| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
ARTICLES |
Biochemistry and Laboratory of Endocrinology (M.B., B.H., A.C., P.H., M.D., E.R., J.C., G.H.) and Molecular Genetics Laboratory (J.P., J.C.), Institute of Pathology B23, avenue de l’Hôpital, 3, University of Liège, 4000 Liège, Belgium
Address all correspondence and requests for reprints to: Jean Closset, University of Liège, Biochemistry and Laboratory of Endocrinology, Molecular Genetics Laboratory Institue of Pathology B23, Avenue de l’Hôpital, Liège 4000, Belgium. E-mail: closset{at}ulg.ac.be
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The complete cDNA is 1187 bp long and codes for a protein of 298 amino acids that contains four potential glycosylation sites and three half cystinyl residues.
The PARM-1 gene was found to be expressed at quite low levels in most rat tissues including those of the urogenital tract. The kinetic of induction of PARM-1 gene in the prostate was highly correlated to the development of apoptosis in the whole organ. Supplementation of castrated animals with androgens reversed both the process of apoptosis and the overexpression of PARM-1 gene. Supplementation with estrogens did not result in an increase in the PARM-1 messenger RNA levels when compared with the castration alone. However, the treatment resulted in a more rapid return to intact levels in the castrated plus estrogen group. When apoptosis of testis and prostate was induced in vivo by hypophysectomy, it was found that PARM-1 was only overexpressed in the prostate. Therefore, PARM-1 seems to be regulated by androgens only in the prostate. Using in situ hybridization and immunohistological techniques, we have shown that PARM-1 gene product is found exclusively in the epithelial cells of involuting prostate. Analysis by flow cytometry of MAT LyLu epithelial cells transiently expressing PARM-1 protein did not allow us to demonstrate a direct effect of PARM-1 gene overexpression on the programmed death of the transfected cells. Treatment of MAT LyLu cells by transforming growth factor-ß induced apoptosis but had no effect on PARM-1 production. However PARM-1 protein has been detected by Western blotting in various cell lines such as MAT LyLu, MAT Lu, and PIF, which are androgen independent. This would suggest that PARM-1 gene product would be a marker for acquired androgen-independence of these tumor cells.
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
While androgen receptor is present in both epithelial and stromal compartments of the gland, epithelium-stroma interactions seems to be involved in the onset of androgen-deprivation-induced prostatic regression. In rats, castration causes increased expression of the gene coding for transforming growth factor ß1 (TGF-ß1) in stromal cells (8, 9) and of the genes encoding type-I and type-II TGF-ß receptors (TßRI and TßRII, respectively) in epithelial cells (10, 11). Androgen-deprivation-induced TGF-ß1 may thus bind to its membrane receptor on epithelial cells and initiate the signal transduction cascade leading to apoptosis. This cascade includes overexpression of the gene encoding the insulin-like growth factor binding protein 3 (IGFBP-3) (12, 13). The protein seems to act as a mediator of the cascade because TGF-ß1-induced prostatic epithelial cell apoptosis is inhibited by IGFBP-3-neutralizing antibodies or IGFBP-3 antisense oligonucleotides (14). On the other hand, several proteases including cathepsins B and D (15, 16, 17), matrix metalloproteinases 2 and 9 (18), tissue- and urokinase-type plasminogen activators (19, 20), and matrilysin (21) are activated in prostatic involution induced by androgen deprivation. These enzymes are elements of a proteolytic cascade responsible for the successive degradation of protease inhibitors and of major components of the basement membrane. Several other genes are also up-regulated in the involuting prostate upon androgen deprivation, notably the genes encoding testosterone-repressed prostatic message 2 (TRPM-2, commonly termed clusterin) (22, 23), activin (24), calmodulin (25), glutathione S-transferase subunit Yb1 (26, 27), par-4 (28), the L7 large ribosomal subunit protein (Bruyninx, M., H. Ammar, E. Reiter, A. Cornet, B. Hennuy, J. Poncin, and J. Closset, submitted for publication), and integrin-associated protein (IAP) (30). It is not yet known, however, if any members of the conserved interleukin-1ß-converting enzyme family (ICE or caspases) are activated in the prostate following castration. Furthermore, it is likely that many other androgen-dependent proteins involved in prostatic regression have yet to be identified.
In this report, we describe the cloning by subtractive hybridization of two complementary DNAs (cDNAs) corresponding to genes whose expression is increased in the rat prostate following androgen deprivation. The first corresponds to the gene encoding the large ribosomal subunit protein L5. The full-length sequence of the second (we have named the corresponding gene PARM-1 for "prostatic androgen-repressed message-1") is highly identical to a mouse expressed sequence tag (EST). The positive regulation of PARM-1 gene expression in the prostate of castrated rats was confirmed by immunodetection. We have also studied by the same technique its expression in different androgen-independent rat prostatic cell lines. To further examine PARM-1 gene expression, we have investigated the tissue distribution of the corresponding messenger RNA (mRNA) and examined its regulation in the prostate by the hormones testosterone, 17ß-estradiol, and cortisol. To assess the possible role of the PARM-1 gene product in the process of prostatic apoptosis, we have next measured its potency to induce programmed death in transiently transfected rat prostatic MAT LyLu cells. We have also studied the effects of TGF-ß-induced apoptosis on PARM-1 expression in the same cell line. Finally, using in situ hybridization and immunohistochemical techniques, we have sought to determine which prostatic cell type(s) is (are) responsible for PARM-1 gene expression in the involuting gland.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
A first set of male rats were castrated by scrotal incision under ether anesthesia. Animals were killed by cervical dislocation 1, 2, 3, 4, 5, 7, and 14 days after surgery (n = 4 per group).
A second set of male rats were castrated under the same conditions and received from the fourth day after orchidectomy sc injections of testosterone proprionate (Sigma-Aldrich Corp., St. Louis, MO) at the dosage of 1 mg/rat·day in 300 µl sesame oil containing 10% ethanol. Animals were killed 12, 24, and 48 h after the first injection (n = 4 per group).
A third and fourth set of male rats were castrated and immediately injected either sc with 17ß-estradiol (Sigma-Aldrich Corp.) in 100 µl sesame oil containing 10% ethanol (dosage: 50 µg/rat·day) or ip with cortisol (Sigma) dissolved in 1 ml PBS (0.01 M phosphate–0.15 MR NaCl) (dosage: 25 mg/rat·day). Animals were killed 1, 2, 3, and 4 days after castration and treatment (n = 4 per group).
A fifth set of intact male rats were injected sc with 17ß-estradiol in 100 µl sesame oil containing 10% ethanol (Sigma-Aldrich Corp.) (dosage 50 µg/rat·day). Animals were killed 1, 2, 3, and 4 days after the first injections (n = 4 per group).
A sixth set of 15 male rats were hypophysectomized and killed by cervical dislocation respectively after 2, 4, 6, 8, and 10 days after surgery (n = 3 per group).
The ventral prostates of normal (n = 4) and treated animals were rapidly removed after rats were killed, frozen in liquid nitrogen, and stored at -70 C until further processing.
Spleen, liver, heart, and lung were removed from normal male and female rats (n = 4) as were the testes, seminal vesicles, and epididymis from male rats and the ovaries from female animals. The different tissues were frozen in liquid nitrogen and kept at -70 C until RNA isolation.
RNA isolation
Total RNA from the different tissues of intact and treated rats
was isolated by the single-step guanidium-thiocyanate method (31).
Poly(A+)mRNA was purified from the ventral prostates of
normal and 3-day-castrated animals by means of the PolyATtract mRNA
isolation system IV (Promega Corp., Madison, WI).
Construction and screening of a prostatic cDNA subtraction library
(castrated vs. normal rats)
Two cDNA libraries were constructed in the 
Uni-ZAP XR
expression vector according to the manufacturer’s instructions
(ZAP-cDNA synthesis kit, Stratagene, La Jolla, CA).
Each was produced from 5 µg of prostatic poly(A+)RNA, but
in one case the poly(A+)mRNA was obtained from
3-day-castrated rats and in the other case from intact rats. The
recombinant phages were encapsidated with the Gigapack II Gold
Packaging Extract (Stratagene) and amplified in SURE
bacterial cells (Stratagene).
The subtractive procedure was derived from that of Jiang and Fisher with minor modifications (32). The method is based on the in vivo excision properties of the pBluescript phagemid and especially on the ability to obtain it in either single- or double-stranded form with the ExAssist helper phage/SOLR strain system (Stratagene).
Briefly, phagemids containing tester and driver prostatic cDNAs
(corresponding to castrated and normal rats respectively) were excised
from the ZAP phage using the mass excision procedure described by
Short and Sorge (33). Double-stranded tester cDNA inserts were then
excised from the phagemid by a 2 h digestion with XhoI
and NotI restriction enzymes. Four hundred nanograms of
these cDNA inserts were next hybridized with 4 µg of biotinylated
single-stranded driver phagemid in 20 µl of a solution containing 0.5
M NaCl, 50 mM HEPES (pH 7.6), 0.2% SDS, 40%
deionized formamide. The mixture was heat-denatured for 5 min, cooled
on ice, and incubated for 48 h at 42 C. Subtraction of the
noncastration-specific cDNA inserts hybridized with biotinylated
phagemid and of excess biotinylated driver phagemid was performed using
streptavidin-coupled paramagnetic beads (Promega Corp.)
according to the manufacturer’s instructions. Castration-specific cDNA
inserts were then inserted into the
XhoI-NotI-digested Uni-ZAP XR expression
vector (Stratagene) and phage particles were packaged
using the Gigapack II Gold Packaging Extract
(Stratagene).
For screening the library after subtraction, two radiolabeled cDNA
probes were prepared by RT. The probes were representative of the
prostatic mRNA populations of 3-day-castrated and normal rats. Two
micrograms of castrated and normal rat prostate
poly(A+)mRNA were incubated for 1 h at 37 C in 50
mM Tris (pH 8.3), 40 mM KCl, 6 mM
MgCl2, 10 mM DTT, 0.5 mM
dATP-dTTP-dGTP, 20 µM dCTP, 20 µg/ml BSA in the
presence of 1 µg oligo(dT) 12–18, 250 µCi
[
32P]dCTP (3000 Ci/mmol, NEN Life Science Products, Boston, MA), and 100 U Moloney-Murine Leukemia Virus
Reverse Transcriptase (Roche Molecular Biochemicals,
Mannheim, Germany). Probes were purified on Sephacryl S-400 columns and
RNAs were subsequently hydrolyzed by incubation for 15 min at 65 C in
the presence of 1 N NaOH. Each clone of the plated
subtraction library was picked up and individually in vivo
excised from the ZAP phage as described above. Each
EcoRI-linearized cDNA-containing phagemid (7.5 µg) was
denatured for 30 min at room temperature in 0.1 N NaOH, dot
blotted in duplicate on reinforced nitrocellulose membranes
(Schleicher & Schuell, Inc., Dassel, Germany) in 10
x SSC, and bound to the membranes by baking for 2 h at 80 C.
Filters were prehybridized at 42 C for 6 h in 50 mM
Tris (pH 7.5), 1 M NaCl, 50% deionized formamide, 10%
dextran sulfate, 1% SDS, 0.2% BSA, 0.2% polyvinylpyrrolidone, 0.2%
Ficoll, 0.1% sodium pyrophosphate, 0.1 mg/ml denatured salmon sperm
DNA and hybridized at 42 C for 16 h in the same buffer (without
dextran sulfate or NaCl) containing 2 x 106 cpm/ml of
the reverse-transcription-generated cDNA probes. Filters were washed
for 2 x 5 min at room temperature in 2 x SSC and for 30 min
at 65 C in 2 x SSC-0.1% SDS, then autoradiographed on Hyperfilm
ßmax (Amersham Pharmacia Biotech, Aylesbury, UK) with an
intensifying screen. Castration-induced up-regulation of the candidate
genes was confirmed by Northern blot analysis (as described below). The
corresponding cDNAs were sequenced with an automated sequencer (Applied
Biosystems, Foster City, CA) and the PRISM Ready Reaction Dyedeoxy
Terminator Cycle Sequencing kit (Applied Biosystems). Clones were
identified by comparing the sequences with those listed in the GenBank
and EMBL databases by means of the FASTA program.
cDNA probe labeling and Northern blot analysis
cDNA fragments were excised from pBluescript phagemids by
digestion for 2 h with XhoI-NotI, then
labeled with [32P]dCTP (NEN Life Science Products) as recommended by the manufacturer of the random
primed DNA labeling kit (Roche Molecular Biochemicals).
Radiolabeled cDNAs were separated from unincorporated dNTPs by
chromatography on Sephadex G50 columns (Pharmacia & Upjohn, Uppsala, Sweden).
Total RNA preparations from the tissues of normal and treated rats were
denatured in 1 x MOPS buffer (pH 7.0) containing 6% formaldehyde
and 50% deionized formamide by heating at 65 C for 15 min. RNA samples
were then separated by electrophoresis on a 1% formaldehyde-agarose
gel and transferred to reinforced nitrocellulose membranes
(Schleicher & Schuell, Inc.) in a Vacugene blotting
apparatus (Amersham Pharmacia Biotech). 28S and 18S
ribosomal RNAs were used as size indicators. After baking at 80 C for
2 h in a vacuum oven, filters were prehybridized in 50%
formamide, 5 x SSC, 5 x Denhardt’s, 50 mM
Na2HPO4, 250 µg/ml heat-denatured salmon
sperm DNA (Sigma) for 2 h at 42 C. Hybridizations
were performed overnight at 42 C in the same solution containing the
[ 32P]-labeled cDNA probe. The membranes were then
washed successively in 2 x SSC, 0.1% SDS at room temperature and
in 0.1 x SSC, 0.1% SDS at 52 C. The blots were finally
autoradiographed on Hyperfilm ßmax (Amersham Pharmacia Biotech) with an intensifying screen at -70 C. Quantitation of
the signals obtained was done by scanning the autoradiograms
(Ultrogel-Scan, LKB, Uppsala, Sweden). The equal
amounts of total RNA loaded in each well were monitored by ethidium
bromide staining of the gel or by rehybridizing the filters with a
radiolabeled rat ß-actin probe.
Cloning of the full-length PARM-1 cDNA
An unknown 382-bp partial 3'-end cDNA corresponding to a gene
whose expression is increased in the ventral prostate of
androgen-deprived rats was cloned from the subtraction library. This
gene was termed PARM-1 for "prostatic androgen-repressed
message-1." The corresponding full-length cDNA was obtained by
5'-rapid amplification of cDNA end PCR (5'-RACE PCR) according to the
manufacturer’s indications (Marathon cDNA Amplification kit,
CLONTECH Laboratories, Inc. Palo Alto, CA). The
long-distance PCR reaction was carried out with a template consisting
of prostate poly(A+)mRNA from 3-day-castrated rats. The
specific downstream PARM-1 antisense primer was
5'-CTTGTCCTCAGCCACCTTCCTTTGT-3' (for the position, see Fig. 2
), and the
Marathon Adaptor specific upstream primer was
5'-CCATCCTAATACGACTCACTATAGGGC-3'. The conditions for << touchdown >>
amplification were as follows: 5 cycles of 94 C for 30 sec, 72 C for 4
min followed by 5 cycles of 94 C for 30 sec, 70 C for 4 min, and
finally 25 cycles of 94 C for 30 sec, 68 C for 4 min. The complete
1166-bp cDNA was inserted into the T/A-type pGEM-T PCR cloning vector
(Promega Corp.) and sequenced entirely in both directions
as described above.
|
Protein homogenates were extracted from about 500 mg of normal and treated rat tissues and from about 3 x 107 cells of the following types rat androgen-independent Dunning tumor cells r3327 MAT LyLu, MATLu, PIF (kindly provided by Dr. De Coster, Beerse, Belgium) and human PC-3 cells (obtained from the ATCC). This was done by sonicating the samples for 30 s in 1 ml ice-cold 50 mM Tris, pH 7.4 in the presence of protease inhibitors (50 U Trasylol (Bayer Corp., Leverkusen, Germany) and 10 mg phenylmethylsulfonyl fluoride). Proteins were denatured by adding 100 µl of 10% SDS and ß-mercaptoethanol to a final concentration of 0.2 M. The homogenates were then centrifuged at 15,000 x g for 10 min at 4 C. Supernatants were collected and 90 µg of proteins were size-fractionated by electrophoresis on a homogenous 12% polyacrylamide gel. Western blotting was performed according to the standard protocol described by the manufacturer of the apparatus (Bio-Rad Laboratories, Inc. Hercules, CA). Immunochemical detection of PARM-1 gene products was carried out with a polyclonal goat antirabbit IgG coupled to horseradish peroxidase, and 4-chloro-1-naphtol as substrate following the supplier’s indications (Sigma-Aldrich Corp.).
Construction of a PARM-1-expressing vector, cell transfections, and
detection of apoptosis
A 929-bp cDNA fragment containing the entire open reading frame
of PARM-1 was excised from pGEM-T by digestion for 2 h with
KpnI-NcoI. The fragment was blunt-ended with calf
intestinal alkaline phosphatase (Roche Molecular Biochemicals) and inserted in frame into the ECORV site of a
pcDNA3 expression vector (Invitrogen, Carlsbad, CA). The
construct was checked by asymmetric restrictions and partial
sequencing.
Dunning r3327 MAT LyLu cells were grown in RPMI-1640 medium supplemented with 10% heat-inactivated FBS (Life Technologies, Inc., Gent, Belgium), 50 U/ml penicillin (Life Technologies, Inc.), and 50 µg/ml streptamycin (Life Technologies, Inc.). Transient transfection assays were performed by the lipofectamine method (Life Technologies, Inc.), with 5 x 105 cells and 4.8 µg of PARM-1-containing pcDNA3 or 4.8 µg of pCMV-ßgal or vector alone as negative controls. MAT LyLU cells were harvested 24 h after transfection for detection of signs of apoptosis. DNA fragmentation was revealed by DNA content analysis. Transfected MAT LyLu cells were washed twice in PBS and fixed in 70% ethanol for 30 min at -20 C. The cells were then incubated with 50 µg/ml RNase (Sigma) and 50 µg/ml propidium iodide for 10 min at room temperature. The relative DNA content was measured by red fluorescence analysis using a fluorescence-activated cell sorter (FACStar Plus, Becton Dickinson and Co., Rutherford, NJ) equipped with an argon laser (ILT air-cooled with 100 mW excitation lines at 488 nm). A conventional scatter gating method was used to eliminate debris from the analysis; cells and nuclei doublets were excluded by means of the pulse processor board (Becton Dickinson and Co.). Propidium iodide emission signals were collected with the help of a 630-nm filter (band pass 22). Ten thousand events per sample were collected in list mode, stored, and analyzed with the Consort 32 system (Becton Dickinson and Co.).
TGF-ß-induced apoptosis and effects on mean PARM-1 protein
level
Dunning r3327 MAT LyLu cells were grown in RPMI-1640 medium
supplemented with 10% heat-inactivated FBS (Life Technologies, Inc.), 50 U/ml penicillin (Life Technologies, Inc.), and 50 µg/ml streptamycin (Life Technologies, Inc.). 5 x 107 cells were rinsed three times
in PBS then incubated for 24 h in the same medium (without serum)
in the presence or in the absence of TGF-ß (2 ng/ml, Promega Corp.). TGF-ß-induced apoptosis was confirmed by DNA
fragmentation analysis. Genomic DNA was extracted from 2 x
107 cells as described previously (35) and loaded onto a
1.5% agarose gel containing 0.1 mg/ml ethidium bromide.
Electrophoresis was carried out in 40 mM Tris acetate, 1
mM EDTA, pH 8.0 and the DNA was visualized under UV light
and photographed.
Proteins were extracted from the remaining 3 x 107 cells, denatured, and size-fractionated as described above. Immunochemical detection of PARM-1 gene products was carried out using the polyclonal anti-PARM-1 antibody.
In situ hybridization
The protocol described by Arce and co-workers (36) was used to
perform in situ hybridizations on 16-µm-thick frozen
sections prepared from the prostates of normal, 3-day-, and
14-day-castrated rats. The prostates were fixed with 4%
paraformaldehyde in 0.12 M phosphate buffer, pH 7.4, and
cryopreserved in 15% sucrose and 0.12 M phosphate buffer,
pH 7.2, before embedding in the same buffer plus 7.5% gelatin.
The PARM-1-containing plasmid pcDNA3 was used as a template to prepare antisense riboprobes. The plasmid was linearized with EcoRI and digoxygenin-UTP (DIG-UTP) labeled probes were transcribed using SP6 RNA polymerase and the Roche Molecular Biochemicals DIG RNA labeling kit. After overnight hybridization at 70 C with 500 ng/ml probe, sections were washed three times for 30 min at 65 C in 50% formamide, 1 x SSC, 0.1% Tween 20 and twice for 30 min in MABT buffer before blocking in buffer A (MABT, 2% blocking reagent from Roche Molecular Biochemicals, 20% sheep serum) for 1 h. Sections were then exposed overnight at room temperature to a 1:5000 dilution of anti-DIG-alkaline phosphatase (AP)-conjugate (Roche Molecular Biochemicals) in buffer A. After washing for 30 min in MABT, the bound DIG-probe was visualized by means of an AP-catalyzed color reaction using NBT/BCIP (Roche Molecular Biochemicals). The color reaction was stopped by washing in water. Sections were dehydrated in graded alcohol series and slides were finally washed twice in xylene and mounted with Eukitt (Poly Labo, Strasbourg, France).
Immunohistochemistry
Ventral prostates from normal and 3-day-castrated rats were
embedded in Tissue-Tek (Sakura, Torrance, CA), frozen on dry ice, and
sectioned at a thickness of 6 µm for immunohistochemical analysis.
Sections were fixed in 1% formaldehyde for 15 min at room temperature,
air dried, and permeabilized in 70% ethanol for 2 min at -20 C. A
30-min incubation at room temperature was then performed with 20%
normal pig serum, 0.1% BSA, and 1% powdered milk in TBS to block
nonspecific sites on the sections, which were next incubated for 2
h at room temperature with either a 1:200 dilution of anti-PARM-1
antibody in TBS or a nonimmune rabbit serum as a negative control. The
immunoreaction was visualized with a biotinylated pig antirabbit
secondary antibody (DAKO Corp., Merelbeke, Belgium)
(diluted 1:1500 in TBS), followed by avidin-biotin-alkaline phosphatase
complex (DAKO Corp.) and neofuchsin staining (DAKO Corp.). After rinsing, the sections were counterstained slightly
with Meyer’s hematoxylin (Sigma) and mounted with
Glycergel (DAKO Corp.).
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Uni-ZAP XR
(Stratagene), starting with poly(A+)mRNA
extracted from apoptotic and control prostates respectively. A cDNA
subtraction library was subsequently generated in the same vector,
according to a method adapted from the original one described by Jiang
and Fisher (32), with a 10-fold excess of normal compared with
castrated rat cDNAs in the subtractive hybridization step (see
Materials and Methods for details). The subtraction library,
enriched in sequences corresponding to genes overexpressed during
prostatic apoptosis, comprised about 4,000 recombinant clones
containing cDNA inserts ranging in size from 200 to 1800 bp. One
thousand inserts were excised individually from the recombinant clones
by means of the ExAssist helper phage system (Stratagene)
and spotted in duplicate onto nitrocellulose filters. These cDNAs were
first screened for differentially expressed gene products using two
radioactive reverse-transcribed cDNA probes generated with
poly(A+)mRNA from the prostates of either normal or
3-day-castrated rats. Candidate clones were then used individually as
probes for secondary screening by Northern blot hybridization. Among 25
candidate clones examined, 6 turned out actually to contain cDNAs
corresponding to genes whose expression was significantly higher in the
prostates of 3-day-castrated rats than in those of normal animals (Fig. 1). Densitometric analysis of the signals
obtained on the autoradiograms showed stimulation factors of 13, 3,
160, 16, 118, and 126, respectively, for transcripts of cDNAs 4, 17,
201, 359, 616, and 725. These cDNA fragments were completely sequenced,
and comparisons of the sequences with existing databases revealed more
than 92% identity between 4 clones (clones 4, 201, 616, and 725) and
the clusterin gene product and approximately 94.8% sequence identity
between clone 17 and large ribosomal subunit protein L5 cDNA (Fig. 1).
On the other hand, clone 359 (henceforth designated as PARM-1 for
"prostatic androgen-repressed message-1") shared 81.5% identity
with 157 bp of a human expressed sequence tag (accession number
Z44658).
|
, the complete PARM-1 cDNA contains a
894-bp open reading frame starting with an ATG initiation codon at
position 202 and ending with a TAA termination codon at position 1096.
The putative encoded protein has a predicted molecular mass of 33 kDa
and a calculated isoelectric point of 5.51. Four possible glycosylation
sites were identified, but comparisons with existing databases revealed
no similarities to consensus sequences or known domains.
Western blot analysis
To investigate whether the level of PARM-1 protein also increases
upon androgen deprivation, polyclonal antibodies were raised against a
synthetic 19-amino acid peptide matching the hydrophilic COOH-end
(residues 279 to 298) of the putative deduced amino acid sequence of
PARM-1. Rabbit antiserum was prepared and used to probe Western blots
of rat prostatic protein extracts. As shown in Fig. 3, no signal was observed in the lane
corresponding to normal animals, but two bands were detected in
prostatic protein extracts from 3-day castrated rats. The size of the
first band (around 30 kDa) is consistent with the predicted molecular
mass of the PARM-1 protein, but the second band corresponds to a
protein with a much higher molecular weight (around 100 kDa) that could
result from glycosylation of the PARM-1 protein combined with
incomplete reduction of interchain disulfide bonds at the
ß-mercaptoethanol concentration used (0.2 M). These
results thus confirm those of the Northern blotting experiments: PARM-1
gene expression is indeed up-regulated in the prostates of castrated
rats, at both transcript and protein level. When used with proteins
extracted from the androgen-independent rat prostatic tumor cell lines
MAT LyLu, MAT Lu, and PIF, anti-PARM-1 antibodies also detected the
protein (Fig. 3), whereas no sign of apoptosis was observed in these
cells. Surprisingly, only the 33-kDa form of the protein was detected
in the different cell lines. Furthermore, no signal was observed when
anti-PARM-1 antibodies were used to test androgen-independent human
prostatic PC-3 cells.
|
). PARM-1 mRNA was detected in a
wide range of tissues, but none was found in the mammary gland, spleen,
or liver. Differences in mRNA levels were seen among the tissues, such
as the relatively high level of PARM-1 transcript in the heart compared
with the quite low level observed in most tissues examined. It is
noteworthy, however, that corresponding male and female tissues
displayed similar PARM-1 transcript levels.
|
shows that PARM-1 transcript levels
were fairly low in the prostates of intact animals and that they
increased gradually after androgen deprivation, peaking between days 3
and 5 post castration (16-fold induction), then decreasing. Such a
curve, paralleling the evolution of the proportion of apoptotic cells,
has already been observed for genes known to be involved in prostatic
apoptosis, such as those encoding clusterin (22, 23), TGF-ß1 (38),
matrilysin (21), or the ribosomal protein L7 (Bruyninx, M., H. Ammar,
E. Reiter, A. Cornet, B. Hennuy, J. Poncin, and J. Closset, submitted
for publication).
|
).
Near-normal transcript levels were restored within 48 h. We also
investigated the effects of 17ß-estradiol injections on prostatic
PARM-1 mRNA levels because estrogen has been shown to inhibit growth of
an androgen-sensitive rat Dunning prostatic tumor (40, 41), and, in
combination with castration, to reduce prostatic epithelial cell number
by an apoptotic mechanism (42). Figure 5C shows that ß-estradiol
treated animals displays no change in the expression of PARM-1 gene.
When combined with castration, on the other hand, estrogen
administration did not cause the PARM-1 mRNA level to rise any higher
than in rats subjected to castration alone, but there was a more rapid
return to intact levels in the castrated plus estrogen group (Fig. 5D).
Glucocorticoids have also been shown to delay the increase in cell
death and to block the rise in clusterin expression in the prostate
when administered to castrated rats (43, 44), but prostatic PARM-1
transcript levels were no different in castrated rats treated for
1, 2, 3, and 4 days with cortisol (25 mg/rat·day) than in rats
subjected to castration alone (Fig. 5E).
Analysis of PARM-1 expression in hypox rat prostate and
testis
To determine whether PARM-1 gene expression is also
regulated in vivo in other apoptotic organs than the
involuting prostate after androgen withdrawal, we have measured its
transcript levels using the following models: first, during thymic
regression after glucocorticoids administration; second, in the mammary
glands after weaning; third, myocytes after heart attack, and finally,
during the atrophy of the kidney after unilateral ligature of the
ureters. In all these in vivo apoptotic models, we didn’t
observe any change in the transcript levels of PARM-1 gene compared
with control animals (data not shown) suggesting that PARM-1 gene
product was not involved in apoptosis. However, to assess whether
PARM-1 expression is specifically associated with an apoptotic process
triggered by androgens, we have studied the PARM-1 mRNA levels in both
prostate and testis of adult hypophysectomized rats (Fig. 6).
|
).
Interstingly, PARM-1 expression was not affected in the regressing
testis of adult hypophysectomized rats (Fig. 6B), suggesting that the
regulation of PARM-1 gene expression by androgens was restricted to the
prostate.
Effects of PARM-1 overexpression in MAT LyLu cells
To determine whether the protein encoded by PARM-1 cDNA is a
potent inducer of apoptosis in prostatic cells, transient expression
experiments were performed with a pcDNA3 vector expressing the PARM-1
gene under the control of the CMV enhancer/promoter. Vector alone and a
pCMV-ßgal construct were used as negative controls. Rat MAT LyLu
cells were chosen for the transfection assays because, as illustrated
in Fig. 3, this cell line displays the lowest PARM-1 protein content of
all the cell lines studied. Overexpression of PARM-1 in transfected
cells was checked by probing Western blots with polyclonal anti-PARM-1
antibody (data not shown). Twenty-four hours after transfection, cells
were harvested and further processed for DNA content analysis by flow
cytometry (data not shown). Although programmed death could be induced
in MAT LyLu cells by various apoptosis inducers, the percentage of
apoptotic cells was about the same among PARM-1- and
controls-transfected cells. A TUNEL assay for DNA fragmentation
confirmed the results of the DNA content analysis (data not shown). We
conclude that PARM-1 is most probably not an inducer of programmed cell
death.
TGF-ß induced apoptosis of MAT LyLu cells and PARM-1 protein
level
Treatment during 24 h of MAT LyLu cells by 2 ng/ml of TGF-ß
induced apoptosis as demonstrated by DNA fragmentation analysis (Fig. 7). However, immunochemical detection of
PARM-1 protein in the cellular extracts did not reveal any difference
in PARM-1 level between TGF-ß treated cells and controls. This
results corroborates well with those obtained in transient transfection
experiments performed on the same cell line.
|
). In
the ventral lobes of androgen-deprived rat prostates, however, almost
all glandular epithelial cells were intensively and specifically
colored in blue, whereas no signal was observed in stromal cells (Fig. 8B). The intensity of the staining was even more pronounced in the
epithelial cells lining the peripheral ducts of the gland. When used
with 14-day-castrated rat prostates sections, PARM-1 antisense
riboprobe still produced a quite strong signal in the remaining
epithelial cells, while these cells don’t undergo massive apoptosis
anymore (Fig. 8C). The data were confirmed by immunohistochemistry: as
depicted in Fig. 8E, PARM-1 gene products were easily detected (fuchsia
stain) by the polyclonal antibody in involuting rat ventral prostates,
exclusively at the level of the epithelial cells. No signal could be
observed however in any cells from normal rat prostate sections (Fig. 8D).
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
PARM-1 expression is detected in other organs besides the prostate, but it is quite low in all of them, including in the various tissues of the urogenital tract. Surprisingly, PARM-1 expression is markedly higher in the heart than in other organs of intact rats, although the heart is not known to be controlled by androgens. On the other hand, androgen deprivation by castration doesn’t increase PARM-1 expression in any other tissue than the prostate (data not shown). Furthermore, PARM-1 expression is not affected in the regressing testis of an adult hypophysectomized rat. This result suggests that the regulation of PARM-1 gene expression by androgens may be exclusively restricted to the prostate. Also surprising is the fact that no PARM-1 transcripts were detected in the rat mammary gland, even though this gland was the source of the corresponding mouse EST.
As shown by Isaacs and co-workers (3, 4, 37), the first morphological signs of apoptosis (DNA ladders and apoptotic bodies) appear in the rat ventral prostate on the second day postcastration. Because apoptosis peaks on day 4, PARM-1 induction kinetics thus closely parallels the increase in the proportion of prostatic epithelial cells undergoing apoptosis after castration. The induction profile of PARM-1 thus resembles that of other genes known to be involved in prostatic apoptosis, such as clusterin (22, 23), L7 (29), TGFß (38), and matrilysin (21). On the other hand, PARM-1 expression in the ventral prostate of a castrated rat is inhibited by testosterone administration. This inhibition parallels regression of apoptosis and resumption of proliferation in the epithelial cell population after androgen treatment (5, 6, 39). This correlation between PARM-1 and the physiological status of epithelial cells is corroborated by the fact that both in situ hybridization and immunohistochemical techniques revealed PARM-1 gene products in the epithelial cells only.
It has been shown in animals receiving a transplant that estrogens combined with castration inhibit growth of androgen-sensitive Dunning tumors (40, 41). In addition to this, estrogen treatment also causes reduction of tumor mass through activation of apoptosis (42). This is accompanied by overexpression of TGFß and its type-I and type-II receptors (51). In our supplementation experiments, estrogen administration did not cause any further increase in PARM-1 expression beyond the level reached in castrated animals receiving no estrogen. We did observe, however, a more rapid decrease in PARM-1 expression after orchidectomy in estrogen-supplemented animals. There is no obvious explanation for this effect. Our data are in agreement with those of Lee and co-workers (52, 53), who had shown that estradiol treatment following castration slows the rate of prostate involution.
Regarding the role of glucocorticoids in the prostate, cortisol is known to delay apoptosis in castrated rats and to alter expression of several genes such as clusterin, Hsp 70, and c-fos (43, 44). Surprisingly, administration of cortisol to castrated animals does not appear to affect PARM-1 overexpression. This may reflect the absence of glucocorticoid response elements in the promoter of the PARM-1 gene.
To investigate the potential role of PARM-1 in prostatic apoptosis, we have conducted transient expression experiments with this cDNA in a rat prostatic cell line (MAT LyLu). This line was chosen among other androgen-independent cell lines because it showed the lowest PARM-1 expression level in Western blotting experiments. DNA content and fragmentation analysis by flow cytometry revealed no significant difference between cells transiently transfected with PARM-1 and control cells. Consequently, overexpression of PARM-1 is not alone sufficient to induce programmed cell death. Moreover, the lack of PARM-1 overexpression during TGF-ß-induced apoptosis suggests that PARM-1 doesn’t participate in the intracellular pathway, leading to the programmed death triggered by the paracrine factor. Because prostate epithelial cells possess potent androgen receptors, it cannot be ruled out, however, that the PARM-1 gene could be involved in an apoptotic pathway mediated by the direct action of androgens at the epithelial level. On the other hand, a quite high level of PARM-1 transcripts was observed by in situ hybridization in the remaining epithelial cells of 14-day-castrated rat prostates. As these cells don’t undergo massive apoptosis anymore, PARM-1 gene products might therefore constitute a survival factor expressed by epithelial cells resisting or trying to resist apoptosis. Such a role has been proposed for clusterin (54). The sometimes high basal expression of PARM-1 in the different rat prostatic cell lines studied corroborates the potential role of PARM-1 as a survival or anti-apoptotic factor. Indeed, these cells do not require androgens for their proliferation and consequently don’t undergo apoptosis following androgen withdrawal. The PARM-1 gene product might also be an indicator of the passage of prostatic cells from an androgen-dependent to a hormone-resistant state. If so, PARM-1 could have some bearing on prostate cancer diagnosis, since prostate cancer inevitably but unpredictably evolves from an androgen-dependent to a hormone-independent stage. Cloning of the human equivalent of PARM-1 might thus prove useful in diagnosing tumors that have become androgen-independent.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
Received March 11, 1999.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
ZAP expression vector. Methods Enzymol 216:495–508[Medline]
This article has been cited by other articles:
 |
|
 |
|
  J. Grimm, A. Mueller, F. Hefti, and A. Rosenthal Molecular basis for catecholaminergic neuron diversity PNAS, September 21, 2004; 101(38): 13891 - 13896. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 K. Dillner, J. Kindblom, A. Flores-Morales, R. Shao, J. Tornell, G. Norstedt, and H. Wennbo Gene Expression Analysis of Prostate Hyperplasia in Mice Overexpressing the Prolactin Gene Specifically in the Prostate Endocrinology, November 1, 2003; 144(11): 4955 - 4966. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 Y. Li, P. J. Friel, D. J. McLean, and M. D. Griswold Cystatin E1 and E2, New Members of Male Reproductive Tract Subgroup Within Cystatin Type 2 Family Biol Reprod, August 1, 2003; 69(2): 489 - 500. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
B. Hennuy, E. Reiter, A. Cornet, M. Bruyninx, M. Daukandt, P. Houssa, V.-H. N'Guyen, J. Closset, and G. Hennen A Novel Messenger Ribonucleic Acid Homologous to Human MAGE-D Is Strongly Expressed in Rat Sertoli Cells and Weakly in Leydig Cells and Is Regulated by Follitropin, Lutropin, and Prolactin Endocrinology, October 1, 2000; 141(10): 3821 - 3831. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| 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 |
