorescently labeled PCR products were separated on 6% denaturing acrylamide gels [50% (wt/wt) urea, 19:1 acrylamide-bisacrylamide,
and 1 3 TBE] and analyzed. Peak areas were calculated with the Genescan program (PE Applied Biosystems, version 1.2.1). The ratio of
native PCR product to standard PCR product was used for the differential determination of gene expression. Initial differences in the
amounts of total RNA that were subjected to RT were corrected by
calculating the ratios of native GAPDH PCR products to standard
GAPDH PCR products.
TABLE 1. Primers used for amplification
GAPDH f clo
GAPDH r clo
GAPDH f
GAPDH r
GAPDH f mut
GAPDH r mut
5aRed1 f
5aRed1 r
5aRed1 f nested
5aRed1 r nested
5aRed1 f mut
5aRed1 r mut
5aRed2 f
5aRed2 r
5aRed2 f nested
5aRed2 r nested
5aRed2 T7 f mut
5aRed2 r mut
T7
TCTCCAGAACATCATCCCTG
TGGGCCATGTGGTCCACCAC
TAMRA TGCCAAGGCTGTGGGCAAGG
GCTTCACCACCTTCTTGATG
GTGGACCTGACCTGCAACCTGCCAAATATGATGAC
CATATTTGGCAGGTTGCAGGTCAGGTCCACCAACTG
ACGGAGTAAGCTGCTCTGCC
TTATATTGATAACAGGTACAGG
TAMRA TGTGCTCTGGTCTGACATGG
TCATTCTTTACACTACAAGGG
CTCTGTCTTCTAGTTAGTTTTTTTGTTCTGTTCCCC
AGAACAAAAAAACTAACTAGAAGACAGAGAGAAACC
CTTCTGCGTACATTACTTCC
CCCAAGCTAAACCGTATGTC
TAMRA TACTCACTGCTCAATCGAGG
CCCAAGCTAAACCGTATGTC
GGATCCTAATACGACTCACTATAGGGAGCTTCTGCGTACATTACTTCC
CCCAAGCTAAACCGTATGTCCATCAGGGTATTCAGCACAG
GTAAAACGACGGCCAGTG
5aRed, 5a-Reductase; f, forward primer; r, reverse primer; clo, primer used for cloning and overlap extension mutagenesis; mut, primer used
for overlap extension mutagenesis or single step mutagenesis; T7 f mut; primer used for single step mutagenesis; abbreviation for fluorescent
dye: TAMRA, N,N,N9,N9-tetramethyl-6-carboxyrhodamine.
3638
STOFFEL-WAGNER ET AL.
JCE & M • 1998
Vol 83 • No 10
FIG. 1. Titration of RNA standards for 5a-reductase 1 and 2. Nested PCR was performed from total RNA of liver tissue (5a-reductase 1; a),
prostate tissue (5a-reductase 2; b), and temporal lobe tissue [5a-reductase 1 (c) and 5a-reductase 2 (d)]. Each lane corresponds to cDNA reversely
cotranscribed from 250 ng total RNA with decreasing standard RNA concentrations of either 5a-reductase 1 or 5a-reductase 2. The amounts
of standard RNAs were, from left to right, 100 pg, 10 pg, 1 pg, 100 fg, 10 fg, 1 fg, 100 ag, 10 ag, and 1 ag. The optimal titration point for 5a-reductase
1 standard RNA is 10 pg in liver tissue and 100 fg in temporal lobe tissue for 250 ng total RNA each. For 5a-reductase 2 standard RNA, the
optimal titration point is 1 pg in prostate tissue and below 1 ag in temporal lobe tissue for 250 ng total RNA each.
Determination of 5a-reductase activity
Statistical analysis
5a-Reductase activity in the biopsy materials was determined with
the following modifications according to methods previously described
(18, 19). In brief, samples were homogenized in ice-cold 10 mmol/L
Tris-chloride buffer (pH 7.4), and 1 mmol/L ethylenediamine tetraacetate with a Douncer homogenizer (Kontes Co., Vineland, NJ). Either
these homogenates were used, or a crude nuclear fraction and a crude
supernatant containing the microsomes were prepared by centrifuging
the homogenates at 1000 3 g for 15 min. The precipitate was then
suspended in the same buffer and rehomogenized to obtain the presumable nuclear fraction. The supernatant containing the microsomes
was not diluted further. 5a-Reductase activity was determined by incubation with [1,2-3H]androstenedione as substrate. Standard assays for
kinetic studies were carried out in triplicate. Standard assays contained
the 1,2-3H-labeled androstenedione at concentrations varying from 0.05–
3.5 mmol/L, 3 mmol/L NADPH, 0.08 mol/L Tris citrate (pH 7.5), and
5 mmol/L MgCl2 and the brain homogenates in a final volume of 200
mL. Incubations, separation of the 5a-androstanes by thin layer chromatography, and calculations of 5a-reductase activity rates were performed as previously described (18 –20).
The pH optimum of 5a-reduction was determined in cortex homogenates using a procedure similar to that described for 5a-reductase
activity, except that a series of buffers (0.08 mol/L Tris citrate) with
different pH values (4.0 –9.5) were used instead of a single buffer. These
experiments were carried out in duplicate.
Results were calculated as the mean 6 sem. The statistical difference
between groups was calculated using the Mann-Whitney U test. P , 0.05
was considered statistically significant.
Results
Determination of titration points for 5a-reductase 1 and 2
mRNA in reference tissues and temporal lobe tissue
To validate our mRNA quantification protocol, we determined the expression of 5a-reductase 1 mRNA in liver tissue
and the expression of 5a-reductase 2 in prostate tissue. The
use of competitive RT-PCR requires the amount of standard
RNA that yields a signal of approximately equal density
when coamplified with total RNA. The optimal titration
point for liver tissue was 10 pg standard RNA for 5a-reductase 1 and 250 pg for GAPDH based on 250 ng total RNA (Fig.
1a). For total RNA from prostate tissue, the equivalent titration point was 1 pg standard RNA for 5a-reductase 2 and 250
pg for GAPDH (Fig. 1b).
In the same way, the titration points of 5a-reductase 1 and
2 in human temporal lobe tissue were determined. The op-
5a-REDUCTASE IN THE HUMAN TEMPORAL LOBE
timal titration point was 100 fg standard RNA for 5a-reductase 1 and 250 pg for GAPDH based on 250 ng total RNA (Fig.
1c). For 5a-reductase 2, even a RNA standard amount of 1 ag
did not yield a detectable fluorescence signal of native RNA
(Fig. 1d). Conclusively, 5a-reductase 2 mRNA is not expressed in the human temporal lobe; only illegitimate transcription was detectable in each sample when the standard
RNA was omitted in the RT step.
Expression of 5a-reductase 1 mRNA in temporal lobe tissue
from children and adults
5a-Reductase 1 mRNA concentrations in the cerebral cortex did not differ significantly between women [25.9 6 7.9
arbitrary units (aU); n 5 28] and men (20.4 6 2.8 aU; n 5 25),
but they were significantly higher in the cerebral cortex of
adults (23.3 6 4.4 aU; n 5 53) than in that of children (6.4 6
2.3 aU; n 5 10; P , 0.005; Fig. 2a). No significant differences
in 5a-reductase 1 mRNA expression were observed between
3639
the cerebral cortex (23.3 6 4.4 aU; n 5 53) and the subcortical
white matter of adults (32.6 6 5.6 aU; n 5 31; Fig. 2). Also,
5a-reductase 1 mRNA concentrations did not differ significantly in the subcortical white matter of women (26.4 6 6.8
aU; n 5 15) and men (38.5 6 8.7 aU; n 5 16; Fig. 2b). As only
two white matter specimens from children were available, a
statistical analysis of the mRNA expression in the white
matter of adults and children is impossible. However, in the
2-yr-old boy studied, the 5a-reductase 1 mRNA concentration was 2.8 aU, and in the 13-yr-old girl, it was 3 aU, which
means a low level of expression compared to the expression
levels in adults.
5a-Reductase activity in the temporal lobe of children
and adults
Studies were then performed to characterize 5a-reductase
activity in temporal lobe tissue. Due to the limited amount
of tissue available, studies of 5a-reductase activity had to be
performed on a smaller number of specimens than studies of
5a-reductase 1 mRNA expression. However, using androstenedione as the substrate, 5a-reductase activity was
present in all studied temporal lobe specimens of children
and adults. As summarized in Table 2, in cortex tissue the
apparent Km of 5a-reduction did not show significant differences between crude nuclear fractions and crude supernatants containing the microsomes or between the two sexes.
Although only two children’s specimen could be studied, no
obvious difference concerning the Km value of 5a-reduction
between children and adults was present; the maximal veloity of 5a-reduction in the two children was higher than that
in almost all adults.
To evaluate possible differences in the kinetics of 5areduction between cortex tissue and the subcortical white
matter, 5a-reduction was studied in cortex and white matter
tissue homogenates from the same individuals. As shown in
Table 3, no obvious differences could be detected in specimens from a man, a woman, and a boy.
In cortex homogenates of a man and a women, 5a-reductase had a broad pH optimum (6.0 – 8.5) centered at pH 8
(Fig. 3).
Discussion
FIG. 2. Expression of 5a-reductase 1 mRNA in human cortex tissue
(a) and subcortical white matter tissue (b) of children and adults.
Over 20 yr ago, a number of investigators demonstrated
the presence of 5a-reductase activity in brain tissue (21, 22).
As most studies dealt with animal tissue, only few data
became available on 5a-reduction in the human brain. Some
investigators documented 5a-reductase activity in human
fetal brain (10 –13), but 5a-reductase activity in the brain of
adults was only demonstrated in a few tissue specimens (13,
14). Androgen metabolism in the human brain at different
ages has not been studied to date.
Two isozymes of 5a-reductase (types 1 and 2) with differential tissue distribution and biochemical and pharmacological differences have been identified in humans (7, 8). Our
study is the first to determine the expression of 5a-reductase
isoforms in a large number of fresh human temporal lobe
tissue specimens. The highly sensitive nested competitive
RT-PCR approach used permitted us to demonstrate that the
almost exclusive 5a-reductase gene expressed in the human
3640
JCE & M • 1998
Vol 83 • No 10
STOFFEL-WAGNER ET AL.
TABLE 2. Kinetic parameters of 5a-reduction in 1000 3 g precipitates presumably containing the nuclear fractions (nucl.) and in the
respective supernatants containing the microsomes (micros.) of cerebral cortex tissue
Men
Women
Children
n
Km, nucl.
(mmol/L)
Km, micros.
(mmol/L)
Vmax, nucl.
(pmol/h z mg protein)
Vmax, micros.
(pmol/h z mg protein)
5
4
2
0.99 6 0.36
1.16 6 0.38
1.04 6 0.24
1.20 6 0.21
1.08 6 0.50
1.34 6 0.23
110 6 19
148 6 53
247 6 5
118 6 20
119 6 43
246 6 16
The fractions were incubated with [3H]androstenedione in varying concentrations ranging from 0.05–3.5 mmol/L and 3 mmol/L NADPH at
37 C for 1 h at pH 7.5. Data are expressed as the mean 6 SEM.
TABLE 3. Comparison between kinetic parameters of 5a-reduction in homogenates of cortex and white matter tissue (white m.)
Man
Woman
Boy
Age
(yr)
Km, cortex
(mmol/L)
Vmax, cortex
(pmol/h z mg protein)
Km, white m.
(mmol/L)
Vmax, white m.
(pmol/h z mg protein)
29
49
5
2.35
1.44
1.63
84
51
211
2.47
2.1
1.28
73
44
230
The fractions were incubated with [3H]androstenedione in varying concentrations ranging from 0.05–3.5 mmol/L and 3 mmol/L NADPH at
37 C for 1 h at pH 7.5. Data are expressed as the mean 6 SEM.
FIG. 3. Effect of pH on 5a-reductase activity in cortex homogenates
of a 34-yr-old man and a 34-yr-old woman. Homogenates were incubated with [3H]androstenedione (1 mmol/L) and 3 mmol/L NADPH at
37 C for 1 h at a range of pH values. Representative data are shown
from single experiments performed in duplicate.
temporal lobe is the type 1 isoform. This is in accordance with
a study of Thigpen and co-workers, who found exclusively
5a-reductase 1 mRNA expression in tissue samples collected
postmortem from cerebellum, hypothalamus, pons, and medulla oblongata (9), and with studies on rat brain tissue
reporting a predominant expression of 5a-reductase 1 (23,
24).
Our data show that 5a-reductase 1 mRNA is present in
cortex tissue as well as in the subcortical white matter of
children and adults. To confirm and extend these experiments, 5a-reductase activity was measured in tissue homogenates. The enzyme activity was present in all tissue
specimens under investigation. The apparent Km values
and the pH profile of 5a-reduction substantiated the predominant expression of the type 1 isoform. The apparent
Km values determined in tissue homogenates (either cortex
tissue or white matter tissue) or in crude nuclear fractions
or crude supernatants containing the microsomes varied
between 0.99 mmol/L (mean) and 2.47 mmol/L (mean).
When COS cells were transfected with human 5a-reductase 1 cDNA, the apparent Km value obtained for androstenedione as the substrate was 1.7 mmol/L (7), so our
results are in accordance with the presence of 5a-reductase
1 in the human brain.
The broad pH optimum (6.0 – 8.5) centered at pH 8 of
5a-reduction indicates the predominant presence of 5areductase 1 in human brain, because in cell extracts prepared
from 293 cells transfected with the 5a-reductase 2 cDNA,
5a-reduction had a sharp pH optimum at pH 5.0, and in 293
cells transfected with the 5a-reductase 1 cDNA, it showed a
broad pH optimum between pH 6 – 8.5 (8).
One report on the subcellular distribution of 5a-reductase
demonstrated that 5a-reductase enzyme activity was highest
in the nuclear fraction in human fetal brain when androstenedione or progesterone was used as the substrate (10),
whereas in another study, using testosterone as the substrate,
the microsomal fraction displayed high activity rates in rat
brain tissue preparations (25). However, using androstenedione as the substrate, we could not find obvious differences
between the kinetic studies of 5a-reductase in 1000 3 g precipitates and the respective supernatants of cerebral cortex
homogenates.
In our study the expression levels of 5a-reductase 1 mRNA
and 5a-reductase activity did not differ significantly in cerebral cortex and subcortical white matter tissue. In the rat
and mouse brain, however, 5a-reductase activity appears to
be highly concentrated in the subcortical white matter,
whereas the cerebral cortex possesses a much lower activity
(14). In other animal species (hamster, bull, pig, and monkey)
and in brain tissue from a 61-yr-old woman, these researchers found the 5a-reductase activity to be more concentrated
in the cerebral cortex than in the white matter. The reasons
for these discrepancies may refer to differences between the
species.
5a-REDUCTASE IN THE HUMAN TEMPORAL LOBE
The expression levels of 5a-reductase 1 mRNA did not
differ significantly between the sexes, nor could obvious sex
differences concerning the kinetics of 5a-reduction be detected. As studies on the mRNA expression in human brain
are still lacking, only a comparison of our results with data
obtained for the enzymatic activity of 5a-reductase is possible. Our findings are consistent with previous studies in
which no significant sex differences concerning 5a-reductase
activity were found in neural tissue of nonhuman primates
during fetal development (26) or in rodents during postnatal
development (27, 28).
An important finding of this study is the fact that 5areductase 1 mRNA expression was significantly higher in
cortex specimens from adults than in those from children as
well as in tissue specimens from two postmenopausal
women (aged 50 and 53 yr) not receiving sex hormone replacement therapy.
Similar results were reported on the expression of 5areductase 1 in human skin tissue (9). At this point it is of
interest that the researchers found a steep increase in 5areductase 1 during puberty by immunoblotting. The data
presented suggest that there is a low 5a-reductase 1 mRNA
expression in the brain during childhood, which is further
induced during puberty, when serum sex steroid hormones
increase.
The physiological significance of 5a-reduction in the
brain remains unclear. The brain is an important target for
the effects of androgens; specific receptors have been identified in several regions of the brain, so androgens could
effect a genomic response (5). Based on differences in
substrate affinities and tissue distribution of the steroid
5a-reductase isozymes observed in the rat, it has been
concluded that type 2 may play an anabolic and type 1 a
catabolic role in the metabolism of androgens and other
steroid hormones (23). However, the physiological role of
steroid 5a-reductase isozymes in most tissues to date
awaits elucidation.
The metabolism of androgens occurring in the human
brain may subserve different physiological purposes at
different times of life, and this may account for the differences in the expression levels of 5a-reductase 1 in children and adults. On the other hand, the ubiquitous distribution of 5a-reductase in animal brain suggests that the
5a-reduced metabolites may be concerned with more general effects rather than exclusively with the regulation of
specific brain mechanisms, such as controlling reproductive function (1, 2). In contrast to reproductive and neuroendocrine actions of steroids via intracellular receptors
that regulate transcriptionally directed changes in protein
synthesis, certain pregnanes and androstanes rapidly alter
central nervous system excitability and produce behavioral effects (29). 5a-Reduced metabolites of progesterone
alter g-aminobutyric acidA receptor function, behavior,
drug metabolism, and neural development (3, 29). Therefore, the effects of those metabolites may involve both
genomic and nongenomic actions.
In conclusion, the present study is the first to determine
5a-reductase isozyme expression in the human temporal
lobe of children and adults. We found mRNA expression
of 5a-reductase 1 and 5a-reductase activity in the temporal
3641
lobe of children and adults. In contrast, 5a-reductase 2
mRNA was not detectable. The expression levels of 5areductase 1 did not differ significantly between the sexes
or between cerebral cortex and subcortical white matter
tissue, but they were significantly lower in children than
in adults (P , 0.005). Many questions regarding the biological role of 5a-reductase in the human brain are still
unanswered, and further efforts are required to delineate
and understand the physiological role of 5a-reductase activity in the brain.
Acknowledgments
We thank Dr. D. W. Russell, University of Texas Southwest Medical
Center (Dallas, TX), for the human 5a-reductase cDNAs; Prof. M. Nuri,
Department of Urology, Waldkrankenhaus Bonn (Bonn, Germany), for
the supply of prostate tissue; and Dr. M. Wolff, Department of Surgery,
University of Bonn (Bonn, Germany), for the supply of liver tissue.
References
1. Martini L. 1982 The 5a-reduction of testosterone in the neuroendocrine structures. Biochemical and physiological implications. Endocr Rev. 3:1–25.
2. Martini L, Melcangi RC. 1991 Androgen metabolism in the brain. J Steroid
Biochem Mol Biol. 39:819 – 828.
3. Lephart ED. 1993 Brain 5a-reductase: cellular, enzymatic, and molecular perspectives and implications for biological function. Mol Cell Neurosci.
4:473– 484.
4. Naftolin F. 1994 Brain aromatization of androgens. J Reprod Med. 39:257–261.
5. Sholl SA, Goy RW, Kim KL. 1989 5a-reductase, aromatase, and androgen
receptor levels in the monkey brain during fetal development. Endocrinology.
124:627– 634.
6. Wilson JD. 1975 Metabolism of testicular androgens. Hand Physiol. 5:491–508.
7. Andersson S, Russell DW. 1990 Structural and biochemical properties of
cloned and expressed human and rat steroid 5a-reductases. Proc Natl Acad Sci
USA. 87:3640 –3644.
8. Andersson S, Berman DM, Jenkins EP, Russell DW. 1991 Deletion of
steroid 5a-reductase 2 gene in male pseudohermaphroditism. Nature.
354:159 –161.
9. Thigpen AE, Silver RI, Guileyardo JM, Casey ML, McConnell JD, Russell
DW. 1993 Tissue distribution and ontogeny of steroid 5a-reductase isoenzyme
expression. J Clin Invest. 92:903–910.
10. Saitoh H, Hirato K, Yanaihara T, Nakayama T. 1982 A study of 5a-reductase
in human fetal brain. Endocrinol Jpn. 29:461– 467.
11. Schindler AE. 1976 Steroid metabolism in foetal tissues. IV. Conversion of
testosterone to 5a-dihydrotestosterone in human foetal brain. J Steroid Biochem. 7:97–100.
12. Mickan H. 1972 Metabolism of 4-14C-progesterone and 4-14-C-testosterone in
brain of the previable human fetus. Steroids. 19:659 – 668.
13. Jenkins JS, Hall CJ. 1977 Metabolism of [14C]testosterone by human foetal and
adult brain tissue. J Endocrinol. 74:425– 429.
14. Celotti F, Melcangi RC, Negri-Cesi P, Ballabio M, Martini L. 1986 A comparative study of the metabolism of testosterone in the neuroendocrine structures of several animal species. Neuroendocrinol Lett. 5:227–236.
15. Watzka M, Waha A, Koch A, et al. 1997 An optimized protocol for mRNA
quantification using nested competitive RT-PCR. Biochem Biophys Res Commun. 231:813– 817.
16. Ho SN, Hunt HD, Horton RM, Pullen JK, Pease LR. 1989 Site-directed
mutagenesis by overlap extension using polymerase chain reaction. Gene.
77:51–59.
17. Celi FS, Zenilman ME, Shuldiner AR. 1993 A rapid and versatile method
to synthesize internal standards for competitive PCR. Nucleic Acids Res.
21:1047.
18. Fisher LK, Kogut MD, Moore RJ, et al. 1978 Clinical, endocrinological and
enzymatic characterization of two patients with 5a-reductase deficiency: evidence that a single enzyme is responsible for the 5a-reduction of cortisol and
testosterone. J Clin Endocrinol Metab. 47:653– 664.
19. Schweikert HU, Totzauer P, Rohr HP, Bartsch G. 1985 Correlated biochemical
and stereological studies on testosterone metabolism in the stromal and epithelial compartment of human benign prostatic hyperplasia. J Urol.
134:403– 407.
20. Bruch HR, Wolf L, Budde R, Romalo G, Schweikert HU. 1992 Androstenedione metabolism in cultured osteoblast-like cells. J Clin Endocrinol Metab.
75:101–105.
21. Massa R, Stupnicka E, Kniewald Z, Martini L. 1972 The transformation of
testosterone into dihydrotestosterone by the brain and the anterior pituitary.
J Steroid Biochem. 3:385–399.
3642
STOFFEL-WAGNER ET AL.
22. Sholiton LJ, Taylor BB, Lewis HP. 1974 The uptake and metabolism of labelled
testosterone by the brain and pituitary of the male rhesus monkey (Macaca
mulatta). Steroids. 24:537–547.
23. Normington K, Russell DW. 1992 Tissue distribution and kinetic
characteristics of rat steroid 5a-reductase isozymes. J Biol Chem.
267:19548 –19554.
24. Lephart ED. 1993 Pituitary and brain 5a-reductase messenger RNA levels in
control, castrated, and dihydrotestosterone-treated rats. Mol Cell Neurosci.
4:526 –531.
25. Rommerts FFG, van der Molen HJ. 1971 Occurrence and localization of
steroid 5a-reductase, 3a- and 17b-hydroxysteroid dehydrogenases in hy-
26.
27.
28.
29.
JCE & M • 1998
Vol 83 • No 10
pothalamus and other brain tissues of the male rat. Biochim Biophys Acta.
248:489 –502.
Resko JA, Connolly PB, Roselli CE. 1988 Testosterone 5a-reductase activity
in neural tissue of fetal rhesus macaques. J Steroid Biochem. 29:429 – 434.
Massa R, Justo S, Martini L. 1975 Conversion of testosterone into 5a-reduced
metabolites in the anterior pituitary and in the brain of maturing rats. J Steroid
Biochem. 19:235–239.
Selmanoff MK, Brodkin LD, Weiner RI, Siiteri PK. 1977 Aromatization and
5a-reduction of androgens in discrete hypothalamic and limbic regions of the
male and female rat. Endocrinology. 101:841– 848.
Paul SM, Purdy RH. 1992 Neuroactive steroids. FASEB J. 6:2311–2322.