| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Original Studies |
Division of Bone Diseases, WHO Collaborating Center for Osteoporosis and Bone Diseases, Department of Internal Medicine, University Hospital, 1211 Geneva 14, Switzerland
Address all correspondence and requests for reprints to: Dr. Danielle Manen, Division of Bone Diseases, Department of Internal Medicine, University Hospital, 1211 Geneva 14, Switzerland. E-mail: manen{at}cmu.unige.ch
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
). In the human gene, the synthesis of PTH-R is under the
control of three distinct promoters, P1, P2, and P3. The
5'-untranslated region (5'UTR) of the messenger ribonucleic acid (mRNA)
generated by P1 comprises the exons U1 and U2 spliced to exon S. The
5'UTR of the mRNA generated by P2 is formed by splicing exon U3 to exon
S. Finally, the proximal promoter P3 generates a 5'UTR including exon
U4 coupled to exon S (10). The regulation of these three promoters is
still poorly elucidated. P1 activity appears to be restricted to the
kidney, whereas P2 and P3 are active in a variety of tissues, including
kidney and bone cells (8, 10, 11, 12). P2 has been reported to be the only
promoter active in fetal tissues at midgestation (13). The recently
described promoter P3 drives the synthesis of abundant mRNA in kidney
and is also active in human osteoblastic SaOS-2 cells (10, 11). The
promoter P3 belongs to the family of TATA-less promoters that lie in
GC-rich regions and generate mRNAs highly structured in their 5'UTR. To
study the important elements contributing to the activity of P3 we used
reporter gene constructs containing various fragments of the
5'-sequence flanking the gene. We identified a region essential for P3
activity and containing two key Sp1 binding sites. In addition, we
found that the 5'-nontranslated exon U4 contains a strong negative
regulator element for PTH-R gene expression.
|
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Cloning, ligations, transformations in Escherichia coli XL1 blue strain, and DNA minipreparations were carried out as reported by Sambrook et al. (14). Plasmid DNA was prepared using QIAGEN kits (Basel, Switzerland). When used in transient transfections, the plasmid DNA was phenol-chloroform extracted and ethanol precipitated before use. Plasmid DNA was sequenced using a T7 sequencing kit (Pharmacia Biotech, Dubendorf, Switzerland). Unless otherwise indicated, PCR was performed using Taq DNA polymerase from QIAGEN. The reactions were carried out as recommended by the supplier, using Q solution. The temperature cycling protocol was 94 C for 1 min, 55 C for 20 s, and 72 C for 1 min for 35 cycles. The reactions were concluded by a 10-min elongation at 72 C. Oligonucleotides were purchased from MWG- Biotech GmbH (Ebersberg, Germany). All reagents were of the highest purity available from laboratory suppliers.
Reporter gene constructs
We inserted various DNA fragments from the 5'-flanking region of
exon S into the multiple cloning site of the promoterless firefly
luciferase expression vector pGL3-enhancer (Promega Corp.,
Zurich, Switzerland). The convention for sequence coordinates is that
+1 is the first base of the coding sequence in exon S (Fig. 1
).
5'-End deletions of P3. Plasmids P3–1 and P3–5 (Fig. 2A) are the previously described plasmids
P3a and P3b (11). The SacI-AvrII fragment was
cloned in pGL3-enhancer (pGL3-en.) and cut by SacI and
NheI, leading to plasmid P3–1. We used the oligos D2, D3,
and D4 as forward primers and the commercial primer GL-pr-2 as reverse
primer to generate PCR fragments (Table 1). The DNA template was plasmid P3–1.
The oligos D2, D3, and D4 contain an additional SacI site in
their 5'-end. We replaced the SacI-NdeI fragment
of P3–1 by the PCR fragments digested by SacI and
NdeI to obtain plasmids P3–2, P3–3, and P3–4 (Table 1 and
Fig. 2A). Plasmid P3–5 resulted from deletion of the restriction
fragment SacI-NdeI from P3–1.
|
|
). We replaced the
NdeI and BglII fragment of P3–1 by the PCR
fragments digested by NdeI and BglII to obtain
plasmids P3–6, P3–7, and P3–8 (Table 1 and Fig. 2A). Deletion of the
restriction fragment NdeI-BglII in plasmid P3–1
and P3–5 leads, respectively, to plasmids P3–9 and P3–10.
Site-directed mutagenesis. To mutate the distal Sp1 site
(Figs. 1A and 3A), we amplified the DNA
fragments located upstream and downstream from the Sp1 site using two
sets of primers: first D4 (see Table 1) and D64
(CTGCAGGGATCCCCCTCCCGCGCCGCCTTT) and then D65
(CTGCAGGGATCCGGGGCGGGCCGGGGGCGGGGGG) and Gl-pr2 (see Table 1). The DNA template was P3–4. Primers D64 and D65 contain a
BamHI site in their 5'-ends (underlined). The PCR
fragment (D4; D64) was digested by SacI and
BamHI, and the PCR fragment (D65; Gl-pr2) was digested by
BamHI and BglII. The fragments were ligated
together and inserted into the pGL3-en. vector cut by SacI
and BglII. Thus, in the resulting plasmid, P3–4 m1, the
wild-type sequence, GGGGGGCGGGGG, was replaced by the mutated
sequence, GGGGGGATCCGG, containing a BamHI
restriction site.
|
) and contains mismatches (BamHI site;
underlined) at the level of the site to be mutated. In the
resulting plasmid, P3–4 m2, the wild-type sequence (GGGGGGCGGGGG)
was replaced by the mutated sequence (GGGGGGATCCGG),
containing a BamHI restriction site. To construct the double mutant we used oligos D4 and D67, (CACATCCATATGGCCGGGCCCCGGATCCCCGGCCCGCCCCGGATCCCCCT). PCR was performed on plasmid P3–4 m1. We replaced the SacI-NdeI fragment of P3–4 by the PCR fragment digested by SacI and NdeI to obtain plasmid P3–4 m3, in which both Sp1 sites are mutated.
To mutate the ATG start codon of open reading frame 4 (ORF4; Figs. 1A and 4A) we amplified DNA fragments
located upstream and downstream of the ATG codon using two sets of
primers: first D4 (see Table 1) and D69
(GCAGCCTAGGCGCTGAGGGCGAGCGA) and then D70
GCAGCCTAGGCCCCGCGCCGGGCCCCGGGG) and Gl-pr2 (see Table 1).
The DNA template was P3–1. Primers D69 and D70 contain an
AvrII site in their 5'-ends. The PCR fragment (D4; D69) was
digested by NdeI and AvrII, and the PCR fragment
(D70; Gl-pr2) was digested by AvrII and BglII. We
replaced the NdeI-BglII fragment of P3–1 by the
PCR fragments digested and ligated together. In the resulting plasmid,
P3–1 m4 (Fig. 4A), the ATG start codon of ORF4 was changed into the
CTA triplet, creating an AvrII restriction site. In each
case the cloned PCR fragments were sequenced to verify the absence of
other mutations.
|
). (On request,
the authors will supply additional details on plasmid constructions.)
The pRL-TK vector (Promega Corp.) provides constitutive
expression of Renilla luciferase and was used in transfection assays as
an internal control.
|
Human osteosarcoma cells SaOS-2 were grown in DMEM/F-12 medium (Life Technologies, Inc., Basel, Switzerland) supplemented with 100 IU/mL penicillin, 100 µg/mL streptomycin, and 10% FCS (Amimed, Muttenz, Switzerland). Seventy-two hours after plating, the cells were seeded in 6-well tissue culture clusters at a density of 200,000 cells/well. The cells were transfected 24 h later, using the standard method of calcium phosphate-DNA precipitation in N,N-bis-(2-hydroxyethyl)-2-aminoethanesulfonic acid buffer (16). Cells were transfected with 2 µg PTH-R promoter constructs and 0.25 µg pRL-TK vector as an internal control. After 24 h in 3% CO2 at 35 C, the cells were washed twice in PBS without Mg2+ and without Ca2+ and put back to standard culture conditions for 48 h. We harvested the cells in 300 µL lysis buffer from the Dual-Luciferase Reporter Assay System (Promega Corp.), and we used 20 µL of the lysate to measure firefly and Renilla luciferase activities (the latter was used to normalize for transfection efficiency). We took the luciferase activity of the reporter plasmid P3–1 as the 100% value. Results are the mean ± SD of triplicate determinations and are representative of at least four independent experiments. Absolute values of the light units of firefly luciferase activity for the control groups are 5578 ± 620 in Exp 2–4 and 7975 ± 918 in Exp 5.
Gel retardation
The binding reaction mixture (15 µL final volume) contained approximately 2 ng 32P end-labeled DNA, 0.1 footprinting unit (fpu) purified Sp1 protein (Promega Corp.), and 50 ng poly(dI-dC) in binding buffer [50 mmol/L Tris-glycine (pH 9.3), 1 mmol/L ethylenediamine tetraacetate (pH 8), 200 µg/mL BSA, 70 mmol/L KCl, 7 mmol/L MgCl2, 3 mmol/L CaCl2, and 1 mmol/L ß-mercaptoethanol]. Incubation was carried out at 16 C for 20 min; 3 µL loading buffer (50% glycerol and 1 mg/mL BSA) were added before loading onto a 7.5% acrylamide nondenaturing gel [running buffer: 50 mmol/L Tris-glycine (pH 9.4) and 0.1 mmol/L ethylenediamine tetraacetate]. Electrophoresis was performed at 4 C for 3 h at 12 V/cm. The gel was dried and autoradiographed.
Computer analysis
A search for consensus transcription factor binding sites was performed using MatInspector V2.2 (17). Evaluation of the U4 RNA folding free energy was performed using Michael Zuker’s FOLDRNA program in the Wisconsin Package (version 9.1, Genetics Computer Group, Madison, WI).
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
In view of discrepancies between the two reported sequences of the
P3 promoter region (10, 11), we repeated the sequencing of some 600 bp
upstream of exon S, using our genomic clone. This analysis mostly
confirmed the sequence we have previously published (11), except for
three errors that are now corrected (at bp -367 a G was missing; at bp
-197 and 191 two additional T have been removed; Fig. 1). Sequencing
of our clone revealed the presence of a small open reading frame (ORF4,
bp -164 to -108; Fig. 1) in the U4 exon, shorter than that described
by Bettoun et al. (10).
Analysis of the P3 promoter region by progressive deletions
We previously reported on the promoter activity of the P3 region
(11). To localize within that region cis-acting elements
that are essential to the transcription of the gene, we fused various
fragments of the 5'-flanking region of exon S to the luciferase gene in
the pGL3 enhancer vector (Figs. 1 and 2). The relative luciferase
activities of the different constructs are shown in Fig. 2.
Progressive 5'-deletions resulted in decreased luciferase activity
(Fig. 2, plasmids P3–1 to P3–5). We found that the fragment between
bp -440 and bp -329 was essential to retain a significant level of P3
expression (plasmids P3–4 and P3–5); when the inserted genomic
fragment started at bp -329 (NdeI restriction site), no
luciferase activity was detected. Remarkable features of this region,
between bp -440 and -329, are a stretch of As and two canonical
binding sites for the transcription factor Sp1. We noticed that
additional elements located upstream of this essential region are
necessary for full expression of P3 activity (Fig. 2, plasmids P3–1 to
P3–4).
Starting from the reporter plasmid P3–1 we then performed 3'-deletions within exon U4. Luciferase activity was not changed when the fusion with the luciferase gene occurred at bp -66 (plasmid P3–6). When, however, the fusion occurred at bp -163 or -237 (plasmids P3–7 and P3–8), the reporter gene activity was markedly increased. These results suggest the existence of a negative regulatory element of PTH-R expression within exon U4. When the fusion with the reporter gene occurred at the NdeI restriction site in the genomic DNA, close to the 5'-end of exon U4 (plasmid P3–9), luciferase activity dropped markedly, indicating that the fragment located between the NdeI restriction site and the beginning of exon U4 contains elements important for P3 activity. Plasmid P3–10 resulted from combined 3'- and 5'-deletions and retained the P3 region between bp -440 and bp -329. It exhibited a low level of luciferase activity, which is consistent with all of the above-mentioned observations.
Mutations of the Sp1 sites located in the essential promoter region
As reported in the previous section, activity of the P3 promoter
needs the presence of cis-acting elements localized between
bp -440 and bp -329 (plasmids P3–4 and P3–5). By computer analysis
we found two perfect consensus binding sites for the transcription
factor Sp1 in this segment (Fig. 1). To evaluate the functional roles
of these Sp1 sites, we replaced them, singly and jointly, by
nonspecific sequences of equal size (plasmids P3–4 m1, P3–4 m2, and
P3–4 m3; Fig. 3A). When either site was mutated, reporter gene
expression decreased markedly; when both were mutated, P3 activity was
completely abolished (Fig. 3, A and B). These results demonstrate that
these Sp1 sites are needed for the activity of the P3 promoter in
SaOS-2 cells.
Bandshift assays
We monitored the binding of purified Sp1 protein to a DNA fragment
containing the two putative Sp1 binding sites, i.e. the
SacI-NdeI restriction fragment (bp -440 to
-329) of plasmid P3–4. We purified and 32P
end-labeled the wild-type fragment (from plasmid P3–4) and the
corresponding mutated fragment (from plasmid P3–4 m3) in which both
Sp1 binding sites were mutated. In a bandshift experiment the wild-type
fragment was shifted by 0.1 fpu purified Sp1 protein (Fig. 3C, lane 2).
Under the same conditions there was no detectable attachment of the
protein to the mutated fragment (lane 4).
Mutation of the ORF4 start codon
Results from the reporter gene constructs showed the presence of a
negative regulatory element located in exon U4 (Fig. 2, P3–1 compared
to P3–7 and P3–8). This inhibition could be due to either a decreased
level of transcription or an impaired translation of the messenger. To
determine whether transcription of ORF4 could impair translation of the
messenger we mutated the start codon of ORF4, replacing ATG by CTA. We
did not detect any significant change in luciferase expression (plasmid
P3–1 m4; Fig. 4), suggesting that translation of ORF4 is not
responsible for the low reporter gene expression.
The U4 region of PTH-R decreases expression of a heterologous gene
To analyze the effect of the U4 region on gene expression in a
different context, we constructed a reporter gene in which the fragment
of the 5'-flanking region of the PTH-R gene containing exon U4 was
inserted between a heterologous promoter and the luciferase-coding
region (Fig. 5). As heterologous promoter, we used the promoter of the
human Glvr-I gene, which codes for a phosphate transporter/retrovirus
receptor and is well expressed in the SaOS-2 cell line (15). The U4
region does not have any promoter activity by itself (Fig. 5, P3–5).
Insertion of the U4 fragment downstream of the Glvr-I gene promoter
markedly inhibited the expression of luciferase (Fig. 5). This
indicates that the negative effect of the U4 region is not specific to
the PTH-R P3 promoter.
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Transient transfection assays in the osteoblastic-like cell line SaOS-2, in which the PTH-R receptor is highly expressed (18), allowed us to identify genetic elements involved in the expression of the protein from the P3 promoter. We found that the DNA sequence between bp -440 and -329 possesses sites essential for P3 activity. This DNA fragment contains a stretch of As followed by two perfect consensus Sp1 binding sites, and we found that it does bind purified Sp1 protein in vitro. The functional roles of these two Sp1 binding elements were established by monitoring the reporter activity of constructs containing single or combined mutations of theses two sites. In each case we saw a significant loss of activity, demonstrating the key roles of these sites. The Sp1 protein belongs to a large family of zinc finger transcription factors interacting with GC box elements and capable of activating transcription in mammalian cells. Therefore, the identification of factors interacting in vivo with putative Sp1 binding sites is complicated by the multiplicity of such proteins (see Ref. 19 for a review). Our findings show that among the putative transcription factors that could be involved in P3 activity in the SaOS-2 cells, Sp1 is a good candidate.
Other features of this region, yet to be investigated, might be implicated in P3 activity. As such sequences are known to distort DNA structure, it is possible that the stretch of As participates in a specific topology necessary for P3 expression. Computer analysis also revealed the presence of a potential binding site for the activating protein-2 transcription factor between the Sp1 sites and exon U4, which might be implicated in P3 activity.
We detected the presence of a negative regulatory element located in exon U4. In our gene reporter system the luciferase gene expression is evaluated by measuring the amount of active luciferase synthesized. Therefore, the observed change in gene expression could reflect changes in mRNA levels or changes in the amount of protein produced from the messenger. As it has been reported that mRNA translation can be impaired by the presence of short ORF in the 5'UTR of transcripts (20, 21) we mutated the ATG start codon of ORF4. This mutation did not abolish the negative effect, eliminating this possibility. To better appreciate the intrinsic effect of the U4 inhibitor element we inserted the 5'-flanking region of PTH-R between a heterologous promoter and the luciferase reporter gene. There was a drastic drop in luciferase expression showing that the negative regulatory is not specific to the PTH-R P3 promoter.
As reported by Kozak (20), many mRNAs with GC-rich leader sequences
include mRNAs for a variety of proteins essential for the cell
viability such as GH receptors, protooncogenes, and others. She
suggested that the production of these critical regulatory protein is
throttled at the level of translation. For instance, the
N-methyl-D-aspartate receptor is a
typical example of down-regulation of translation by the presence of a
highly structured 5'UTR (22). For PTH-R, the mRNA generated by the P3
promoter presents an encumbered leader sequence: computer analysis
revealed for the folding of U4 RNA an elevated free energy in the range
of 
G of -80 kcal/mol. We, therefore, favor the hypothesis of a
translation impaired by mRNA secondary structures located in exon U4.
We cannot, nevertheless, exclude the presence in the leader sequence of
a site responsible for a transcriptional down-regulation.
In conclusion, our data showed that in SaOS-2 cells the activity of promoter P3 depends on the integrity of two Sp1 binding sites. In addition, we demonstrated the presence of a negative regulatory element of PTH-R expression in the 5'-untranslated exon 4 generated by the P3 promoter.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
Received April 12, 2000.
Revised June 5, 2000.
Accepted June 13, 2000.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
 |
|
 |
|
  P.-Q. Liu, S. Tan, M. C. Mendel, R. J. Murrills, B. M. Bhat, B. Schlag, R. Samuel, J. J. Matteo, R. de la Rosa, K. Howes, et al. Isogenic Human Cell Lines for Drug Discovery: Regulation of Target Gene Expression by Engineered Zinc-Finger Protein Transcription Factors J Biomol Screen, June 1, 2005; 10(4): 304 - 313. [Abstract] [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 |
