| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
ARTICLES |
Departments of Medicine (L.C., S.B., A.J.M., R.P.R., G.N.H.), Physiology (D.G.W., J.H.W., G.N.H.), and Human Genetics (G.NH.), McGill University and Royal Victoria Hospital, Montréal, Québec H3A 1A1, Canada
Address all correspondence and requests for reprints to: Dr. Geoffrey N. Hendy, Calcium Research Laboratory, Room H4–67, Royal Victoria Hospital, 687 Pine Avenue West, Montréal, Québec H3A 1A1, Canada. E-mail: gnhendy{at}medcor.mcgill.ca
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
We previously isolated and characterized the human (h) CgA gene (5). Its organization is very similar to that of the mouse (6) and bovine (7) genes, and the proximal promoter region is strongly conserved across species (5, 6, 7, 8). We (5) and others (6) showed that 2-kb of the 5'-flanking region of the human and mouse CgA genes conferred neuroendocrine cell-specific transcriptional activity upon reporter genes in transient transfection studies. We defined a minimal promoter construct containing 87 bp of the human CgA gene (nucleotides -55 to +32 relative to the transcriptional start site) directed neuroendocrine cell specific activity of a chloramphenicol acetyl transferase (CAT) gene. This portion of the hCgA gene promoter contains a consensus cAMP response element (CRE) and a TATA-box motif. These two elements are conserved in the CgA gene across species and are also found in the gene promoters of other members of the granin family such as chromogranin B (CgB) (9) and secretogranin II (SgII) (10, 11). This indicated that the CRE in addition to the TATA box may play a critical role in the expression of the granins. Evidence in support of this has been presented by Wu et al. (12) for the mouse (m) CgA gene and by us (13, 14) for the hCgA gene.
In the present study, we carried out gene transfer experiments with hCgA gene promoter/CAT reporter gene constructs in a variety of human and rodent neuroendocrine cell lines to further understand the role of the CRE and another conserved sequence (the TG-box) just downstream of the CRE in the minimal promoter region in the regulation of CgA gene expression. Our results demonstrate that the CRE plays a key role in basal as well as in cAMP-stimulated expression of the hCgA gene in neuroendocrine cells, whereas the TG-box plays a much lesser role and only contributes to basal transcriptional activity in selected neuroendocrine cell types. However, we found by gel retardation assays and immunoblot analyses equivalent amounts of PKA-sensitive CREB in neuroendocrine and nonendocrine cells. Expression of the CREB-binding protein (CBP) was similar in neuroendocrine and nonendocrine cells. In addition, we have determined that the distal promoter contains additional positively acting, cell-specific regulatory elements that function independently of the CRE. Finally, we have identified a sequence downstream of the transcriptional start site, which is not conserved across species, and contributes positively to basal expression of the CgA/CAT reporter gene.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Plasmid constructs
Human CgA promoter deletion constructs, p2300CAT, p700CAT,
p450CAT, p260CAT, and p55CAT contained hCgA gene fragments, from 2316
bp to 55 bp upstream to 32 bp downstream of the start of transcription,
cloned into the CAT expression vector, pBLCAT3. These constructs were
prepared as described previously (5). The hCgA promoter constructs,
p55(-CRE)CAT, p55(-TG)CAT, p55(-CRE/-TG)CAT, and p31CAT were created by
ligating BamHI-XhoI fragments generated by PCR
into pBLCAT3. The forward primer for the p55(-CRE)CAT fragment
5'-AAGTCAGGATCCGCCAGTGTCATTTCCGG-3', is composed of six nucleotides
flanking a BamHI site followed by nucleotides -53 to -37,
but with nucleotides -51 to -48 (TGAC) replaced by CAGT. The forward
primer for the p55(-TG)CAT fragment
5'-AACTAGGGATCCGCTGACGTCATTGATCTGGTC-3', is composed of six nucleotides
flanking a BamHI site followed by nucleotides -53 to -33,
but with nucleotides -41 to -37 (TCCGG) replaced by GATCT. The
forward primer for the p55(-CRE/-TG)CAT fragment
5'-AACGCGGGATCCGCCAGTGTCATTGATCTGGTCG-3' is composed of six nucleotides
flanking a BamHI site followed by nucleotides -53 to -32,
but with nucleotides -51 to -48 (TGAC) and nucleotides -41 to -37
(TCCGG) replaced by CAGT and GATCT, respectively. The forward primer
for the p31CAT fragment 5'-TTTCCGGGATCCGGGTATATA-3' is composed of six
nucleotides flanking a BamHI site followed by nucleotides
-31 to -23. For all fragments the reverse primer was
5'-CACGGGGCTCGAGCAC-3', which is the complement of +25 to +41 and
contains the XhoI site. Fragments were amplified by PCR,
cleaved with BamHI and XhoI, gel-purified and
ligated into BamHI-XhoI cut pBLCAT3.
Human CgA promoter constructs p2300(-CRE)CAT and p700(-CRE)CAT were the same as constructs p2300CAT and p700CAT, respectively, except that in the CRE motif, the sequence TGAC was mutated to CAGT. This mutation in the CRE was the same as that made in the shorter hCgA-CRE mutant constructs. Construct p700(-CRE)CAT was created by ligating a HindIII-XhoI fragment of 753 bp, generated by PCR, into HindIII and XhoI cleaved pBLCAT3. Recombinant p2300(-CRE)CAT was created by ligating a HindIII-HindIII fragment of 1481-bp (cleaved from p2300CAT, spanning from the HindIII site of pBLCAT3, just upstream from the hCgA insert, to the HindIII site at nucleotide -721 of the hCgA promoter) into P700(-CRE)CAT digested with HindIII. The HindIII-XhoI fragment of 753 nucleotides was generated by site-directed mutagenesis using PCR and spanned nucleotides -721 to +32 of the hCgA gene.
Human CgA gene constructs 3' extended to nucleotide +198 were made as follows. Recombinants p2300/198CAT, p450/198CAT, p55/198CAT, and p31/198CAT were made by inserting a hCgA gene fragment (+33 to +198) into the XhoI site at hCgA gene position +32 of the plasmids p2300CAT, p450CAT, p55CAT, and p31CAT, respectively. Construct p55/198–33CAT was made by inserting the +33 to +198 fragment in the antisense direction, into the XhoI site at hCgA gene position +32 of p55CAT. Recombinant p33–198/55CAT was made by inserting the hCgA +33 to +198 fragment in the sense direction, into the SalI site 28 bp upstream to the -55 to +32 hCgA sequence of p55CAT. The hCgA gene +33 to +198 fragment was created by PCR. The forward primer was 5'-CTGCAGTGCTCGAGCCCCGT-3'. This sequence is hCgA gene +20 to +39 with an additional XhoI site. The reverse primer was 5'-AGTTGCCTCGAGCAGCTGGCGGTGT-3'. This comprises six nucleotides flanking an XhoI site, followed by complementary hCgA gene sequence +194 to +182. The PCR amplified fragment was digested with XhoI, gel-purified, and ligated into hCgA-CAT plasmids as described above.
Plasmid DNAs were prepared by the alkaline lysis/cesium chloride gradient method or by alkaline lysis followed by affinity chromatography (Qiagen Inc., Chatsworth, CA). All hCgA-CAT constructs were verified by dideoxynucleotide sequencing.
Cell culture and transient transfection
Five neuroendocrine cell types were used: AtT20, a mouse
pituitary corticotroph cell line; BEN, a human lung carcinoma cell
line; GH4C1, a rat pituitary cell line; PC-12, a rat pheochromocytoma
cell line and TT, a human medullary thyroid carcinoma cell line. Two
nonendocrine cell lines were used: NIH 3T3, a mouse fibroblast cell
line and Rat-2, a rat fibroblast cell line. BEN cells were maintained
in RPMI with the addition of 10% FBS and L-glutamine.
PC-12 cells were maintained in DMEM with the addition of 10% FBS and
5% horse serum. TT cells were maintained in Ham’s F12 medium with
10% FBS. All other cell lines were maintained in DMEM with 10% FBS.
All maintenance media contained 100 U/ml penicillin, 100 µg/ml
streptomycin, and 250 ng/ml amphotericin B.
For transfections, cells were trypsinized, seeded at 5 x 105 cells/35 mm tissue culture dish and incubated overnight. Medium was aspirated and cells were washed twice with serum-free DMEM. Transfections with 6 µg of the phCgACAT constructs, with the positive control SV40 early region promoter-driven plasmid pSV2CAT or the positive control Rous Sarcoma Virus LTR promoter/enhancer-driven plasmid pRSVCAT and the negative control promoterless plasmid, pBLCAT3, were done in parallel experiments in all cell lines. Each transfection involved cotransfection with 2 µg of control plasmid, pCH110, in which the gene for ß-galactosidase is under the control of the SV40 early promoter. DNA was transfected by lipofection (15) in the majority of cells. However, TT cells (10 x 106 cells) were transfected by electroporation with a single pulse of 250 V/4 mm, 960 microfarads using the Bio-Rad (Mississauga, Ontario, Canada) Cell Porator with an extended capacitance unit. Forty-eight hours after transfection, cells were washed once with ice-cold PBS and harvested in 1 ml of PBS by scraping into 1.5 ml Eppendorf tubes. Cells were spun at 500 x g at 4 C for 7 min and resuspended in 250 µl of 0.5 M Tris-HCl, pH 7.5. Cells were lysed by three freeze-thaw cycles, and the DNA and cell debris were removed by centrifugation at 16,000 x g for 20 min at 4 C. Cell lysates were assayed for protein using the BioRad protein assay kit and assayed for ß-galactosidase and CAT activity as described below. CAT activities were normalized to ß-galactosidase activity, and the data are presented as the fold-activity relative to that of the promoterless pBLCAT3 vector.
CAT and ß-galactosidase assays
CAT assays were performed with 5 to 20 µg of protein
essentially as described (16). Reactions were incubated at 37 C for 0.5
to 4 h. For the longer incubations, 1 µg of 4 mM
acetyl-CoA was added after 2 h of incubation. After thin layer
chromatography of the samples, acetylation of chloramphenicol was first
visualized with autoradiograph film and then quantitated by liquid
scintillation counting. The CAT measurements were within the linear
part of the response range. ß-galactosidase assays were done by
measuring an increase in optical density at 420 nm using
O-nitrophenyl-ß-D-galactopyranoside as
substrate (16). For a single experiment duplicate or triplicate
estimations were made, and the data were expressed as mean ±
SEM of three to four independent experiments.
DNA mobility shift assays
Cells were harvested for gel retardation assays by washing with
5 ml ice-cold PBS followed by scraping the cells in 1 ml PBS. Cells
were centrifuged at 2,000 rpm for 10 min at 4 C, and the pellets were
resuspended in 100 µl high salt extraction buffer [25 mM
Tris (pH 7.9), 0.3 mM DTT, 0.1 mM EDTA, 420
mM NaCl, 10% glycerol]. Cells were lysed by three cycles
of freeze/thaw and cell debris removed by centrifugation at 10,000 rpm
for 10 min at 4 C. Cell nuclear extracts were made by the protocol of
Dignam et al. (17).
For gel retardation assays, cell nuclear extracts were added to
incubations as indicated in the figures. Samples were incubated for 15
min on ice in 10 µl 25 mM Tris-HCl (pH 8.0), 1
mM DTT, 50 mM KCl, 20% glycerol containing 1
µg poly (dI-dC) and then for a further 20 min at 23 C after the
addition of 50,000–100,000 cpm (5–10 fmol) of
32P-end-labeled double-stranded oligonucleotide. The
nucleotide sequences of the CgA-CRE, CgA-CREmut 1, CgA-CREmut 2,
CgA-CREmut 3, SS-CRE, AP-1, and SP-1 oligonucleotides are shown in
Table 1
. Specific antibodies to the
CRE-binding protein (CREB) (New England Biolabs, Beverly, MA) were also
added to some reactions. Samples were loaded on 5% polyacrylamide gels
equilibrated in 25 mM Tris-HCl/31.1 mM boric
acid/1 mM EDTA (pH 8.0) and electrophoresed at 8 V/cm. Gels
were dried before autoradiography.
|
RT. Five-microgram RNA samples were reverse transcribed with recombinant superscript II RNAse H (GIBCO-BRL) using oligo (dT) 15–18 (Pharmacia) in a total volume of 20 µl. Five microliters of the RT mixture was subjected to standard PCR procedures.
Samples were assessed for CBP using the primers 5'-AGTGGAATTCAAAACACAATTGGTTCTGTTGGTGCAGGGCA-3' (forward) and 5'-TAAAGCTGGCTGGTTACCCAGGATGCCTTGCTTATGTAAACG-3' (reverse). These sequences are highly conserved between human, mouse, and rat CBP.
The ß-actin primers generated a 500-bp fragment whereas the CBP primers amplified a 704-bp fragment encoding a.a. 450 to 684 of the CREB binding region. The PCR mixture contained 2 mM MgCl2, 0.2 mM of each dNTP, 10 mM Tris-HCl (pH 8.3), 50 mM KCl, 25 pmol of forward and reverse primers and 2.5U/100 µl of Taq polymerase (GIBCO-BRL). Thirty cycles (94 C, 45 sec; 58 C, 45 sec; 72 C, 1 min) were performed with a programmable thermocycler (GeneAmp PCR system 9600, Perkin-Elmer Cetus, Mississauga, Ontario, Canada). Aliquots were taken after 15, 20, and 25 cycles.
Western immunoblots
Cells were lysed in triple detergent buffer [50 mM
Tris-HCl (pH 8.0), 150 mM NaCl, 0.02% NaN3,
0.1% SDS, 100 µg/ml PMSF, 1 µg/ml aprotonin, 0.1% NP-40, 0.5%
sodium deoxycholate for 5 min at 0 C. The cell lysates were spun at
12,000 x g, 2 min, 4 C and the supernatants were
stored at -80 C. Aliquots were electrophoresed through 12% SDS-PAGE
and blotted onto polyvinylidene difluoride membranes (Bio-Rad).
Membranes were rinsed in TBST (10 mM Tris-HCl, pH 8.0, 150
mM NaCl, 0.05% Tween 20), blocked with 5% dried milk
powder in TBST for 1–2 h and incubated with antisera for either
phospho-specific CREB (Ser 133) or total CREB (Ser 133 phosphorylation
state-independent) according to manufacturer’s instructions (NEB Inc.,
Beverly, MA) or antibodies directed against epitopes in either the
NH2-terminus or COOH-terminal portion of CBP (Upstate
Biotechnology, Lake Placid, NY) or antihuman ß-tubulin monoclonal
antibody (Cedarlane Laboratories Limited, Hornby, Ontario, Canada).
Antibody-antigen complexes were detected by chemiluminescence using the
Phototope-Star Western blot detection kit (NEB Inc., Beverly, MA), or
the LumiGlo chemiluminescent substrate (Kirkegaard and Perry
Laboratories, Gaithersburg, MD).
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
),
those of the human and rodent CgA genes are not (Fig. 1B). In a
previous study of the mouse (m) CgA gene promoter (8), it was found
that inclusion of exon I sequence extending to close to the translation
initiation codon in promoter constructs led to enhanced reporter gene
activity in the single cell line (of neuroendocrine origin) tested. In
our present study, we examined this issue further by extending hCgA
gene promoter constructs to position +198, just upstream of the
translation initiation codon. Transient transfection studies were
carried out in human neuroendocrine BEN cells, rodent neuroendocrine
GH4C1 cells, and the nonendocrine Rat2 cell line using the series of
constructs shown in Fig. 2A.
|
|
).
Importantly, the p31/198CAT construct demonstrated only baseline
activity. Therefore the +33 to +198 sequence does not function
independently of the proximal promoter elements but rather, cooperates
with them to augment their activity. Enhanced activity of the shortest
3'-extended construct (p55/198CAT), but none of the others, was seen in
the Rat2 nonendocrine cell line (Fig. 2B).
In BEN cells, the stimulatory effect of +33 to +198 was abolished or
greatly reduced in constructs in which the sequence was either inverted
(p55/198–33CAT) or moved upstream of position -55 (p33–198/55CAT).
This suggests that this element does not function as a classical
transcriptional enhancer in these cells. It is likely therefore that
the major effect of the +33 to +198 sequence in BEN cells is either to
enhance the stability of the messenger RNA (mRNA) or to increase its
ability to be translated. However, in GH4C1 cells, the +33 to +198
sequence retained 
50% of its activity after inversion or placement
upstream of the CRE. This suggests that this element may act as an
enhancer to regulate transcription in certain cell types.
Neuroendocrine cell-specific transcriptional activity of the human
CgA gene promoter
Putative binding sites for transcription factors within the
5'-flanking region of the hCgA gene are shown in Table 2. Constructs containing portions (from
-2300 to -31) to +32 of the hCgA gene promoter inserted upstream of
the CAT gene (Fig. 3)
were transiently transfected into a variety of neuroendocrine and
nonendocrine cell lines to determine the minimal portion of the
5'-flanking region of the hCgA gene necessary for neuroendocrine
cell-specific transcriptional activity. Constructs with 2300 bp to 55
bp of 5'-flanking sequence were several-fold more active than a
promoterless CAT construct in all neuroendocrine cell lines (Fig. 3B),
but showed activity no greater than pBLCAT3 in nonendocrine cells (Fig. 3C). The p31CAT construct, which retains the TATA-box but lacks the
CRE-containing region was inactive in neuroendocrine as well as
nonendocrine cells (Fig. 3, B and C). The observation that the hCgA-CAT
constructs were inactive in nonendocrine cells indicated that the
neuroendocrine cell-specific activity is not due to a nonendocrine cell
repressor mechanism.
|
|
), indicating that portions of the distal promoter may
have modest enhancer or repressor activity. However, no consistent
pattern was observed among cell types, suggesting that these effects
are cell specific. Most importantly, these observations emphasized the
fact that for all neuroendocrine cell-types the proximal promoter, -55
to -32 sequence containing the CRE, was critical for basal
transcriptional activity.
The CRE is critical for basal transcriptional activity in
neuroendocrine cells
The hCgA gene promoter from -55 to +1 is shown in Fig. 4A. In addition to the TATA-box, this
region contains the CRE (-51 to -44) and another well conserved
sequence designated the TG-box (-43 to -34) (see Fig. 7 in Ref.5).
To determine the importance of these motifs in basal transcriptional
activity, hCgA (-55 to +32)CAT constructs mutated in either the CRE,
or the TG-box, or both (Fig. 4A) were transiently transfected into
neuroendocrine and nonendocrine cell-lines. In all neuroendocrine cell
lines, mutation of the TG-box [p55(-TG)CAT] either had no effect (in
BEN, GH4C1, and AtT20 cells), or resulted in an approximately 2-fold
reduction in activity (PC12 and TT cells) (Fig. 4B). However, in all
cell lines mutation of the CRE [p55(-CRE)CAT] or combined
mutation of the CRE and TG-box [p55(-CRE/-TG)CAT]
dramatically reduced transcriptional activity (Fig. 4B). In
nonendocrine Rat2 (Fig. 4B) and NIH3T3 (data not shown) cells the
wild-type and mutated hCgA constructs were no more active than pBLCAT3.
These results emphasized the critical role played by the CRE, but not
the TG-box, in basal transcription in all neuroendocrine cells
tested.
|
|
Role of the CRE in basal transcriptional activity in the
5'-extended hCgA promoter
Having shown that the CRE is important for neuroendocrine
cell-specific transcriptional activity in the context of the minimal
promoter construct, p55CAT, we next assessed the importance of the CRE
in 5'-extended constructs. Human CgA (-2300 to +32) and (-700 to +32)
gene promoter constructs mutated in the CREs (see Fig. 5A) were transiently transfected into
neuroendocrine and nonendocrine cell lines. In all neuroendocrine cell
lines tested the mutated CRE constructs [p2300(-CRE)CAT and
p700(-CRE)CAT] were less active than the wild-type p2300CAT and
p700CAT constructs (Fig. 5B). The relative reductions in activity were
50% in BEN cells, 65% in AtT20 cells, and 75% in GH4C1 cells (Fig. 5B). In nonendocrine Rat2 cells, the activities of wild-type p2300CAT,
p700CAT, mutants p2300(-CRE)CAT, and p700(-CRE)CAT were similar to
that of the promoterless construct, pBLCAT3 (Fig. 5B). We therefore
concluded that the CRE is important for basal activity in
neuroendocrine cells even in the context of the 5'-extended hCgA
promoter. The results also indicate that other elements between -700
and -55 act independently of the CRE, and contribute to full basal
activity to varying extents depending upon the particular
neuroendocrine cell type.
|
). The control pSS-70 construct
containing a CRE, showed high basal activity which was only slightly
stimulated by 1 mM dibutyryl cAMP (Fig. 6A).
|
). In an experiment in which 1
mM dibutyryl cAMP stimulated transcriptional activity of
p55CAT over 2-fold, addition of the PKA inhibitor reduced activity to
46% of unstimulated control (Fig. 6C). Neither 1 mM
dibutyryl cAMP nor 10 µM H-89 had any effect on the
activities of p31CAT and pBLCAT3. Taken together, our results show that
the protein kinase A pathway is important for both basal and stimulated
expression of the hCgA gene under these conditions.
Similar protein-DNA complexes form on the hCgA CRE in
neuroendocrine and nonendocrine cells
To determine whether the same or different protein-DNA complexes
form on the hCgA gene CRE in neuroendocrine and nonendocrine cells, we
performed a series of gel retardation assays. As shown in Fig. 7A similar protein-DNA complexes formed
with nuclear extracts from neuroendocrine GH4C1, TT and nonendocrine
NIH3T3, Rat-2 cell lines and a [32P]-labeled CgA-CRE-WT
oligonucleotide. Similar results were obtained with extracts from
neuroendocrine AtT-20, BEN, PC-12, and nonendocrine COS-7 cells (data
not shown). Addition of CREB antibody supershifted the same predominant
complex in both neuroendocrine and nonendocrine cells. The specificity
of complex formation was demonstrated by the inability of
[32P]-labeled CRE mutant oligonucleotides to form the
same protein-DNA complexes as [32P]-labeled CRE-CgA-WT
oligonucleotide with neuroendocrine AtT-20 and nonendocrine Rat-2
extracts (Fig. 7B). Similar results were obtained with neuroendocrine
GH4C1 nuclear extracts (data not shown). When several oligonucleotides
containing either CRE, CRE-like, or unrelated sequences were tested for
their ability to compete with the [32P]-labeled
CgA-CRE-WT oligonucleotide only the somatostatin gene CRE was able to
do so as shown with neuroendocrine BEN and PC-12 extracts in Fig. 7C.
Similar results were obtained with neuroendocrine AtT-20 and GH4C1
cells and nonendocrine Rat-2 cells.
Basal and forskolin-stimulated phosphorylated CREB is expressed
equivalently in neuroendocrine and nonendocrine cells
Gel retardation experiments suggested that similar amounts of CREB
were expressed in neuroendocrine and nonendocrine cells. These
observations were supported by Western blot analysis (Fig. 8A). Similar amounts of total CREB and
phospho-CREB (Ser 133) were found in neuroendocrine AtT-20, BEN, GH4C1,
PC-12, and nonendocrine N1H3T3 and Rat-2 cells. The basal levels of
phospho-CREB (Ser 133) were markedly stimulated in all cells by
combined forskolin and IBMX treatment and, as shown in Fig. 8B, this
stimulation was blocked by H89, a PKA-selective inhibitor. Thus, we
find no correlation between cell-specific CgA promoter activity and
levels of functional CREB.
|
). Protein levels were assessed by
Western blotting using antibodies specific to the N-terminal or
C-terminal regions of CBP. Abundant levels of full length CBP (265 kDa)
were detected with both antibodies (Figs. 10, A and B) in neuroendocrine (AtT-20,
BEN, GH4CI, PC-12, TT) and nonendocrine cells (Cos-7, N1H3T3, Rat-2,
Ros).
|
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Mutation of the CRE in the minimal promoter construct p55CAT resulted in complete loss of the neuroendocrine cell-specific activity. In 5'-extended constructs, p2300CAT and p700CAT, mutation of the CRE resulted in a 50–75% decrease in basal activity in neuroendocrine cells. This emphasizes the important role played by the CRE in basal transcription but also suggests that other elements between -700 and -55 of the human CgA gene may act independently of the CRE to contribute to full basal activity in some neuroendocrine cell types. The importance of the CRE was also shown by the demonstration that the -55 to -32 sequence confers strong basal neuroendocrine-cell specific transcriptional activity on a heterologous adenovirus-2 major late promoter driving a CAT reporter gene, and this activity is stimulated further by cAMP (our unpublished results).
Human CgA gene transcription was stimulated by dibutyryl cAMP in the neuroendocrine BEN cell line and this effect was abolished by mutation of the CRE. In addition, the stimulation was completely blocked by addition of a PKA inhibitor, which also had a marked inhibitory effect on basal activity. The relatively modest induction of hCgA gene expression in BEN cells by cAMP is likely to be a reflection of a system already at high basal activity, and therefore unable to be stimulated further to any great extent. In fact, the majority of studies that have examined the responsiveness of CgA biosynthesis to cAMP or forskolin, which directly stimulates adenylate cyclase, have either shown a modest increase (7, 20) or no effect (20, 21, 22, 23). An exception to this is a recent study of the mouse CgA gene in which up to several-fold increases in its expression were noted in neuroendocrine PC12 cells treated with dibutyryl cAMP (12). This may be due to differences in responsiveness of one cell type relative to another, or to the specific tissue culture conditions employed. Therefore, as is the case for other genes like those for tyrosine hydroxylase (24), prohormone convertase 1 (25), prodynorphin (26) SgII (27), and renin (28, 29) basal, and tissue-specific transcriptional activity of the CgA gene is dependent upon the CRE.
A large number of proteins, which all belong to the basic-leucine zipper class of transcription factors, bind to the consensus CRE sequence, and modulate transcriptional activity (30, 31, 32). These include the cAMP response element binding protein (CREB), activating transcription factors (ATFs), and cAMP response element modulator (CREM). Activation of the PKA-pathway leads to phosphorylation of CREB/ATF-1 family members (e.g. phosphorylation of CREB at Ser 133), which is required for them to activate transcription. In our study, the results obtained with the PKA selective inhibitor H89, support those obtained with the mutated CRE constructs, suggesting that the activity comes from phosphorylated CRE-binding proteins. It is noteworthy that mutation of the CRE or addition of H89 lead to similar reductions in hCgA gene expression. This suggests that ubiquitously expressed PKA-regulated CRE-binding proteins such as CREB or ATF-1, are important for CgA gene expression. It is likely, however, that no single CRE-binding protein is solely responsible for mediating CgA gene transcription.
CREB binds the CREs of mouse CgA (12), mouse CgB (33) and human CgA (the present studies, and Ref.34) genes and ATF-1 binds the human CgA gene (unpublished data). Our gel retardation assays showed that the protein-DNA complexes formed with an oligonucleotide spanning the hCgA gene CRE and nuclear extracts from neuroendocrine and nonendocrine cells were similar with no clear-cut differences between cell types. The predominant complex was supershifted with CREB antibody in all cells, and the specificity of binding to the CRE was shown by the inability of CRE mutants to form the same complex and competition experiments with other CRE, AP-1 and SP-1 containing oligonucleotides. Nolan et al. (34) found similar CREB-containing protein-DNA complexes with an oligonucleotide spanning the human CgA gene promoter CRE and nuclear extract from neuroendocrine BEN cells but did not examine nonendocrine cell extracts. Wu et al. (12) compared results in gel retardation assays with an oligonucleotide spanning the mouse CgA gene promoter CRE and nuclear extracts from neuroendocrine AtT20 and nonendocrine N1H3T3 cells. They found that the entire complex formed with AtT20 nuclear extract supershifted to a new single complex of slower mobility with CREB antibody, whereas only a portion of the complex formed with N1H3T3 nuclear extract was supershifted by CREB antibody, and two supershifted complexes were observed. Therefore, while there is general agreement that the predominant protein-DNA complexes formed with either neuroendocrine or nonendocrine nuclear extracts and oligonucleotides spanning either the human or mouse CgA gene promoter CRE contains CREB, the importance of selective binding to the CRE by other factors in a nonendocrine cell as suggested by Wu et al. (12) was not supported in our study using several different neuroendocrine and nonendocrine cells. Recent studies on the neuroendocrine cell-specific expression of the proprotein convertase 1 (PC1) gene suggest that the CRE-elements in its proximal promoter bind novel neuroendocrine factors in addition to known CRE binding proteins like CREB (35).
The mechanism of tissue specificity of CgA gene transcription potentially could be due to either the amount or phosphorylation status of CREB, or the relative amount or activity of so-called coactivators or cointegrators such as CBP, which are recruited by phosphorylated CREB, and couple this transactivation to the basal transcriptional machinery. In the present study, we have excluded the former possibility by demonstrating that there was no difference in total CREB or in phosphorylated CREB levels in neuroendocrine and nonendocrine cells. Moreover, phosphorylated CREB was increased by forskolin, and the effect of this treatment was blocked by a PKA-inhibitor in both types of cells. With respect to the latter possibility, we show that expression of CBP at both the mRNA and protein levels is similar in neuroendocrine and nonendocrine cells, thus making it unlikely that the amount of CBP is the critical factor in determining neuroendocrine cell expression of CgA. However, it will be important in future studies to assess whether in nonendocrine cells there is a block in the correct association of CBP with the transcriptional machinery.
The human and bovine CgA genes show almost perfect homology in the
5'-flanking sequence between the CRE and TATA-box, which we have
designated the TG-box, suggesting that this region may be important in
binding trans-acting factors which act either independently
or in association with proteins bound to the CRE. The rodent genes also
exhibit very high homology with this TG-box region just downstream of
the CRE; however, they also have an insertion of more than twenty
nucleotides just upstream of the TATA-box relative to the human and
bovine genes. We found that mutation of the TG-box in the human gene
had a modest or no effect on transcriptional activity in either
neuroendocrine or nonendocrine cells. In the mouse CgA gene, mutation
of this region caused a 50% reduction in expression in PC12 and AtT20
cells (21). This can be compared with our present study in which
mutation of the TG-box had either no effect (BEN, GH4C1, AtT20, and
Rat2 cells) or resulted in a 50% decrease in transcriptional activity
(TT and PC12 cells) (see Fig. 2B). In all neuroendocrine cells, with
the exception of TT cells, mutation of the CRE alone was sufficient to
reduce transcriptional activity to basal, indicating that the promoter
effects of the CRE predominated over those of the TG-box. However, in
TT cells mutation of both the CRE and TG-box was required to obtain
complete reduction to basal transcriptional activity. Thus, while the
CRE is the preeminent modulator of basal CgA gene expression, a
contribution is also made by the TG-box in some neuroendocrine cell
types.
Transcriptional regulation of the human CgA gene has been studied by
others (19, 34). In an initial report that focused on distal sequences
from -726 to -455, Nolan et al. (19) noted that in
transient transfection studies in neuroendocrine BEN cells, whereas
hCgA/GH reporter constructs containing sequence from -726 to -590
demonstrated high basal activity, deleted constructs from -566 to
-455 showed 5-fold lower activity. Constructs deleted to positions
closer to the transcriptional start site were not studied. These
authors concluded at this time (19), that elements within the -590 to
-566 sequence were important for basal transcriptional activity. In a
later study (34), these authors suggested that the sequence from -570
to -555, rather than sequences upstream of -566 as emphasized in
their initial report (19), acted as a specific neuroendocrine cell
enhancer. It can be concluded from our studies that other elements
between -700 and -55 may contribute either enhancer or repressor
activities but, with the exception of BEN cells, the magnitude of these
effects relative to overall promoter activity is slight, and the
sequences involved vary from one neuroendocrine cell type to the other
(see Fig. 5B). Similar conclusions can be drawn from studies of the
mouse CgA gene promoter (12).
In summary, we have shown that the CRE of the human CgA gene plays a pivotal role in mediating both the basal and cAMP-induced expression of this gene in neuroendocrine cells. Other elements that play a lesser role include the TG-box lying just downstream of the CRE and more distal promoter sequences which may contribute to full transcriptional activity in particular neuroendocrine cell types. The mechanism of tissue-specific control mediated by the CRE is presently unknown. It could be due to a PKA-responsive CRE-binding protein other than CREB in neuroendocrine cells, expression of a repressor molecule that binds CREB in nonendocrine cells, or to the relative amount or activity of coactivators or cointegrators other than CBP, which are required to couple transactivators to the basal transcriptional machinery (36, 37, 38).
|
Acknowledgments |
|---|
|
Footnotes |
|---|
2 These authors contributed equally to this study.
3 Supported in part by the Royal Victoria Hospital Research
Institute.
4 Chercheur-boursier of the Fonds de la Recherche en Santé du
Québec (FRSQ).
5 Scientist of the Medical Research Council of Canada.
Received August 18, 1997.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
 |
|
 |
|
  X. Deng, H. Liu, J. Huang, L. Cheng, E. T. Keller, S. J. Parsons, and C.-D. Hu Ionizing Radiation Induces Prostate Cancer Neuroendocrine Differentiation through Interplay of CREB and ATF2: Implications for Disease Progression Cancer Res., December 1, 2008; 68(23): 9663 - 9670. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. Leja, H. Dzojic, E. Gustafson, K. Oberg, V. Giandomenico, and M. Essand A Novel Chromogranin-A Promoter-Driven Oncolytic Adenovirus for Midgut Carcinoid Therapy Clin. Cancer Res., April 15, 2007; 13(8): 2455 - 2462. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 G. N. Hendy, T. Li, M. Girard, R. C. Feldstein, S. Mulay, R. Desjardins, R. Day, A. C. Karaplis, M. L. Tremblay, and L. Canaff Targeted Ablation of the Chromogranin A (Chga) Gene: Normal Neuroendocrine Dense-Core Secretory Granules and Increased Expression of Other Granins Mol. Endocrinol., August 1, 2006; 20(8): 1935 - 1947. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. A. Williams, D. Mondal, and K. C. Agrawal The HIV-1 Tat Protein Enhances Megakaryocytic Commitment of K562 Cells by Facilitating CREB Transcription Factor Coactivation by CBP Experimental Biology and Medicine, December 1, 2005; 230(11): 872 - 884. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. J. Keller, D. G. Wheeler, E. Cooper, and J. L. Meier Role of the Human Cytomegalovirus Major Immediate-Early Promoter's 19-Base-Pair-Repeat Cyclic AMP-Response Element in Acutely Infected Cells J. Virol., June 15, 2003; 77(12): 6666 - 6675. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. Raychowdhury, G. Schafer, J. Fleming, S. Rosewicz, B. Wiedenmann, T. C. Wang, and M. Hocker Interaction of Early Growth Response Protein 1 (Egr-1), Specificity Protein 1 (Sp1), and Cyclic Adenosine 3'5'-Monophosphate Response Element Binding Protein (CREB) at a Proximal Response Element Is Critical for Gastrin-Dependent Activation of the Chromogranin A Promoter Mol. Endocrinol., December 1, 2002; 16(12): 2802 - 2818. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 L. Canaff and G. N. Hendy Human Calcium-sensing Receptor Gene. VITAMIN D RESPONSE ELEMENTS IN PROMOTERS P1 AND P2 CONFER TRANSCRIPTIONAL RESPONSIVENESS TO 1,25-DIHYDROXYVITAMIN D J. Biol. Chem., August 9, 2002; 277(33): 30337 - 30350. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. d'Herbomez, V. Gouze, D. Huglo, M. Nocaudie, F. Pattou, C. Proye, J.-L. Wemeau, and X. Marchandise Chromogranin A Assay and 131I-MIBG Scintigraphy for Diagnosis and Follow-Up of Pheochromocytoma J. Nucl. Med., July 1, 2001; 42(7): 993 - 997. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. K. Mahata, M. Mahata, C. V. Livsey, H.-H. Gerdes, W. B. Huttner, and D. T. O’Connor Neuroendocrine Cell Type-Specific and Inducible Expression of the Secretogranin II Gene: Crucial Role of Cyclic Adenosine Monophosphate and Serum Response Elements Endocrinology, February 1, 1999; 140(2): 739 - 749. [Abstract] [Full Text]
|
|
|
|
|
|
S. Pugazhenthi, T. Boras, D. O'Connor, M. K. Meintzer, K. A. Heidenreich, and J. E.-B. Reusch Insulin-like Growth Factor I-mediated Activation of the Transcription Factor cAMP Response Element-binding Protein in PC12 Cells. INVOLVEMENT OF p38 MITOGEN-ACTIVATED PROTEIN KINASE-MEDIATED PATHWAY J. Biol. Chem., January 29, 1999; 274(5): 2829 - 2837. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D. G. Wheeler and E. Cooper Depolarization Strongly Induces Human Cytomegalovirus Major Immediate-Early Promoter/Enhancer Activity in Neurons J. Biol. Chem., August 17, 2001; 276(34): 31978 - 31985. [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 |
) to one of slower mobility (S
