| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
ARTICLES |
Division of Reproductive Biology, Department of Biochemistry and Molecular Biology, School of Hygiene and Public Health, The Johns Hopkins University, Baltimore, Maryland 21205
Address all correspondence and requests for reprints to: Dr. William W. Wright, Division of Reproductive Biology, School of Hygiene and Public Health, The Johns Hopkins University, 615 North Wolfe Street, Baltimore, Maryland 21205. E-mail: bwright{at}jhmi.edu
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Despite considerable data demonstrating that germ cells regulate steady state levels of many transcripts expressed by Sertoli cells, there has been no direct determination of whether such regulation reflects changes in gene transcription by these somatic cells and/or alterations in the rate of turnover of the transcripts. Consequently, the experiments in this paper directly test the hypothesis that germ cells regulate gene transcription by Sertoli cells. These experiments are conducted from the vantage point of the gene encoding cathepsin L (cath L). This gene was chosen because mature rat Sertoli cells in stage VI–VII seminiferous tubules express high steady state levels of this transcript, whereas at other stages there is minimal or undetectable expression of this transcript (5, 11). Additionally, there is evidence that this stage-specific expression results from multiple interactions between Sertoli cells and their adjacent germ cells (13). Studies of testes undergoing germ cell depletion and repletion led us to propose that at stages I–IV and stages IX–XIV, pachytene and diakinetic spermatocytes as well as step 9–15 spermatids repress expression of cath L messenger RNA (mRNA) (13). However, the increased expression of cath L mRNA by Sertoli cells within normal, stage V tubules argues that at midcycle either the repressive signal is no longer produced by the germ cells or its effect is counteracted by a different, derepressive signal from germ cells. Finally, at stages VI–VII, step 18 and 19 spermatids appear to further stimulate steady state levels of cath L mRNA in Sertoli cells (5, 13, 14). This stimulation must cease before spermiation, however, because cath L mRNA levels are reduced 4-fold in stage VIII tubules (15).
The experiments in this study use two experimental approaches to test the hypothesis that germ cells regulate transcription of the cath L gene by rat Sertoli cells. The first experiment compares the amounts of cath L mRNA that are synthesized in 1 h by mature rat Sertoli cells in stage VI–VII tubules and in stage IX–XII tubules. In the remaining experiments we isolate and characterize the 5'-end of the rat cath L gene and then examine the ability of germ cells to regulate the expression of cath L reporter constructs in mature Sertoli cells. The results of these experiments provide the first direct proof that germ cells regulate gene transcription by Sertoli cells.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Analysis of transcription of the cath L gene by Sertoli cells in
stage VI–VII and stage IX–XII seminiferous tubules
Stage-specific transcription of the cath L gene was assayed by a
modification of published methods for analysis of metabolically labeled
RNA (16). Forty centimeters of stage VI–VII or stage
IX–XII tubules were incubated for 1 h in 600 µl Ham’s
F-12/DMEM supplemented with 1 mM 4-thiouridine and 110
nM [3H]uridine (SA, 45 Ci/mmol;
Amersham Pharmacia Biotech, Piscataway, NJ). Tubules were
then solubilized in 4.2 M guanidine isothiocyanate, 25
mM sodium citrate, and 0.7% ß-mercaptoethanol and
stored frozen at -80 C. This procedure was repeated three more times,
the samples from the same stages were pooled, and total RNA was
isolated (11). As a negative control, another 1.5 m
of stage VI–VII tubules were incubated without 4-thiouridine, and the
RNA was isolated. Two hundred micrograms of total RNA from each set of
tubules were then dissolved in 1.2 ml 0.15 M LiCl in NES
[50 mM NaOAc (pH 5.6), 4 mM EDTA, and 0.1%
SDS]. This mixture was heated for 5 min at 70 C and incubated
batchwise for 2 h at 4 C with 0.6 ml (packed volume)
organomercurial agarose (Affi-Gel 501, Bio-Rad Laboratories, Inc., Hercules, CA) (16). The resin was then
sedimented, and RNA in the supernatant was precipitated. The resin was
washed three times for 5 min each time at 4 C with 1.3 ml 0.5
M LiCl in NES and once with 1.3 ml 0.1 M LiCl
in NES. and bound RNA was eluted with 200 µl 20 mM
dithiothreitol/0.1 M LiCl in NES. Aliquots containing RNA
were identified by measuring incorporated radioactivity, these aliquots
were pooled, and the RNA was precipitated in the presence of 1 µl
glycogen. The same 200-µl fractions were collected from the set of
tubules incubated without 4-thiouridine. One microgram of the bound RNA
(
50% total sample) from stage VI–VII and stage IX–XII tubules
incubated with 4-thiouridine and 60% of the total sample from stage
VI–VII tubules incubated without 4-thiouridine were fractionated on
denaturing agarose gels. Two micrograms of RNA that did not bind to the
resin were also fractionated, and the integrity of all samples of RNA
was demonstrated by staining the RNA in the gel with ethidium bromide.
RNA was then blotted to a nylon membrane and probed sequentially for
cath L mRNA and clusterin mRNA, which is expressed at all stages of the
cycle by Sertoli cells (17, 18, 19). Radioactivity was
detected with x-ray film, the film was scanned, and the intensities of
individual bands were quantified using IP Lab Imaging Software
(Scanalytics, Inc., Fairfax, VA). Control experiments demonstrated that
the intensity of the image on the film was linear and proportional to
the amount of RNA analyzed.
Analysis of the turnover of cath L mRNA in Sertoli cells in stage
VI–VII and stage I-IV tubules
Ten centimeters of stage VI–VII or stage I-IV seminiferous
tubules were collected immediately or cultured for 7 h in Ham’s
F-12/DMEM supplemented with 0.1% dimethylsulfoxide (DMSO; control) or
10 µg/ml actinomycin D. Preliminary experiments demonstrated that
this dose of inhibitor quantitatively blocked RNA synthesis. RNA was
then isolated, 3 µg RNA were fractionated on denaturing agarose gels,
and cath L mRNA was analyzed as described above.
Cloning of the 5'-end of the rat cath L gene and generation of cath
L-firefly luciferase reporter constructs
A Sprague Dawley rat genomic library (CLONTECH Laboratories, Inc., Palo Alto, CA) was probed with
[32P]cath L complementary DNA (cDNA) (11, 20). To identify genomic clones that contained the 5'-end of the
cath L gene, we rescreened cath L cDNA-positive clones with
32P-labeled 5'-CCTCAGGTGTTTGAACCATGACCCCTTTAC-3',
which encodes the most 5'-region of cath L cDNA (21). The
largest clone that hybridized to this oligonucleotide was analyzed
further. A restriction map of this clone was generated
(22), restriction fragments were subcloned, and ends were
sequenced (23). To obtain contiguous sequence starting
approximately 2 kb upstream of exon 1 and extending into intron 2,
nested deletions in a maximum of 150-bp increments were created and
sequenced (24).
Cath L genomic fragments were ligated into the promoterless firefly
luciferase (Luc) reporter construct, pGL-2 basic (Promega Corp., Madison, WI). The construct, cath L (-2060/+33)-Luc
contained a SacI to BamHI genomic fragment (Fig. 2
). The SacI site was converted to a blunt end by digestion
with Klenow polymerase. The construct, cath L (-244/+33)-Luc was
generated by PCR using primers that encoded KpnI or
HindIII restriction sites at the 5'- and 3'-ends of the
genomic fragment, respectively (Fig. 2). Constructs were purified by
ion exchange chromatography (endotoxin-free maxi columns,
QIAGEN, Chatsworth, CA) followed by
CsCl2 centrifugation (25). The
genomic fragment in both constructs was sequenced.
|
Transcription start sites were identified by both primer extension and
S1 nuclease protection analyses (31, 32). The
32P-labeled oligonucleotide used for primer
extension analysis was antisense to nucleotides +41 to +71 of cath L
mRNA (see Fig. 2
, A and B). S1 nuclease protection analysis was
performed using a 270-mer single strand cDNA. This probe was
synthesized using the same oligonucleotide used for primer extension
analysis. Products of both primer extension and S1 nuclease analyses
were fractionated on 7% denaturing acrylamide gels along with known
sequencing reactions as size standards.
Transfection of rat Sertoli cells with cath L reporter
constructs
Sertoli cells were isolated from mature rats as previously
described (33) and cultured at a density of 1.5 x
105 cells/cm2 on 30-mm
Millicell-HA culture chambers (Millipore Corp., Bedford
MA) coated with 280 µl Matrigel (Becton- Dickinson Laboratories,
Bedford, MA.). Cells were cultured in F-12/DMEM supplemented with 8H
[human transferrin (5 µg/ml), insulin (10 µg/ml), epidermal growth
factor (1 ng/ml), retinol acetate (3.5 x
10-8 M),
testosterone (10-7
M), human recombinant or highly purified ovine FSH (50
ng/ml), 2.1 µM vitamin E, and 200 µM
vitamin C]. Residual germ cells were removed after 24 h of
culture by a 2-min incubation in 50 mM Tris-HCl, pH 7.4.
The next day, cells were transfected for 5–6 h with 0.5 pmol of the
appropriate cath L-Luc reporter construct or pGL-2 basic, 0.09 pmol
cytomegalovirus-Renilla luciferase (pRL-CMV; Promega Corp.), 15 µl Lipofectamine (Life Technologies, Inc., Gaithersburg, MD), and sufficient F-12/DMEM to bring the
final volume to 1.35 ml. After transfection, cells were washed twice
with F-12/DMEM and cultured for an additional 18–20 h in F-12/DMEM
plus 8H. Cells were then collected by digestion of the Matrigel with
dispase (Becton Dickinson Laboratories, Bedford, MA),
washed twice with HEPES-saline, and lysed in 200 µl Passive Lysis
Buffer (Promega Corp.) supplemented with 5 µg/ml
leupeptin and 20 µg/ml aprotinin. Lysates were frozen in dry ice and
stored at -80 C until measurement of firefly and Renilla
luciferase activities (see below). Preliminary experiments demonstrated
that cotransfection with pRL-CMV did not change the apparent activities
of cath L-Luc constructs in Sertoli cells.
To study the effects of germ cells on cath L (-2060/+33)-Luc and cath L (-244/+33)-Luc activities, mature Sertoli cells were isolated, cultured, and transfected as described above. During transfection, a pool of spermatogenic cells from all stages of the cycle was obtained by digesting mature testes twice for 20 min each time with 0.1% collagenase, 0.2% hyaluronidase, 0.03% deoxyribonuclease, and 0.03% soy trypsin inhibitor. The germ cell suspension was filtered sequentially through 50-, 30-, and 20-µm pore size nylon cloth, clumps of cells were allowed to settle out in 2% BSA, and the suspended germ cells were pelleted by centrifugation and washed three or four times with culture medium. Microscopic examination of the pelleted cells revealed all types of male germ cells except elongated spermatids, which were removed during filtration of the germ cells through nylon cloth. After transfection, 0–4 x 106 germ cells were added to the Sertoli cells, and the cocultures were maintained for 18 h in 8H supplemented with 3.5 x 10-8 M retinoic acid, 1 mM sodium pyruvate, and 13 mM sodium lactate. Two control experiments were performed to characterize the morphology of germ cells in coculture and the specificity of the response to Sertoli cells to germ cells. In the first, cocultures containing 4 x 106 germ cells were fixed in 4% buffered glutaraldehyde, imbedded in Epon, sectioned either parallel or at right angles to the Millipore membrane, and stained with toluidine blue. The morphology of the germ cells was then examined by light microscopy. In the second control experiment, transfected Sertoli cells were cocultured with 2 x 106 K562 cells (American Type Culture Collection, Manassas, VA).
Luciferase assays
Firefly and Renilla luciferase activities were
measured in 5–20 µl Sertoli cell extracts using the dual luciferase
assay from Promega Corp. To correct for differences in
transfection efficiency, the activity of each cath L-Luc luciferase
construct in Sertoli cells was expressed as a ratio of firefly
luciferase to Renilla luciferase enzyme activities.
Titration of extracts demonstrated that the amounts of both luciferase
activities in cell extracts were in the linear ranges of the
assays.
Statistical analysis
Data were analyzed by ANOVA, and differences between individual
means tested by Fisher’s multiple range test using StatView
(SAS Institute, Inc., Cary, NC). Differences were defined
as significant at P 
0.05.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Densitometric analysis of the Northern blots demonstrated that stage
VI–VII tubules contained 7-fold more 4-thiouridine-labeled cath L mRNA
than did stage IX–XII tubules (Fig. 1A, compare the amounts of 4-thiouridine-labeled cath L RNA synthesized
by stage VI–VII tubules and by stage IX–XII tubules; this RNA bound
to the organomercurial agarose). Only 4-thiouridine-containing RNA
bound organomercurial agarose, as none of the cath L mRNA in stage
VI–VII tubules incubated without 4-thiouridine bound this resin (Fig. 1A, compare bound with not bound fractions from stage VI–VII tubules;
4-thiouridine -). In contrast to the stage-specific transcription of
the cath L gene, stage VI–VII and stage IX–XII tubules contained
similar amounts of metabolically labeled clusterin mRNA (Fig. 1A).
Results from the clusterin mRNA analysis confirm that metabolically
labeled RNA from both stage VI–VII and stage IX–XII tubules was
intact.
|
). Stage I–IV
tubules were studied because it was our experience that they contained
higher concentrations of cath L mRNA then stage IX–XII tubules. Thus,
we anticipated that it would be easier to examine transcript stability
at stages I–IV than at stages IX–XII. Actinomycin D treatment reduced
the amount of total RNA recovered from stage VI–VII and stage I–IV
tubules by 38% and 32%, respectively. However, Northern blot analysis
demonstrated that 7 h of actinomycin D treatment did not change
the amounts of cath L mRNA measured in 3 µg RNA from stage VI–VII
tubules and from stage I–IV tubules (Fig. 1B). Thus, taken together,
our data support the hypothesis that the stage-specific expression of
cath L mRNA results primary, if not solely, from stage-specific
transcription of the cath L gene.
Isolation of the 5'-end of a rat cath L gene
Although the above studies proved that the cath L gene was
transcribed in a stage-specific manner, they did not directly prove
that germ cells regulate gene transcription by Sertoli cells. To
generate this proof we decided to determine whether germ cells
regulated the expression of cath L reporter constructs that had been
transfected into mature Sertoli cells. We expected that this
demonstration would require the isolation of a new cath L genomic
clone, as published clones for the rat and murine cath L genes encode
only 270 bp of sequence upstream from the transcription start site
(34, 35). Those clones might lack important elements
mediating the effect of germ cells on cath L gene transcription by
Sertoli cells. We, therefore, cloned a 12-kb genomic fragment encoding
the 5'-end of the cath L gene. This fragment contained approximately 9
kb upstream from exon 1 through part of intron 4 (data not shown). A
series of nested deletions was generated for selected subclones of this
genomic fragment, and 3213 bp of continuous sequence were obtained
(Fig. 2). This sequence contained 2066 bp
of sequence upstream from the transcription start site (defined in Fig. 3), the first and second exons, the first
intron, and 43 bp of the second intron. As previously reported, the
translation start site is present in the second exon
(34).
|
When rat liver DNA was digested with BamHI,
EcoRI, PvuII, or SacI, fractionated by
agarose gel electrophoresis, and hybridized to the 747-bp genomic
probe, single restriction fragments of the predicted sizes were
detected (Fig. 3A). Single restriction fragments were also detected
after digestion with HhaI, HindIII,
KpnI, and PstI (data not shown).
We next asked whether the cloned genomic fragment contains the
transcription start site used by Sertoli cells. This question was
prompted by the possibility that transcription of the cath L gene by
Sertoli cells might be initiated from a cell-specific promoter that
used a novel transcription start site (38). However,
primer extension analysis identified the same two adjacent residues (C
and A) as transcription start sites in RNA from testis, kidney, liver,
and small intestine (Fig. 3B). S1 nuclease protection analysis was then
used to confirm this finding. The 3'-end of the 270 nucleotide single
stranded antisense DNA probe was 69 and 70 nucleotides downstream from
the nucleotides identified as start sites by primer extension analysis.
When this probe was hybridized to RNA from testis, kidney, liver, and
small intestine and then digested with S1 nuclease, fragments of 69 and
70 nucleotides were, in fact, generated. These nucleotides are labeled
nucleotides +1 and +2 in Fig. 2. Taken together, the data in Figs. 1 and 3 indicate that a single copy cath L gene is transcribed in a
stage-specific manner by rat Sertoli cells using the same transcription
start sites; thus, it is likely that Sertoli cells use the same core
promoter as other cell types.
Demonstration that germ cell repress expression of a cath
L-luciferase reporter construct in Sertoli cells and identification of
a region of the cath L gene that mediates the repressive effects of
germ cells
Two cathepsin L-luciferase reporter constructs, cath L
(-244/+33)-Luc and cath L (-2060/+33)-Luc, were used to test the
hypothesis that germ cells regulate gene expression by isolated Sertoli
cells. We chose these constructs for the following reasons. Preliminary
experiments demonstrated that cath L (-244/+33)-Luc was expressed in
Sertoli cells isolated from both mature (60-day-old) and immature
(30-day-old) rats and that expression of this construct was 3-fold
greater in the mature Sertoli cells (Charron, M., and W. W.
Wright, unpublished). Thus, this construct contains the core promoter
as well as regulatory elements active in these cells. We reasoned that
these regulatory elements might also mediate the effect of germ cells
on cath L gene transcription. We also tested cath L (-2060/+33)-Luc
because we recognized that the effects of germ cells might be mediated
by elements which were upstream from -244 of the cath L gene.
The analysis of the effect of germ cells on Sertoli cell gene
transcription entailed three experiments. In the first experiment, we
compared the expression of cath L (-2060/+33)-Luc, cath L
(-244/+33)-Luc and the negative control, pGL-2 basic in mature Sertoli
cells which were cultured in the absence of germ cells. Results
demonstrate that cath L (-244/+33)-Luc and cath L (-2060/+33)-Luc
produced similar levels of luciferase activity in mature Sertoli cells
and that these levels were at least 100-fold higher than the luciferase
activities in cells transfected with the negative control construct,
pGL-2 basic (Fig. 4A). These results
demonstrate that both cath L-Luc constructs produce promoter-dependent
luciferase expression in mature Sertoli cells. Additionally, they show
that elements upstream from -244 of the cath L gene do not affect
expression of the reporter gene when Sertoli cells are cultured without
germ cells.
|
In the second experiment we asked whether germ cells would repress the
activity in Sertoli cells of cath L (-244/+33)-Luc. In this experiment
Sertoli cells were transfected with cath L (-244/+33)-Luc for 5 h
and were then either cultured alone for an additional 18 h or were
cocultured with 1–4 x 106 germ cells.
Results demonstrated that activity of cath L (-244/+33)-Luc in Sertoli
cells cultured without germ cells did not differ from the activity of
that construct in Sertoli cells cultured with 1–4 x
106 germ cells (Fig. 4B).
In the third experiment, which was conducted simultaneously with the
second, we asked whether elements upstream from -244 mediate the
effects of germ cells on cath L gene transcription. Thus, Sertoli cells
were transfected with cath L (-2060/+33)-Luc and were then cultured
alone or were cocultured with 1–4 x 106
germ cells as described above. In contrast to what was observed when
Sertoli cells were transfected with cath L (-244/+33)-Luc, 1 x
106 germ cells caused a significant, 30%
reduction in expression of cath L (-2060/+33)-Luc (Fig. 4B).
Increasing the numbers of germ cells above 1 x
106 had no additional effect, demonstrating that
the effect of germ cells was saturable (Fig. 4B). Additionally,
statistical analysis confirmed that all doses of germ cells had a
significantly greater repressive effect on expression of cath L
(-2060/+33)-Luc than on cath L (-244/+33)-Luc. For example, 1 x
106 germ cells had a significantly greater effect
on expression of cath L (-2060/+33)-Luc (75% of control) than on
expression of cath L (-244/+33)-Luc (90% of control).
We next asked whether the germ cells remained viable in the cocultures.
Light microscopic analysis of 1 µm plastic sections of the cocultures
revealed that at the end of the experiment, greater than 95% of the
germ cells in the cocultures were morphology normal (data not shown).
Finally, we asked whether the effect of germ cells would be mimicked by
another cell type. In this experiment germ cells were transfected with
cath L (-2060/+33)-Luc and then cultured alone, with 2 x
106 germ cells or with 2 x
106 of control, K562 cells. Consistent with the
results of Fig. 4B, germ cells caused a 30% reduction in the activity
of cath L (-2060/+33)-Luc (expressed as the ratio of firefly
luciferase to Renilla luciferase activities) in mature
Sertoli cells (Table 1). This reduction
was due to a decrease in firefly luciferase activity encoded by cath L
(-2060/+33)-Luc and not to an increase in Renilla
luciferase activity (Table 1). In contrast, an equal number of K562
cells increased the ratio of cath L (-2060/+33)-Luc to
CMV-Renilla luciferase activities by 60% (Table 1). This
increase did not result from increased firefly luciferase activity, but
to decreased Renilla luciferase activity in the Sertoli
cells (Table 1). Taken together, the data from all of the coculture
experiments support the hypothesis that germ cells regulate gene
transcription by mature Sertoli cells. The repressive effect of germ
cells on expression of cath L (-2060/+33)-Luc is a response of the
Sertoli cells to viable germ cells and is not observed when Sertoli
cells are cocultured with control, K562 cells. The repressive effect of
germ cells is mediated by regulatory elements that are upstream from
-244 of the cath L gene.
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Our data prove that the cath L gene expressed in rat Sertoli cells is the same gene that is expressed in many different cell types of the rat. Our data also argue that all cells use the same cath L core promoter. In contrast to what has been reported for human cath L mRNA, we detected no heterogeneity in the 5'-untranslated region of rat cath L mRNA (39). This difference between the rat and human transcripts may stem from the fact that the first intron and the first exon of the rat gene share no area of significant sequence identity with the human gene. In summary, our data indicate that the effect of germ cells on cath L gene transcription by Sertoli cells involves the same gene and same core promoter that are active in many other cell types in the rat.
Stage-specific expression of cath L mRNA by Sertoli cells results
from changes in the rate of transcription of the cath L gene
In this study we used metabolic labeling to prove that
stage-specific changes in steady state levels of cath L mRNA result
primarily from changes in the rate of gene transcription. As with all
metabolic labeling studies, the validity of this experimental approach
was predicated on the equal incorporation of exogenous uridine at both
sets of stages. As tubules from both sets of stages exhibited similar
levels of incorporation of [3H]uridine and
yields of 4-thiourdine-labeled RNA, this requirement was met. The
validity of this experiment also required that the half-life of cath L
mRNA was similar at both sets of stages and substantially longer than
1 h. Otherwise, measurement of stage-specific differences in
transcription of the cath L gene would be confounded by the rate of
transcript turnover. However, our examination of the turnover of cath L
mRNA in stage VI–VII and stage I–IV tubules indicates that at both
sets of stages, the transcript turns over slowly, and the half- life of
the transcript appears to be longer than 7 h. We realize, however,
that, actinomycin D treatment can alter the turnover of some
transcripts (40), raising concerns about interpretation of
our data. In this regard, it is important to note that other
observations support the conclusion that the half-life of cath L mRNA
is longer than 7 h. Quantitative Northern blot analysis
demonstrates that after stage VIIa,b there is a logarithmic decrease in
cath L mRNA levels (15). This logarithmic decrease
continues until stage IX, and half-maximal levels of cath L mRNA are
reached in the middle of stage VIIc,d, which occurs 13.4 h after
stage VIIa,b. Assuming that transcription of the cath L gene ceases at
the end of stage VIIa,b, this observation is consistent with the slow
turnover of cath L mRNA that was observed in actinomycin D-treated
tubules. Therefore, taken together our studies demonstrate that
stage-specific expression of cath L mRNA by Sertoli cells results
primarily from changes in the rate of gene transcription. These studies
provide the first direct proof of stage-specific gene transcription by
Sertoli cells. The results of these studies, when placed in the context
of how germ cells regulate steady state levels of cath L mRNA,
suggest that the regulation of stage-specific transcription is
complex. At stages I–VI and stages IX–XIV, transcription is
repressed. At stages V–VII this repression is lost or derepressed, and
at stages VI–VII transcription is stimulated. This cycle of
transcriptional repression and stimulation appears to be due to a
sequence of interactions between mature Sertoli cells and the
surrounding germ cells that progress in synchrony through the stages of
the cycle of the seminiferous epithelium (13).
A region of the cath L gene upstream from the core promoter is
required for germ cells to repress cath L gene transcription
There are numerous reports that germ cells, when cocultured with
Sertoli cells, regulate the expression of a number of different
transcripts (9, 10, 41). However, this paper presents the
first direct proof that germ cells regulate gene transcription by
cultured Sertoli cells and demonstrates that the region of the cath L
gene that mediates the response to germ cells is upstream from the core
promoter. Our observation that a pool of total germ cells collected
from all stages of the cycle repressed the expression of cath L
(-2060/+33)-Luc is consistent with the fact that cath L mRNA
expression is repressed at stages I–IV and stages IX–XIV, which
constitute 55% of the duration of one cycle of the seminiferous
epithelium (42). This repressive effect of germ cells was
clearly saturable, as the addition of more than 1 x
106 germ cells did not further decrease cath L
(-2060/+33)-Luc activity in Sertoli cells. Such saturation is a
hallmark of a receptor-mediated process. This repressive effect may be
germ cell specific, as the addition of 2 x
106 K562 cells did not recapitulate the effect of
germ cells. However, why was there not a greater repressive effect of
germ cells? Two aspects of our experimental design provide possible
explanations. Firstly, in vitro, the germ cells are cultured
on top of the Sertoli cells, whereas in vivo, germ cells
surround the Sertoli cells. As such, our in vitro
experiments do not replicate the numbers of germ cells that bind
Sertoli cell in vivo or the distribution of the germ cells
around the Sertoli cells. Secondly, the added germ cells were from all
stages of the cycle of the seminiferous epithelium. This mixing of germ
cells from different stages is important, because we have previously
postulated that germ cells in stage V–VIII tubules may trigger a
derepressive signal within Sertoli cells, a signal that counteracts the
repression caused by germ cells at other stages (13). If
this is the case, then in the coculture experiments, 55% of the germ
cells were potentially inhibitory, whereas 45% of the germ cells were
potentially disinhibitory to cath L gene transcription
(42). This mixture of germ cells with potentially opposite
effects on cath L gene transcription could produce the 30% inhibition
of cath L (-2060/+33)-Luc expression that we observed.
In contrast to the repressive effect of germ cells on cath L (-2060/+33)-Luc expression, they had no effect on expression of the smaller construct, cath L (-244/+33)-Luc. These data demonstrate that regulatory elements upstream from -244 are essential for the response of cath L to the repressive effects of germ cells. Obviously, our experiments do not address whether elements downstream from -244 are also required for germ cells to repress transcription of the cath L gene.
An important next step in this research is to identify the specific elements that mediate the effects of spermatogenic cells on cath L gene transcription. Some of these elements reside between -2060 and -244. It is possible that these elements are shared with another gene that is also maximally expressed by Sertoli cells at stages VI and VII. Such a gene encodes the RII receptor for Mullerian inhibiting substance (MIS) (8). Comparison of the upstream sequences of both the cath L and MIS RII genes demonstrate that five areas of the cath L gene, ranging in size from 28–137 bp, exhibit 76–84% sequence identity with upstream areas of the MIS RII gene (43). All of these areas in the cath L gene are in the broad region that contains the elements mediating repressive effects of spermatogenic cells. Interestingly, for four of these six areas, the corresponding sequences in the MIS RII gene contain inverted Alu repeats. This is of particular interest because Alu and other retroviral inserts have become important regulatory elements in other genes (44). Thus, one or more of the regions in the cath L gene that contain the putative retroviral inserts may contain functional cis-acting elements that mediate responses to germ cells. If these regions contain those regulatory elements, they may mediate the sequential inhibitory, disinhibitory, and stimulatory effects of spermatogenic cells that are responsible for the stage-specific transcription of the cath L gene by mature rat Sertoli cells.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
2 S.Z. and M.C. contributed equally to this work.
3 Current address: Novartis Pharmaceuticals, Summit, New Jersey
07091.
4 Current address: Biology Department, Northeastern University,
Boston, Massachusetts 02115.
Received October 3, 2000.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
-glutamyl transpeptidase messenger ribonucleic acid-IV
in the rat epididymis. Biol Reprod 60:664–673This article has been cited by other articles:
 |
|
 |
|
  P J O'Shaughnessy, L Hu, and P J Baker Effect of germ cell depletion on levels of specific mRNA transcripts in mouse Sertoli cells and Leydig cells Reproduction, June 1, 2008; 135(6): 839 - 850. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. Charron, J.-Y. Chern, and W. W. Wright The Cathepsin L First Intron Stimulates Gene Expression in Rat Sertoli Cells Biol Reprod, May 1, 2007; 76(5): 813 - 824. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R.-S. Fujino, K. Tanaka, M. Morimatsu, K. Tamura, H. Kogo, and T. Hara Spermatogonial Cell-Mediated Activation of an I{kappa}B{zeta}-Independent Nuclear Factor-{kappa}B Pathway in Sertoli Cells Induces Transcription of the Lipocalin-2 Gene Mol. Endocrinol., April 1, 2006; 20(4): 904 - 915. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. D. Anway, M. D. Show, and B. R. Zirkin Protein C Inhibitor Expression by Adult Rat Sertoli Cells: Effects of Testosterone Withdrawal and Replacement J Androl, September 1, 2005; 26(5): 578 - 585. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. M. Hill, M. D. Anway, B. R. Zirkin, and T. R. Brown Intratesticular Androgen Levels, Androgen Receptor Localization, and Androgen Receptor Expression in Adult Rat Sertoli Cells Biol Reprod, October 1, 2004; 71(4): 1348 - 1358. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
K.-F. Lee, W. S.B. Yeung, J. F.C. Chow, C. K. Shum, and J. M. Luk Different Testicular Gene Expression Patterns in the First Spermatogenic Cycle of Postnatal and Vitamin A-Deficient Rat Testis Biol Reprod, April 1, 2004; 70(4): 1010 - 1017. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. D. Anway, W. W. Wright, B. R. Zirkin, N. Korah, J. S. Mort, and L. Hermo Expression and Localization of Cathepsin K In Adult Rat Sertoli Cells Biol Reprod, March 1, 2004; 70(3): 562 - 569. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 V. Sriraman and J. S. Richards Cathepsin L Gene Expression and Promoter Activation in Rodent Granulosa Cells Endocrinology, February 1, 2004; 145(2): 582 - 591. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. Charron, J. N. DeCerbo, and W. W. Wright A GC-Box Within the Proximal Promoter Region of the Rat Cathepsin L Gene Activates Transcription in Sertoli Cells of Sexually Mature Rats Biol Reprod, May 1, 2003; 68(5): 1649 - 1656. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. Charron, J. S. Folmer, and W. W. Wright A 3-Kilobase Region Derived from the Rat Cathepsin L Gene Directs In Vivo Expression of a Reporter Gene in Sertoli Cells in a Manner Comparable to That of the Endogenous Gene Biol Reprod, May 1, 2003; 68(5): 1641 - 1648. [Abstract] [Full Text] [PDF]
|
|
