help button home button Endocrine Society Molecular Endocrinology
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
Molecular Endocrinology, doi:10.1210/me.2003-0461
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Supplemental Data File
Right arrow All Versions of this Article:
18/7/1740    most recent
Author Manuscript (PDF)
Right arrow Purchase Article
Right arrow View Shopping Cart
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Lee, L. T.-O.
Right arrow Articles by Chow, B. K.-C.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Lee, L. T.-O.
Right arrow Articles by Chow, B. K.-C.
Molecular Endocrinology 18 (7): 1740-1755
Copyright © 2004 by The Endocrine Society

Regulation of the Human Secretin Gene Is Controlled by the Combined Effects of CpG Methylation, Sp1/Sp3 Ratio, and the E-Box Element

Leo Tsz-On Lee, Kian-Cheng Tan-Un, Ronald Ting-Kai Pang, David Tai-Wai Lam and Billy Kwok-Chong Chow

Department of Zoology, University of Hong Kong, Hong Kong, People’s Republic of China

Address all correspondence and requests for reprints to: Billy Kwok-Chong Chow, Department of Zoology, University of Hong Kong, Pokfulam Road, Hong Kong, People’s Republic of China. E-mail: bkcc{at}hkusua.hku.hk.


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 RESULTS
 DISCUSSION
 MATERIALS AND METHODS
 REFERENCES
 
To unravel the mechanisms that regulate the human secretin gene expression, in this study, we have used secretin-expressing (HuTu-80 cells, human duodenal adenocarcinoma) and non-secretin-expressing [PANC-1 (human pancreatic ductile carcinoma) and HepG2 (human hepatocellular carcinoma) cells] cell models for in vitro and in vivo analyses. By transient transfection assays, within the promoter region (–11 to –341 from ATG, relative to the ATG initiation codon), we have initially identified several functional motifs including an E-box and 2 GC-boxes. Results from gel mobility shift and chromatin immunoprecipitation assays confirmed further that NeuroD, E2A, Sp1, and Sp3 bind to these E- and GC-boxes in HuTu-80 cells in vitro and in vivo, whereas only high levels of Sp3 is observed to bind the promoter in HepG2 cells. In addition, overexpression of Sp3 resulted in a dose-dependent repression of the Sp1-mediated transactivation. Collectively, these data suggest that the Sp1/Sp3 ratio is instrumental to controlling secretin gene expression in secretin-producing and non-secretin-producing cells. The functions of GC-box and Sp proteins prompted us to investigate the possible involvement of DNA methylation in regulating this gene. Consistent with this idea, we found a putative CpG island (–336 to 262 from ATG) that overlaps with the human secretin gene promoter. By methylation-specific PCR, all the CpG dinucleo-tides (26 of them) within the CpG island in HuTu-80 cells are unmethylated, whereas all these sites are methylated in PANC-1 and HepG2 cells. The expressions of secretin in PANC-1 and HepG2 cells were subsequently found to be significantly activated by a demethylation agent, 5'-Aza-2' deoxycytidine. Taken together, our data indicate that the human secretin gene is controlled by the in vivo Sp1/Sp3 ratio and the methylation status of the promoter.


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
RESULTS
 DISCUSSION
 MATERIALS AND METHODS
 REFERENCES
 
SECRETIN EXHIBITS DIVERSE physiological actions in the gastrointestinal tract. The primary action of secretin is to stimulate the release of bicarbonate, electrolytes, and water from pancreatic ductule epithelial cells. The major site for secretin expression is the endocrine S-cells location in the proximal small intestinal (1). During embryonic development, secretin is also produced in pancreatic ß-cells (2, 3). In addition to its peripheral expression, high levels of secretin immunoreactivity, comparable with those found in the duodenum, were also detected in the brain (4). By Northern blot and RT-PCR analyses, secretin mRNA transcripts were widely detected in human and rat brains (5, 6, 7). Recently, secretin and its receptor were found to express in specific neuronal populations within the cerebellar cortex. In addition, secretin was able to selectively facilitate {gamma}-aminobutryic acid (GABA)-ergic inputs onto Purkinje cells via a presynaptic and cAMP-dependent mechanism (8), implicating a direct electrophysiological action of secretin in central neurons. As a result of this work, the idea that secretin serves as a neuropeptide in the brain has received a revived attention (5).

Recently, the transcriptional regulation of the rodent secretin gene was investigated (9, 10). In this study, consistently, we found that within the human secretin gene promoter region, an E-box and two GC-rich motifs are functionally important. E-box binding proteins belong to the basic helix-loop-helix family of transcription factors and can be classified into A and B subclasses (11, 12, 13). Class A proteins such as E12 and E47 are ubiquitously expressed. They can bind onto the E-box motif either by forming a homodimer or heterodimer (with a Class B protein). Class B proteins including NeuroD, MyoD, and Myogenin are cell-specific transcription factors. These proteins can only form heterodimers with Class A proteins.

The Sp family of protein factors contains eight members, designated Sp1 to Sp8. These proteins contain conserved zinc finger DNA binding domains close to the C termini and glutamine-rich domains adjacent to the serine/threonine stretches in the N-terminal regions. Among them, Sp1, Sp3, and Sp4 are closely related. They bind GC motifs and are required for the expression of housekeeping, tissue-specific, and viral genes (14, 15). The Sp protein-regulated promoters are often associated with GC-rich regions of the genome known as CpG islands. CpG island has an average length of 1 kb, and it overlaps with the promoters of 60% of the RNA polymerase II genes. The CpG islands are transcriptionally active as the CpG dinucleotides within these regions are usually unmethylated. In vertebrates, the somatic genomes are generally methylated. CpG methylation represents a mechanism of transcriptional regulation, and it has an inhibitory influence on transcription by assembling the chromatin structure enriched in deacetylated histones (16, 17).

In this report, by gel mobility shift and chromatin immunoprecipitation (ChIP) assays, we have identified the in vitro and in vivo binding of Sp1, Sp3, NeuroD, and E2A to the GC-box and E-box motifs of the human secretin promoter region. Using Drosophila SL2 and HuTu-80 cells, the promoter was found to be dependent on NeuroD and the in vivo ratio of Sp1 and Sp3. Cotransfection studies suggest that NeuroD and Sp1 are additive in action, and this effect can be repressed by Sp3. In addition, the significance of methylation and its relationships between Sp-proteins and NeuroD were further revealed by treating secretin and non-secretin-expressing cells with a demethylating agent, 5'-Aza-2' deoxycytidine (5'-Aza-dC).


   RESULTS
TOP
 ABSTRACT
 INTRODUCTION
 RESULTS
DISCUSSION
 MATERIALS AND METHODS
 REFERENCES
 
Functional Characterization of the Human Secretin Gene 5'-Flanking Region
The transcription initiation sites are located at nucleotides (nt) –60 and –61 (relate to ATG) as determined by a primer extension analysis using poly A+ RNA prepared from HuTu-80 cells (see supplemental data S1, published on The Endocrine Society’s Journals Online web site at http://mend.endojournals.org). The human secretin gene 5'-flanking region (3 kb) was analyzed by the CpGplot program (http://www.ebi.ac.uk/emboss/cpgplot/). Two CpG islands that overlap with the immediate 5'-flanking region and the first exon of the gene were identified (Fig. 1AGo). Within these CpG islands, several putative consensus sequences including a CCAAT-box, a TATA-box, and four GC-boxes (Fig. 1BGo) were found. To initially locate the functional regions responsible for regulating secretin gene expression in various cells [HepG2 (human hepatocellular carcinoma), PANC-1 (human pancreatic ductile carcinoma), and HuTu-80], several 5'-deletion constructs with luciferase as the reporter gene were used for transient transfection studies (Fig. 2AGo). In HuTu-80 cells, p341, which contains 331 bp of 5'-flanking region from the translation start site has the highest luciferase activity, there is a 52.4 ± 5.9-fold increase in luciferase activity when compared with the control (pGL2-Basic). Promoter activities of the constructs with the tandem repeat region (TRR) (p1000 and p5000; 8.2 ± 1.4 and 18.5 ± 0.9-fold, respectively) are significantly lower than that of p341. The importance of this observation is unclear, although our data suggest that the TRR may have a putative silencing effect. 5'-Deletion from p341 to p137 led to a drastic reduction (14.4 ± 2.5-fold) in promoter function, indicating that the region between –137 and –341 contains essential element(s) for gene expression. In agreement with the findings that, in the periphery, secretin is mainly expressed in the duodenum (6), this DNA region has the highest activity in the duodenal cell line, HuTu-80 (Fig. 2AGo). Consistently, our results from quantitative real-time PCR (Fig. 2BGo) also showed that secretin is expressed in much higher levels in the HuTu-80 cells; the secretin mRNA transcript levels in HuTu-80 are 37-fold and 8.8-fold more then those in HepG2 and PANC-1 cells, respectively. The relative expression levels of secretin peptide in these cells are confirmed by immunofluorescence staining (Fig. 3Go). In conclusion, secretin is highly expressed at both peptide and mRNA levels in the human HuTu-80 cell line, and for this reason, it is used subsequently as a cell model to investigate the transcriptional regulation of the human secretin gene expression.



View larger version (31K):
[in this window]
[in a new window]
  		Fig. 1. Schematic Representation of the Human Secretin Gene and Nucleotide Sequence of Human Secretin Promoter

A, Schematic representation of the human secretin gene. The open, shaded, and solid boxes are the TRR, the exons and the predicted CpG islands as indicated in the diagram. The GC-boxes and the E-box are highlighted by gray and black lines, respectively. The deduced transcription start sites were indicated by arrows. B, Nucleotide sequences of the human secretin gene 5' flanking region. Putative binding sites for several transcription factors are underlined, including the TATA-, CCAAT-, GC-, and E-box. The open boxes (M1 to M4) indicate the positions of the site-directed mutations.

 


View larger version (19K):
[in this window]
[in a new window]
  		Fig. 2. 5' Deletion Analysis of Human Secretin Promoter and Expression Levels of Human Secretin mRNA in HuTu-80, PANC-1, and HepG2 Cells

A, 5'-Deletion analysis of the human secretin 5'-flanking region in various cell lines. Progressive 5'-deletion constructs were transiently transfected into HepG2, PANC-1, and HuTu80 cell lines. The relative promoter activity of each construct is shown as the fold of induction over the promoterless control pGL2-Basic. B, Real-time quantitative PCR analysis to compare the human secretin transcript levels in various cell types. Five micrograms total RNA prepared from HuTu-80, PANC-1, and HepG2 were reverse-transcribed to first-strand cDNAs. One twentieth of the first-strand products were used in the quantitative PCR by an Assay on Demand System (human secretin, Applied Biosystems) with Taqman Universal Master Mix. The amount of secretin mRNA was calculated based on a standard curve generated by 100 pg to 10 fg of human secretin cDNA fragment. The transcript level was normalized by the total RNA input. The correlation coefficient of this quantitative PCR is 1.00.

 


View larger version (27K):
[in this window]
[in a new window]
  		Fig. 3. Immunofluorescence Staining of Secretin in Various Cell Types

Methanol fixed cells were visualized by incubating with the rabbit antihuman secretin antibody (Phoenix Pharmaceuticals, 1:1000 dilution) and the antirabbit Alexa antibody (Molecular Probes, 1:1000 dilution). All the pictures are captured by a confocal microscope (Carl Zeiss, Jena, Germany; LSM 510) under the same exposure condition.

 
Among the four putative Sp binding motifs (GC-box1 to 4, located at –132, –165, –293 and –300, Fig. 1BGo), only the mutations of GC-box1 and 2 results in significant decreases of promoter strength (Fig. 4Go; M1, 7.5 ± 1.3-fold; M2, 19.8 ± 3.2-fold). It seems that GC-box1 plays a more important role as mutation of this motif results in the loss of most of the promoter activity (83% reduction). Similar to the rodent secretin gene promoter (10), the E-box is also important in the human promoter as mutation of this motif also leads to a drastic loss of promoter strength (78% reduction).



View larger version (12K):
[in this window]
[in a new window]
  		Fig. 4. Mutation Analysis of the Human Secretin Promoter in HuTu-80 Cells

The boxes indicate the locations of the GC-boxes (open boxes), E-box (shaded boxes) and the mutated GC/E-box sites (crossed boxes). The relative promoter activity of each construct was shown as the fold of induction over the promoterless control pGL2-Basic. *, P < 0.001 vs. p341.

 
Identification of Sp1, Sp3, NeuroD, and E2A Factors Binding to the Human Secretin Gene Promoter
To identify the transcription factors that bind the E-box and GC-box1 motifs within the secretin gene promoter, competitive and supershift gel retardation assays were performed. For the E-box motif, using the HuTu-80 nuclear extract and the cold E-box and nonspecific oligo (NonS) as competitors, three specific DNA-protein complexes (complexes 1–3; Fig. 5Go) were identified (Fig. 5AGo, lanes 3–6). In addition, only the oligo with the core E-box element (EBmutB), but not the oligo with the mutated E-box (EBmutA), can abrogate complex formation (see supplemental data S2). Our data, thus, clearly show the specificity of protein-DNA interactions. Both NeuroD and E2A (against E12 and E47 proteins) antibodies were able to decrease the intensity of complex 1 in the antibody supershift assays (Fig. 5AGo, lanes 7 and 8). Our results are consistent with the findings in rat (9) that NeuroD and E2A form a heterodimer to interact with the E-box to regulate secretin gene expression. The Sp4 antibody was used as a control, and it has no observable effect on complex formation.



View larger version (82K):
[in this window]
[in a new window]
  		Fig. 5. Identification of Sp1, Sp3, NeuroD, and E2A Factors Binding to the Human Secretin Gene Promoter

A, Gel mobility shift assays of the E-box in the human secretin gene promoter. Nuclear extract (5 µg) from HuTu-80 cells was preincubated with the competitor (cold DNA oligo EB or nonspecific oligo NonS; lanes 3–6) or an antibody (1 µg; lanes 7–9) specific for NeuroD, E2A or Sp4. Arrows indicate specific DNA-protein complexes using the E-box as a probe. B, Competitive gel mobility shift assays of the GC-box1 motif in the human secretin gene promoter. Nuclear extract (5 µg) from HuTu-80 cells was preincubated with 25x or 100x cold competitor, (oligoGC1, lanes 3 and 4; nonspecific oligo NonS, lanes 5 and 6; Sp consensus oligo, lanes 7 and 8; GCmutA, lanes 9 and 10; and GCmutB, lanes 11 and 12). Arrows indicate specific protein-DNA complexes. C, Supershift assays of the GC-box1 motif in the human secretin gene promoter. Nuclear extract (5 µg) from HuTu-80 cells was preincubated with an antibody (2 µg; lanes 3–5) specific for Sp1, Sp3, or Sp4. Arrows indicate protein-DNA complexes containing Sp proteins: Sp1 (Complex I), Sp3 (Complex II and III), and supershifted complexes (Sp1 and Sp3 supershift). D, ChIP assay. Genomic DNA and proteins from HuTu-80 (upper panel) and HepG2 (lower panel) cells were cross-linked by formaldehyde and immunoprecipitated with a Sp1, Sp3, NeuroD, or E2A antibody (lanes 4–7, respectively). After immunoprecipitation, the human secretin gene promoter (–284 to –57) was amplified by PCR using NS and NAS primers. PCR products were analyzed by an agarose gel. Immunoprecipitation without an antibody (No) and using a nonspecific antibody against rabbit Ig were carried out as negative controls.

 
Using GC-box1 as a probe (GC1), three DNA-protein complexes (I–III) were observed (Fig. 5BGo). The interactions between the proteins and the probe are specific because they can be competed by the unlabeled GC-box1 and Sp consensus oligos but not the nonspecific oligo (Fig. 5BGo, lanes 3–8). In addition, mutation of the core sequences of GC-box1 in GCmutA makes it ineffective as a competitor (Fig. 5BGo, lanes 9 and 10), whereas mutation at the 5'-region outside of the core GC-box1 in GCmutB reduces its efficiency in competing with the GC1 oligo (Fig. 5BGo, lanes 11 and 12). These studies clearly indicate the specificity of the proteins interacting with the GC-box1 sequences.

In the supershift assay, the addition of Sp1 antibody was able to significantly reduce the intensity of complex I, whereas a weak supershifted complex was observed (Fig. 5CGo, lane 3). When Sp3 antibody was used, the formation of both complexes II and III were abrogated and a supershifted complex was observed (Fig. 5CGo, lane 4). These data show that Sp1 and Sp3 are responsible for forming complex I, and complexes II and III, respectively. As a control, the Sp4 antibody fails to supershift these complexes (Fig. 5CGo, lane 5), suggesting that the Sp4 factor does not bind to the GC-box1.

To show the in vivo binding of Sp1, Sp3, NeuroD, and E2A, we have also performed the ChIP experiment (Fig. 5DGo) with the promoter region of the human secretin gene in HuTu-80 and HepG2 cells. In this study, we observed no PCR signal from the negative controls including: 1) no antibody (lane 3), 2) no template (lane 9), and 3) use of antirabbit IgG (lane 8). The controls show neither nonspecific precipitations nor PCR contamination in this assay. Positive PCR signals are found in HuTu-80 cells, supporting the idea that these proteins interact with the secretin gene promoter in in vivo situations. In HepG2 cells, the signal for Sp3 is comparable with that in HuTu-80, whereas much weaker signals are present in Sp1 and no band is detected in NeuroD and E2A. Our data indicate that Sp1, NeuroD, and E2A appear not to bind the secretin promoter in HepG2 cells and this could be a mechanism to control secretin gene expression in this cell.

Mithramycin Inhibits the Human Secretin Promoter Activity
Mithramycin selectively inhibits gene expression by modifying GC-rich regions of the DNA and inhibiting the binding of Sp proteins (18). To investigate the in vivo roles of the GC-box1 within the p341 construct, the transiently transfected HuTu-80 cells were treated with 100, 200, and 300 nM of mithramycin, and the effects of drug treatment were monitored by luciferase assays. As shown in Fig. 6AGo, mithramycin treatment caused a significant and dose-dependent decrease in the promoter activity. At the highest dose of mithramycin (300 nM), there was a specific reduction (70%) in promoter function, whereas no change in luciferase activity was observed in the non-GC-rich cytomegalovirus (CMV) promoter (Fig. 6AGo). The inhibitory effect of mithramycin on the binding of Sp proteins (Sp1 and Sp3) with the GC-box1 was further supported by gel mobility shift assays. There were drastic reduction in the intensities of complexes I–III when the nuclear extracts were incubated with 100–300 nM mithramycin (Fig. 6BGo, lanes 1–4). As a control, a non-GC-rich Oct-1 oligo was used in the binding reactions, and the same concentrations of mithramycin were unable to change the intensities of the complexes (Fig. 6BGo, lanes 5–8). In summary, our data strongly support the interactions of Sp1 and Sp3 with the GC-box1 motif in the promoter of the human secretin gene.



View larger version (47K):
[in this window]
[in a new window]
  		Fig. 6. Effects of Mithramycin to the Human Secretin Gene Promoter

A, Left panel, Mithramycin inhibits the human secretin gene promoter activity. In the transient transfection assays, mithramycin (0–300 nM) was added to the p341-transfected HuTu-80 cells. The relative promoter activity is shown as the percentage activity of the sample with 0 nM mithramycin (*, P < 0.001; **, P < 0.05). Right panel, A control vector (CMV-Luc) with a non-GC-rich promoter was transfected into HuTu-80 cells with or without mithramycin (300 nM). The results were expressed as percentage over control. There was no significant difference (n = 9) between the control and mithramycin-treated cells. B, In vitro inhibition of mithramycin to the binding of Sp1 and Sp3 with the GC-box1 of the human secretin gene promoter. Nuclear extract (5 µg) prepared from HuTu-80 cells was preincubated with 0–300 nM mithramycin on ice 1 h before adding the oligo probes GC1 or Oct-1.

 
Interactions of Sp1, Sp3, and NeuroD to Regulate the Human Secretin Gene Promoter
To investigate the in vivo function and the cooperativity of the Sp proteins in gene regulation, we employed a Sp protein-deficient cell model, the SL2 cell, for transfection studies. The Sp1 and Sp3 expression vectors, pPacSp1 and pPacSp3, were previously used and were found expressive in the transfected SL2 cells (19, 20). In this experiment, we cotransfected various amounts of pPacSp1 and/or pPacSp3 (from 0.5–3.0 µg) with the p341 promoter construct. As shown in Fig. 7AGo, overexpression of pPacSp1 led to a dose-dependent increase in the promoter activity [ranged from 2.9 ± 0.9 relative luciferase units (RLU) to 16.8 ± 2.6 RLU]. Overexpression of Sp3, on the other hand, had minimal effects on the promoter (ranged from 0.15 ± 0.04 RLU to 0.2 ± 0.02 RLU). Cotransfection of pPacSp1 (1.0 µg) with pPacSp3 (1.0 and 2.0 µg) resulted in the reduction of promoter activity (Fig. 7BGo); the promoter strength was reduced from 6.6 ± 0.7 RLU (no pPacSp3) to 2.3 ± 0.4 RLU (1.0 µg pPacSp3) and 1.5 ± 0.3 RLU (2.0 µg pPacSp3). These results reveal that Sp1 is a strong, whereas Sp3 is a weak, activator of the human secretin gene promoters, but in the presence of both Sp1 and Sp3, Sp3 functions to down-regulate the Sp1-mediated transactivation.



View larger version (17K):
[in this window]
[in a new window]
  		Fig. 7. Effects of Overexpressing Sp1 and Sp3 on the Human Secretin Promoter in Sp-Deficient SL2 Cells

A, Transactivation of the human secretin gene promoter by Sp1 and Sp3 in Drosophila SL2 cells. p341(1.5 µg) and various amounts of pPacSp1 or pPacSp3 (0, 0.5, 1.0, 2.0, and 3.0 µg) were cotransfected into SL2 cells. Control cells were transfected with the pPac0 vector alone. B, The Sp1 mediated transactivation was down-regulated by Sp3. 1.5 µg p341 was cotransfected with a constant amount of pPacSp1 (1 µg) and 1.0 or 2.0 µg pPacSp3. The total amount of DNA used in all the transfections was topped to 5 µg by the pBluescript KS+ DNA. The luciferase activity was normalized by the protein concentration. Data represent the mean ± SEM of three experiments performed in triplicates (*, P < 0.001 vs. pPacSp1, 1 µg).

 
After the pilot study in SL2 cells, the relationships of Sp1, Sp3, and NeuroD in regulating the secretin gene expression were further studied by overexpression of these transcription factors in the human duodenal HuTu-80 cell model. Consistently, overexpression of Sp1 or NeuroD up-regulated the secretin promoter, to 3.4 ± 0.2-fold or 2.8 ± 0.2-fold, respectively, when compared with the control p341. The stimulatory effects were found to be dose dependent with respect to the concentrations of the expression plasmid used in the transfection experiments. Also, overexpression of Sp1 or NeuroD was not able to activate the promoter when the GC-box (M1 or M2) or the E-box motif was mutated (see supplemental data S3). Again, no significant effect was observed when Sp3 was overexpressed (1.3 ± 0.2-fold, Fig. 8AGo). Coexpression of Sp1 together with NeuroD resulted in an observed 5.1 ± 0.4-fold (or 4.1-fold increase) in promoter activity, which is about the same as the combined effects from overexpression of individual Sp1 (2.4-fold increase) and NeuroD (1.8-fold increase). These data strongly suggest that the effects of Sp1 and NeuroD are additive in stimulating the human secretin gene promoter. Overexpression of Sp3 consistently and significantly suppressed the Sp1-mediated transactivation in Sp1/Sp3 and Sp1/Sp3/NeuroD cotransfection studies. Western blot analysis was performed to confirm the overexpression of Sp1, Sp3, and NeuroD proteins in the transfected HuTu-80 cells (Fig. 8BGo).



View larger version (18K):
[in this window]
[in a new window]
  		Fig. 8. Effects of Sp1, Sp3, and NeuroD to Human Secretin Promoter in HuTu-80 Cells

A, Coexpression study of Sp1, Sp3, and NeuroD in HuTu-80 cells. Together with p341 (1.5 µg), different combinations of Sp1/CMV, Sp3/CMV, and NeuroD/CMV were transfected into HuTu-80 cells. Data represent the mean ± SEM of three experiments performed in triplicate (*, P < 0.001 vs. transfection with 1.0 µg Sp1/CMV). B, Western blot analysis to show the overexpression of Sp1, Sp3, and NeuroD in the transfected cells. Whole cell lysate (60 µg) was electrophoresed in a 10% SDS-PAGE. The proteins were transferred to the Hybond-C extra membrane which was subsequently incubated with the polyclonal antibodies against Sp1, Sp3, or NeuroD followed by antirabbit HRP antibody for Sp1 and Sp3 or antigoat HRP antibody for NeuroD. The signal was detected by the ECL system (Amersham Biosciences). C, Sp3 down-regulates the Sp1 mediated transactivation in HuTu-80 cells. p341 (1.0 µg) was cotransfected with a constant amount of Sp1/CMV (1.0 µg) and an increasing amounts of Sp3/CMV (0, 0.5, 1.0, and 2.0 µg; *, P < 0.001 vs. Sp1/CMV, 1.0 µg). D, Intracellular level of Sp3 regulates the human secretin gene promoter activity. p341 (1.0 µg) was cotransfected with an increasing amounts of the antisense Sp3 vector, {alpha}Sp3/CMV (0, 0.25, 0.5, 1.0, and 2.0 µg). The lower panel is the Western blot analysis to show the intracellular Sp3 protein levels after the antisense Sp3 vector transfection. Western blotting was performed essentially as described earlier with the exception that 100 µg total cell lysate was used in each lane.

 
Based on the above studies including the in vivo ChIP assay, we hypothesize that the Sp1/Sp3 ratio plays a critical role to regulate the human secretin gene expression in the SL2, HepG2, and HuTu-80 cell models. To test this hypothesis in the human HuTu-80 cell, we showed 1) Sp3 can dose-dependently suppress the Sp1-mediated activation (Fig. 8CGo); and 2) overexpression of the Sp3 antisense transcript can dose-dependently increase the promoter activity (Fig. 8DGo, upper panel). The reduction in the intracellular Sp3 protein levels in the transfected cells were confirmed by Western blotting analysis (Fig. 8DGo, lower panel). In summary, our data indicate that the difference in Sp3 levels in the HepG2 and HuTu-80 cells changes the Sp1/Sp3 ratio, and this could be a key mechanism to control the expression of the human secretin gene in secretin-expressing and nonexpressing cells.

5'-Aza-dC Induction of the Human Secretin Gene
The functional involvement of GC-box and Sp proteins suggests a putative role of methylation/demethylation of the CpG dinucleotide(s) in regulating human secretin gene expression. For this reason, we investigated the effects of demethylation on the in vivo expression of the human secretin gene by using a demethylating reagent, 5'-Aza-dC. Different cells were treated with 5'-Aza-dC for 2 d before they were assayed for the expression of secretin transcripts by 1) RT-PCR coupled to Southern blot analysis (Fig. 9AGo) and 2) real-time RT-PCR (Fig. 9BGo). Treatment of PANC-1 and HepG2 cells with 5'-Aza-dC led to an augmentation of the secretin transcript levels as shown in both assays. In real-time quantitative PCR, secretin transcript levels were increased to 7.8-fold in PANC-1 cells and 9.0-fold in HepG2 cells. However, the drug had little effects (0.3-fold increase) on secretin gene expression in the secretin-expressing HuTu-80 cell. This observation could be explained by the hypothesis that the CpG dinucleotides in the human secretin promoter are already demethylated in the secretin-producing HuTu-80 cells. To test this hypothesis, we looked at the cell-specific methylation profiles of the human secretin promoter region in secretin- and non-secretin-producing cells by methylation-specific DNA modification and PCR (Fig. 9CGo). PCR primer sets specific for methylated and unmethylated DNA were designed and used to amplify the secretin gene promoter region (–285 to –75) from different cells. The amplified DNAs were sequenced to monitor the methylation status of all the CpG dinucleotides within the promoter region. We found that all the CpG dinucleotides (total 26 CpG dinucleotides) within the amplified region are unmethylated in HuTu-80 cells, whereas they are all methylated in PANC-1 and HepG2 cells. These data clearly indicate that the methylation status of the CpGs in the promoter region of the human secretin gene is cell type specific, and it plays a critical role in the control of secretin gene expression.



View larger version (32K):
[in this window]
[in a new window]
  		Fig. 9. Methylation-Specific PCR and Effects of 5'-Aza-dC on the Human Secretin Expression

A, Effects of 5'-Aza-dC on the in vivo expression of human secretin gene in different cell types. Total RNAs (5 µg) prepared from cells with (+) or without (–) 5'-Aza-dC treatment were reverse-transcribed, and the first-strand cDNAs were used for the amplification of partial human secretin (16–420 nt) and GAPDH (149–369 nt) cDNAs. The identify of the PCR products were confirmed by Southern blotting using the human secretin cDNA internal probe (111–263 nt). B, Real-time quantitative PCR to show the relative levels of human secretin transcript in different cell types with (+) or without (–) 5'-Aza-dC treatment. Five percent of the first-strand cDNAs prepared as described above were used for the quantitative PCR. The secretin mRNA/GAPDH mRNA ratio was calculated by the 2-{Delta}{Delta}Ct method using the GAPDH mRNA concentration measured by quantitative PCR as the internal control. The secretin mRNA level in HuTu-80 without 5'-Aza-dC treatment is defined as 1.0. Numbers on the bars indicate the fold changes in secretin mRNA levels after 5'-Aza-dC treatment in each cell type. C, MSP analysis of the human secretin gene promoter in various cells. Genomic DNAs prepared from HuTu-80, PANC-1, and HepG2 cells were used in the MSP analysis. Lane N shows the control amplification using primers NS, NAS, and genomic DNA before bisulfite modification as the template. Lane U represents the PCR using the unmethylated primers; US and UAS. Lane M represents the PCR using the methylated primers; MS and MAS. Lanes U and M are PCR products using genomic DNA after the bisulfite modification as the template. All the PCR products were subsequently sequenced to find out the methylation statuses of all the CpG dinucleotides (26 of them) within the amplified region.

 
To investigate the putative synergistic actions between DNA methylation and the overexpression of various transcription factors, we transfected Sp1/CMV, {alpha}Sp3/CMV, or NeuroD/CMV vector into PANC-1 cells in the presence and absence of 5'-Aza-dC. Without 5'-Aza-dC, overexpression of Sp1, antisense Sp3 or NeuroD does not activate human secretin gene expression (Fig. 10Go). On the other hand, hypomethylation of the endogenous secretin gene promoter caused by 5'-Aza-dC treatment led to significant induction in secretin gene expression from 7.0-fold (control pcDNA3.1) to 13.9-fold (Sp1/CMV), 11.6-fold ({alpha}Sp3/CMV), or 11.3-fold (NeuroD/CMV). Our data show that Sp1/Sp3 ratio and NeuroD can synergistically activate gene transcription, but a hypomethylated promoter is the key and a prerequisite for human secretin gene expression.



View larger version (17K):
[in this window]
[in a new window]
  		Fig. 10. Overexpression Studies of Sp1, Antisense Sp3, or NeuroD in PANC-1 Cells with or without 5'-Aza-dC Treatment

pcDNA3.1, Sp1/CMV, {alpha}Sp3/CMV, or NeuroD/CMV (1 µg) was transfected into PANC-1 cells by Lipofectamine. Six hours after transfection, the cells were treated with (+) or without (–) 5'-Aza-dC (0.25 µM) for 4 d. First strand cDNAs were prepared from the cells as described before and used (5%) for quantitative PCR. The secretin mRNA/GAPDH mRNA ratio was calculated by the 2-{Delta}{Delta}Ct method using the GAPDH mRNA level as the internal control. The mRNA level in the pcDNA3.1-transfected PANC-1 cells without 5'-Aza-dC treatment is defined as 1.0. Data represent the mean ± SEM of three experiments performed in duplicates. *, P < 0.001 vs. pcDNA3.1 (with 5'-Aza-dC) control.

 
Although a direct relationship between DNA methylation and secretin gene expression was observed, we cannot rule out that activation of the secretin gene expression is a result of the global change in cellular gene expression pattern, particularly the changes of Sp1, Sp3, and NeuroD levels after 5'-Aza-dC treatment. For this reason, we measured the transcript levels of these genes before and after drug treatment by real-time quantitative PCR. The Sp1 transcript levels were unchanged in all the cells tested before and after drug treatment (Fig. 11AGo). The Sp3 transcript levels dropped 36% in HuTu-80 cells and 39% in HepG2 cells (Fig. 11BGo). The NeuroD transcript levels were decreased by 61% in the HuTu-80 cells, but a 3-fold induction was observed in both PANC-1 and HepG2 cells (Fig. 11CGo). Taken together, the 5'-Aza-dC-induced secretin gene expression in PANC-1 and HepG2 cells could be achieved by demethylation of the promoter and up-regulation of NeuroD expression.



View larger version (20K):
[in this window]
[in a new window]
  		Fig. 11. Real-Time Quantitative PCR Analysis of Sp1, Sp3, and NeuroD mRNA Levels with (+) or without (–) 5'-Aza-dC Treatment

First-strand cDNAs (5%) prepared as described before were used for quantitative PCR. The Sp1, Sp3, or NeuroD mRNA/GAPDH mRNA ratio were calculated by the 2-{Delta}{Delta}Ct method using the GAPDH mRNA level measured by quantitative PCR as the internal control. The mRNA level in HuTu-80 without 5'-Aza-dC treatment is defined as 1.0. Data represent the mean ± SEM of three experiments performed in duplicates.

 

   DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 RESULTS
 DISCUSSION
MATERIALS AND METHODS
 REFERENCES
 
In the present study, we showed the in vitro and in vivo functions of Sp1 and Sp3 interacting with a GC-box present in the proximal promoter region of the human secretin gene. To investigate the differential activities and the cooperativity of Sp1 and Sp3 in gene regulation, we have used the Drosophila melanogaster Schneider (SL2) cell for our study. SL2 cells lack Sp-like activity and hence are routinely used as a model to examine the function of Sp proteins on promoter activity (19, 20). The data from SL-2 cells were subsequently confirmed by overexpressing Sp1, Sp3, and NeuroD in the human HuTu-80 cell, which expresses high levels of secretin. We concluded that Sp1 functions as a strong activator, although Sp3 alone has a very weak stimulatory effect. Sp3, however, is able to suppress the Sp1-mediated transactivation by presumably competing with Sp1 for the same binding sites (21). We have also shown that Sp3 represses Sp1-mediated transactivation by altering the Sp1/Sp3 ratio (Figs. 7BGo and 8CGo). This is further confirmed by the expression of the antisense Sp3 to reduce endogenous Sp3 levels, which eventually leads to an up-regulation of the human secretin gene. Because Sp1/Sp3 ratio varies in different cells and also at different stages of the cell cycle (22, 23), it reflects the importance of the stochastic ratio of Sp1 and Sp3 in the control of secretin gene expression. Depending on the promoter or the cell type, Sp3 can function either as an activator or a repressor. Sp3 represses the Sp1-mediated activation of kinase domain receptor (KDR/flk-1), calmodulin, and ADH5/FDH (alcohol dehydrogenase 5/formaldehyde dehydrogenase) genes through their promoters (24, 25, 26). In nicotinic acetylcholine receptor ß4-subunit, fibulin-1, and fibroblast growth factor receptor genes (27, 28, 29), Sp3 acts as a transactivator. The bifunctional property of Sp3 suggests that Sp3 could be a target for signal transduction pathways via the change of Sp1/Sp3 (14).

The functions of the GC motif and Sp proteins as well as the presence of two CpG islands in the promoter region of the human secretin gene strongly suggest an involvement of methylation/demethylation in regulating this gene. CpG islands are usually present in the 5'-flanking regions of housekeeping genes, but around 40% of CpG islands are also found in tissue- or cell-specific genes. Recently, the relationship between CpG methylation and gene silencing was evaluated by Schubeler et al. (30). Their study shows that the methylated transgene does not result in a widespread change in chromatin structure and methylation status of the targeted locus. Instead, gene repression is mediated by a mechanism that involves the localized histone hypoacetylation within the promoter. For the human secretin promoter, we found that all the CpG sites within the promoter region in PANC-1 and HepG2 are methylated, whereas all these CpG sites are unmethylated in HuTu-80 cells. These data clearly showed a relationship between methylation and gene repression. 5'-Aza-dC inhibits the enzyme DNA methyltransferase and leads to a progressive loss of CpG methylation. The fact that a demethylating agent can transform a secretin-non-producing cell into a secretin-producing strongly argues for the role of the methylation is regulating human secretin gene expression. Meanwhile, increase in the Sp1/Sp3 ratio or NeuroD level can only augment secretin gene expression when the PANC-1 cells are treated with the demethylating agent. Thus, the unmethylated status of the secretin promoter is required to initially establish and then subsequently to maintain the expression of the secretin gene. Sp1 is probably involved in both processes as it can protect CpG islands from de novo methylation (31). MeCP2 protein, which binds specifically to CpG-methylated DNA and acts as a global transcriptional repressor (32), is also under the regulation of Sp1 (33). Therefore, the Sp1/Sp3 ratio does not only directly regulate the secretin promoter activity, but also affect the CpG methylation status by protecting CpG from de novo methylation and by regulating MeCP2 expression.

Consistent with the findings in rat (9, 10), the human secretin gene promoter is strongly influenced by the interactions of NeuroD and E2A (E12/E47) with a functional E-box within the active promoter region. Given the fact that Sp proteins can interact with a number of transcription activators including c-Jun, E2F, GATA-1, and YY1 (33, 34, 35, 36, 37), the close distance between the E-box and GC-box1 suggests a possible cooperative effect between E-box and Sp binding proteins in the human secretin gene. This is possible as shown in another basic helix-loop-helix transcription factor, sterol-regulatory element binding protein-1, that can synergistically interact with Sp1 to activate low-density lipoprotein receptor gene transcription (38). By cotransfection assays (Fig. 8AGo), in this report, we have shown for the first time the additive cooperative nature and effects of NeuroD with Sp1 in the human secretin gene.

Recently, secretin has aroused considerable interests, primarily attributed to its controversial therapeutic effects on autistic patients (39, 40, 41). Nevertheless, it is now quite clear that secretin is also a neuropeptide, and a model for the neuroactive function of secretin in the cerebellum is proposed (5). In this model, secretin is synthesized in the Purkinje cells and released from the somatodendritic region upon depolarization. Secretin functions as a retrograde messenger to regulate GABA release from the basket cells via a cAMPdependent mechanism. The released secretin binds and activates the GABAA receptors on the Purkinje cell membrane leading to the influx of chloride ions. Thus, the overall neuromodulatory role of secretin in the cerebellum is to stabilize and to prevent overstimulation of the Purkinje cells. To date, there is no convincing evidence to show the genetic linkage between GABAA receptor subunit (GABRB3) or secretin with autism (42, 43). However, the genomic loci of NeuroD (2q32) and Sp3 (2q31) are located in regions that show linkage with autism in genomewide screenings (44, 45, 46). It has also been shown that about 80% of Rett cases have a de novo mutation of the transcriptional silencer MeCP2 gene (47). Rett syndrome and autism have overlapping phenotypes. This mutation leads to an abnormality in the global methylation profile of the chromatin, and MeCP2 gene expression is also under the regulation of Sp proteins (32). It is therefore possible that transcription regulators of secretin, such as, Sp1, Sp3, and NeuroD, may serve as modifiers in the pathogenesis of certain subgroups of autistic patients. Such a bold hypothesis, however, remains to be scrutinized.


   MATERIALS AND METHODS
TOP
 ABSTRACT
 INTRODUCTION
 RESULTS
 DISCUSSION
 MATERIALS AND METHODS
REFERENCES
 
Cell Culture
All cell lines were purchased from American Type Culture Collection (ATCC, Manassas, VA). HuTu-80 (HTB-40) was cultured in MEM (Invitrogen, Carlsbad, CA) with nonessential amino acids and 10% fetal bovine serum (FBS). PANC-1 (CRL-1469) and HepG2 (HB-8065) were grown in DMEM (Invitrogen) with 10% FBS. HuTu-80, PANC-1 and HepG2 were cultured at 37 C with 5% CO2 in a medium supplemented with 100 U/ml penicillin G and 100 µg/ml streptomycin (Invitrogen). Drosophila melanogaster Schneider SL2 cells (CRL-1963) were cultured at 25 C with 5% CO2 in the Schneider medium (Invitrogen) supplemented with 10% FBS, 100 U/ml penicillin G, and 100 µg/ml streptomycin at 25 C.

Fluorescence Immunocytochemical Staining
Cells were plated on glass coverslips coated with poly-D-lysine in six-well plates (Costar, San Diego, CA). Two days later, cells were fixed with cold methanol for 5 min at –20 C. The coverslips were washed in PBS three times for 5 min at room temperature. The fixed cells were incubated with rabbit antihuman secretin antibody (H-067-03; Phoenix Pharmaceuticals, Belmont, CA) in the GDB buffer [30 mM phosphate buffer (pH 7.4), 0.2% gelatin, 0.5% Triton X-100, and 0.8 M NaCl] overnight at 4 C. The cells were then washed three times in the wash buffer [20 mM phosphate buffer (pH 7.4) and 0.5 M NaCl] for 10 min at room temperature. Secondary antibody in GDB buffer was added for 1 h at room temperature, and the coverslips were washed in wash buffer for 10 min at room temperature. The secretin immunostained cells were visualized by a confocal microscope. As a control, the primary antibody was preabsorbed with an excess concentration of secretin (0.1 mM) for 24 h. Another negative control was performing by omitting the primary antibodies in the GDB buffer.

Plasmid Construction
A 6-kb DNA fragment containing the first two exons and the 5'-flanking region of the human secretin gene was isolated by screening a human genomic library (CLONTECH, Palo Alto, CA). The 3' region –11 (relative to ATG) was excluded by exonuclease III/S1 nuclease digestion (Amersham Biosciences, Buckinghamshire, UK). The resulting 5-kb fragment containing the 5' flanking region of the human secretin gene was sequenced and subcloned into the KpnI/SstI sites of pGL2-Basic (Promega, Madison, WI) to generate the promoter construct p5000. Deletion mutants p1000, p137, and p41 were constructed by digesting p5000 with XhoI/BssHII, PvuII, and ApaI, respectively, followed by either blunt-end or sticky-end self-ligation. p485 was prepared by exonuclease III/S1 nuclease digestion. p341 was prepared by inserting a 300-bp ApaI fragment into p41. Site-directed mutants of p341 were constructed by the Altered Sites II in vitro Mutagenesis System (Promega) and the mutagenic oligonucleotides (Table 1Go). Mutations were confirmed by DNA sequencing before subcloning into the pGL2-Basic vector. Expression vectors, pPac0 and pPacSp1 were gifts from Prof. R. Tjian (Howard Hughes Medical Institute, Chevy Chase, MD). pPacSp3 was kindly provided from Prof. G. Suske (Institut fuör Molekularbiologie and Tumorforschung). Normal cell expression vectors Sp1/CMV and Sp3/CMV were obtained from Prof. C. Paya (Mayo Clinic, Rochester, MN). The antisense Sp3 expression vector, {alpha}Sp3/CMV, was constructed by inserting a NotI/EcoRV cut Sp3 fragment from Sp3/CMV into pcDNA3.1(+). To generate the NeuroD expression vector, NeuroD/CMV, full-length NeuroD cDNA fragment was obtained by PCR and then subcloned into pcDNA3.1(+) vector.


View this table:
[in this window]
[in a new window]
  		Table 1. Oligonucleotides Used in Site-Directed Mutagenesis

 
Transient Transfection Assay
HuTu-80, PANC-1, and HepG2 cells were plated at a density of 2.5 x 105 cells/35 mm well (six-well plate, Costar). After 2 d of incubation, promoter-luciferase construct and ß-gal control vector were cotransfected into cells using 6 µl (2 mg/ml) Lipofectamine reagent (Invitrogen) according to the manufacturer’s protocol (Invitrogen). Cells were harvested 48 h after the transfection, and cell extracts were assayed for luciferase and ß-galactosidase activities as described previously (48). For Mithramycin treatment, mithramycin (100–300 nM) was added to the transfected cells 1 d before harvesting. For Sp proteins and NeuroD expression assays, 5 x 105 cells of SL2 cells or 2.5 x 105 of HuTu-80 cells per well were plated. After 2 d, p341 and various amounts of expression vector (pPac0, pPacSp1 or pPacSp3 for SL2 cells and Sp1/CMV, Sp3/CMV, NeuroD/CMV or {alpha}Sp3/CMV for HuTu-80) in a total of 5 µg DNA (adjusted to by pBluescript KS+) were cotransfected into the cells by Lipofectamine reagent. The transfected cells were incubated for an additional 2 d, and the luciferase activity was determined. For SL2 cells, the luciferase activities were normalized by measuring the protein concentration using a Bradford protein assay kit (Bio-Rad, Hercules, CA).

Western Blotting
Western blotting was performed essentially according to a protocol as described earlier (49). Transfected HuTu-80 cells were lysed and were resolved by SDS-PAGE and transferred onto the Hybond-C extra membrane (Amersham Biosciences). After blocking, the presence of Sp1, Sp3, and NeuroD in the cell lysate were detected by adding the primary antibody (Santa Cruz Biotechnology, Santa Cruz, CA) and subsequently the horseradish peroxidase (HRP) conjugated secondary antibody against rabbit-IgG (Sp1 and Sp3) or goat-IgG (NeuroD) and finally visualized by the ECL system (Amersham Biosciences).

Gel Mobility Shift Assays
HuTu-80 nuclear extract was prepared as described earlier (50). Double-stranded oligonucleotide probes (DNA sequences listed in Table 2Go) were end-labeled with [{gamma}-32P] by the Ready-To-Go T4-polynucleotide kinase labeling kit (Amersham Biosciences). Gel mobility shift assays were carried out at room temperature for 20 min in a 20-µl reaction mixture containing 10 mM Tris (pH 7.5), 50 mM NaCl, 2.5 mM MgCl2, 0.5 mM dithiothreitol, 4% glycerol, 2 µg poly(deoxyinosine:deoxycytosine), and 1 pmol probe. Free and bound probes were separated by electrophoresis in a 5% polyacrylamide gel. For competition assays, various concentrations of the unlabeled DNA were added with the labeled probe. In the supershift assay, antibody against Sp1 (sc-59X), Sp3 (sc-644X), Sp4 (sc-645X), NeuroD (sc-1084X), or E2A (sc-416X) (Santa Cruz Biotechnology) was included in the reaction mix. For in vitro mithramycin treatment, mithramycin (100–300 nM) was incubated with the labeled probe for 30 min at 4 C before the addition of the nuclear extract.


View this table:
[in this window]
[in a new window]
  		Table 2. Oligonucleotides Used in Gel Mobility Shift Assay

 
ChIP Assays
The ChIP assays were performed essentially according to Baek et al. (51). HuTu-80 or HepG2 cells were cross-linked with 1% formaldehyde. Cells were harvested by centrifugation and resuspended in the lysis buffer [1% sodium dodecyl sulfate (SDS), 10 mM EDTA, 50 mM Tris (pH 8.1), 1 mM phenylmethylsulfonyl fluoride, 1 µg/ml Aprotinin, and 1.5 µg/ml Pepstatin A]. After sonication in Sonifier 450 (Branson, Danbury, CT), 10 µg antibody and 20 µl Protein A/G plus agarose were added to precipitate the DNA-protein complex. Precipitated DNA-protein complex was washed in the ChIP buffer (0.1% SDS, 1% Triton X-100, 0.1% Na deoxycholate, 140 mM NaCl, 1 mM phenylmethylsulfonyl fluoride, 1 µg/ml Aprotinin, and 1.5 µg/ml Pepstatin A) and eluted in the elution buffer (1% SDS and 0.1 M NaHCO3). The mixture was incubated at 65 C for 4 h to reverse the formaldehyde cross-linking. Protein was removed by proteinase K digestion (200 µg/ml) and phenol/CHCl3 extraction. The extracted DNA was used for PCR using forward normal sense (NS) and reverse [normal antisense (NAS)] primers (Table 3Go).


View this table:
[in this window]
[in a new window]
  		Table 3. PCR Primers List

 
CpG Islands and Methylation Analysis
The presence of putative CpG islands in the proximity to the human secretin gene was analyzed by the default setting (windows size = 10, GC% over 50% and observed CpG/expected CpG ratio over 0.6, for a minimum of 200 bases) of the CpGplot program from the European Molecular Biology Open Software Suite (52).

Bisulfite conversion of genomic DNA (1 µg) was carried out by a CpG Genome DNA Modification Kit (Intergen, Purchase, NY). Twenty-four CpG sites (within –262 to –80 bp) within the human secretin gene promoter were examined by the methylation-specific PCR (MSP) and DNA sequencing. For MSP, the bisulfite converted DNA (4 µl) was amplified by the following primers (Table 3Go): 1) normal primers: NS (sense) and NAS (antisense); 2) unmethylate primers: US (sense) and UAS (antisense); and 3) methylate primers: MS (sense) and MAS (antisense). The PCR products were sequenced with the ABI Prism Big Dye Terminator Cycle Sequencing Ready Reaction Kit and the ABI Prism 3100 Genetic analyzer (Applied Biosystems, Foster City, CA).

5'-Aza-dC Treatment
HuTu-80, PANC-1, and HepG2 cells were seeded onto 150-cm2 plates. After 2 d, the cells were treated with 0.5 µM 5'-Aza-dC (Sigma, St. Louis, MO) and incubated for an additional 4 d. Total RNAs from the cells were isolated using the TriPure Isolated reagent (Roche Molecular Biochemicals, Basel, Switzerland), and 5 µg of the total RNA was reverse-transcribed by employing the oligo-deoxythymidine primer and Superscript II (Invitrogen). One tenth of the first-strand cDNA was used for the amplification of secretin and control [glyceraldehyde-3-phosphate dehydrogenase (GAPDH)] DNA fragments using primers hSCT-F and hSCT-R or GAPDH-F and GAPDH-R (primers sequence listed in Table 3Go). PCR products were analyzed by Southern blotting using the partial human secretin cDNA as a probe, and hybridization signals were detected by the STORM860 (Molecular Dynamics, Sunnyvale, CA). The endogenous Sp1, Sp3 and NeuroD transcript level changes in the treatment were further analyzed by real-time quantitative PCR. For overexpression studies, 1 µg expression vector (pcDNA3.1, Sp1/CMV, {alpha}Sp3/CMV or NeuroD/CMV) was transfected into PANC-1 cells by the Lipofectamine reagent (Invitrogen). The transfected cells were treated with or without 5'-Aza-dC (0.25 µM, 4 d). Total RNA (5 µg) was isolated and reverse-transcribed. Secretin and control GAPDH transcript levels were determined by real-time PCR.

Real-Time Quantitative PCR Analysis
The secreting levels in various cells were detected by an Assay on Demand System (Applied Biosystems, assay ID: Hs00360814_g1) with the Taqman Universal Master Mix (Applied Biosystems) using the iCycler iQ detection system (Bio-Rad). Sp1, Sp3 and NeuroD levels were monitored by SYBR Green PCR (Applied Biosystems). The specificity of the SYBR Green PCR signal was confirmed by melting curve analysis and agarose gel electrophoresis. The relative expression levels were calculated by the 2-{Delta}{Delta}Ct method (53) using GAPDH as the endogenous control.

Data Analysis
Data from the transfection assays were shown as the mean ± SEM of triplicate assays in at least three independent experiments. Data from the quantitative PCR analysis were shown as the means ± SEM of duplicate assays in at least two independent experiments. All data were analyzed by one-way ANOVA and followed by Dunnett’s test using the computer software PRISM (version 3.0, GraphPad Software Inc., San Diego, CA).


   ACKNOWLEDGMENTS
 
We thank Prof. R. Tjian (Howard Hughes Medical Institute, Chevy Chase, MD) for the pPacSp1 and pPac0 clones, Prof. G. Suske (Institut fur Molekularbiologie und Tumorforschung, Philipps-Universitat Marburg, Germany) for pPacSp3 clone, and Prof. C. Paya, Mayo Clinic (Rochester, MN), for Sp1/CMV and Sp3/CMV vector. We thank Dr. S. Philipsen (Erasmus University Rotterdam, The Netherlands) for helpful advice and suggestions.


   FOOTNOTES
 
This work was supported by a research grant from Hong Kong University SPACE, the University of Hong Kong (to K.C.T.-U.), and Research Grants Council HKU 7219/02M (to B.K.C.C.).

Abbreviations: 5'-Aza-dC, 5'-Aza-2' deoxycytidine; ChIP, chromatin immunopreciptiation; CMV, cytomegalovirus; FBS, fetal bovine serum; GABA, {gamma}-aminobutryic acid; GAPDH, glyceraldehyde-3-phosphate dehydrogenase; HepG2, human hepatocellular carcinoma; HRP, horseradish peroxidase; MSP, methylation-specific PCR; NAS, normal antisense; NonS, nonspecific oligo; NS, normal sense; nt, nucleotide; PANC-1, human pancreatic ductile carcinoma; RLU, relative luciferase units; SDS, sodium dodecyl sulfate; TRR, tandem repeat region.

Received for publication December 6, 2003. Accepted for publication April 19, 2004.


   REFERENCES
TOP
 ABSTRACT
 INTRODUCTION
 RESULTS
 DISCUSSION
 MATERIALS AND METHODS
 REFERENCES
 

  1. Chey WY, Chang T 1989 Secretin. In: Makhlouf GM, ed. Handbook of physiology. Sect 6. GI system. Vol II. Neural and endocrine biology. New York: Oxford University Press; 359–402
  2. Chey WY, Escoffery R 1976 Secretion cells in the gastrointestinal tract. Endocrinology 98:1390–1395[Abstract]
  3. Straus E, Yalow RS 1978 Immunoreactive secretin in gastrointestinal mucosa of several mammalian species. Gastroenterology 75:401–404[Medline]
  4. O’Donohue TL, Charlton CG, Miller RL, Boden G, Jacobowitz DM 1981 Identification, characterization, and distribution of secretin immunoreactivity in rat and pig brain. Proc Natl Acad Sci USA 78:5221–5224[Abstract/Free Full Text]
  5. Ng SS, Yung WH, Chow BK 2002 Secretin as a neuro-peptide. Mol Neurobiol 26:97–107[CrossRef][Medline]
  6. Whitmore TE, Holloway JL, Lofton-Day CE, Maurer MF, Chen L, Quinton TJ, Vincent JB, Scherer SW, Lok S 2000 Human secretin (SCT): gene structure, chromosome location, and distribution of mRNA. Cytogenet Cell Genet 90:47–52[CrossRef][Medline]
  7. Itoh N, Furuya T, Ozaki K, Ohta M, Kawasaki T 1991 The secretin precursor gene. Structure of the coding region and expression in the brain. J Biol Chem 266:12595–12598[Abstract/Free Full Text]
  8. Yung WH, Leung PS, Ng SS, Zhang J, Chan SC, Chow BK 2001 Secretin facilitates GABA transmission in the cerebellum. J Neurosci 21:7063–7068[Abstract/Free Full Text]
  9. Mutoh H, Fung BP, Naya FJ, Tsai MJ, Nishitani J, Leiter AB 1997 The basic helix-loop-helix transcription factor BETA2/NeuroD is expressed in mammalian enteroendocrine cells and activates secretin gene expression. Proc Natl Acad Sci USA 94:3560–3564[Abstract/Free Full Text]
  10. Nishitani J, Rindi G, Lopez MJ, Upchurch BH, Leiter AB 1995 Transcriptional regulation of secretin gene expression. J Clin Gastroenterol 21(Suppl 1):S50–S55
  11. Lassar AB, Davis RL, Wright WE, Kadesch T, Murre C, Voronova A, Baltimore D, Weintraub H 1991 Functional activity of myogenic HLH proteins requires hetero-oligomerization with E12/E47-like proteins in vivo. Cell 66:305–315[CrossRef][Medline]
  12. Murre C, McCaw PS, Vaessin H, Caudy M, Jan LY, Jan YN, Cabrera CV, Buskin JN, Hauschka SD, Lassar AB 1989 Interactions between heterologous helix-loop-helix proteins generate complexes that bind specifically to a common DNA sequence. Cell 58:537–544.[CrossRef][Medline]
  13. Murre C, Voronova A, Baltimore D 1991 B-cell- and myocyte-specific E2-box-inding factors contain E12/E47-like subunits. Mol Cell Biol 11:1156–1160[Abstract/Free Full Text]
  14. Suske G 1999 The Sp-family of transcription factors. Gene 238:291–300[CrossRef][Medline]
  15. Philipsen S, Suske G 1999 A tale of three fingers: the family of mammalian Sp/XKLF transcription factors. Nucleic Acids Res 27:2991–3000[Abstract/Free Full Text]
  16. Bird AP, Wolffe AP 1999 Methylation-induced repression—belts, braces, and chromatin. Cell 99:451–454[CrossRef][Medline]
  17. Ng HH, Bird A 1999 DNA methylation and chromatin modification. Curr Opin Genet Dev 9:158–163[CrossRef][Medline]
  18. Miller DM, Polansky DA, Thomas SD, Ray R, Campbell VW, Sanchez J, Koller CA 1987 Mithramycin selectively inhibits transcription of G-C containing DNA. Am J Med Sci 294:388–394[Medline]
  19. Courey AJ, Tjian R 1988 Analysis of Sp1 in vivo reveals multiple transcriptional domains, including a novel glutamine-rich activation motif. Cell 55:887–898[CrossRef][Medline]
  20. Majello B, De Luca P, Suske G, Lania L 1995 Differential transcriptional regulation of c-myc promoter through the same DNA binding sites targeted by Sp1-like proteins. Oncogene 10:1841–1848[Medline]
  21. Hagen G, Dennig J, Preiss A, Beato M, Suske G 1995 Functional analyses of the transcription factor Sp4 reveal properties distinct from Sp1 and Sp3. J Biol Chem 270:24989–24994[Abstract/Free Full Text]
  22. Apt D, Watts RM, Suske G, Bernard HU 1996 High Sp1/Sp3 ratios in epithelial cells during epithelial differentiation and cellular transformation correlate with the activation of the HPV-16 promoter. Virology 224:281–291[CrossRef][Medline]
  23. Discher DJ, Bishopric NH, Wu X, Peterson CA, Webster KA 1998 Hypoxia regulates beta-enolase and pyruvate kinase-M promoters by modulating Sp1/Sp3 binding to a conserved GC element. J Biol Chem 273:26087–26093[Abstract/Free Full Text]
  24. Hata Y, Duh E, Zhang K, Robinson GS, Aiello LP 1998 Transcription factors Sp1 and Sp3 alter vascular endothelial growth factor receptor expression through a novel recognition sequence. J Biol Chem 273:19294–19303[Abstract/Free Full Text]
  25. Pan X, Solomon SS, Shah RJ, Palazzolo MR, Raghow RS 2000 Members of the Sp transcription factor family regulate rat calmodulin gene expression. J Lab Clin Med 136:157–163[CrossRef][Medline]
  26. Kwon HS, Kim MS, Edenberg HJ, Hur MW 1999 Sp3 and Sp4 can repress transcription by competing with Sp1 for the core cis-elements on the human ADH5/FDH minimal promoter. J Biol Chem 274:20–28[Abstract/Free Full Text]
  27. Castoldi M, Chu ML 2002 Structural and functional characterization of the human and mouse fibulin-1 gene promoters: role of Sp1 and Sp3. Biochem J 362:41–50[CrossRef][Medline]
  28. Bigger CB, Melnikova IN, Gardner PD 1997 Sp1 and Sp3 regulate expression of the neuronal nicotinic acetyl-choline receptor ß4 subunit gene. J Biol Chem 272:25976–25982[Abstract/Free Full Text]
  29. Parakati R, DiMario JX 2002 Sp1- and Sp3-mediated transcriptional regulation of the fibroblast growth factor receptor 1 gene in chicken skeletal muscle cells. J Biol Chem 277:9278–9285[Abstract/Free Full Text]
  30. Schubeler D, Lorincz MC, Cimbora DM, Telling A, Feng YQ, Bouhassira EE, Groudine M 2000 Genomic targeting of methylated DNA: influence of methylation on transcription, replication, chromatin structure, and histone acetylation. Mol Cell Biol 20:9103–9112[Abstract/Free Full Text]
  31. Brandeis M, Frank D, Keshet I, Siegfried Z, Mendelsohn M, Nemes A, Temper V, Razin A, Cedar H 1994 Sp1 elements protect a CpG island from de novo methylation. Nature 371:435–438[CrossRef][Medline]
  32. Jones PL, Veenstra GJ, Wade PA, Vermaak D, Kass SU, Landsberger N, Strouboulis J, Wolffe AP 1998 Methylated DNA and MeCP2 recruit histone deacetylase to repress transcription. Nat Genet 19:187–191[CrossRef][Medline]
  33. Marin M, Karis A, Visser P, Grosveld F, Philipsen S 1997 Transcription factor Sp1 is essential for early embryonic development but dispensable for cell growth and differentiation. Cell 89:619–628[CrossRef][Medline]
  34. Kardassis D, Papakosta P, Pardali K, Moustakas A 1999 c-Jun transactivates the promoter of the human p21(WAF1/Cip1) gene by acting as a superactivator of the ubiquitous transcription factor Sp1. J Biol Chem 274:29572–29581[Abstract/Free Full Text]
  35. Rotheneder H, Geymayer S, Haidweger E 1999 Transcription factors of the Sp1 family: interaction with E2F and regulation of the murine thymidine kinase promoter. J Mol Biol 293:1005–1015[CrossRef][Medline]
  36. Merika M, Orkin SH 1995 Functional synergy and physical interactions of the erythroid transcription factor GATA-1 with the Kruppel family proteins Sp1 and EKLF. Mol Cell Biol 15:2437–2447[Abstract]
  37. Lee JS, Galvin KM, Shi Y 1993 Evidence for physical interaction between the zinc-finger transcription factors YY1 and Sp1. Proc Natl Acad Sci USA 90:6145–6149[Abstract/Free Full Text]
  38. Naar AM, Beaurang PA, Robinson KM, Oliner JD, Avizonis D, Scheek S, Zwicker J, Kadonaga JT, Tjian R 1998 Chromatin, TAFs, and a novel multiprotein coactivator are required for synergistic activation by Sp1 and SREBP-1a in vitro. Genes Dev 12:3020–3031[Abstract/Free Full Text]
  39. Horvath K, Stefanatos G, Sokolski KN, Wachtel R, Nabors L, Tildon JT 1998 Improved social and language skills after secretin administration in patients with autistic spectrum disorders. J Assoc Acad Minor Phys 9:9–15[Medline]
  40. Molloy CA, Manning-Court, Swayne S, Bean J, Brown JM, Murray DS, Kinsman AM, Brasington M, Ulrich CD 2002 Lack of benefit of intravenous synthetic human secretin in the treatment of autism. J Autism Dev Disord 32:545–551[CrossRef][Medline]
  41. Unis AS, Munson JA, Rogers SJ, Goldson E, Osterling J, Gabriels R, Abbott RD, Dawson G 2002 A randomized, double-blind, placebo-controlled trial of porcine versus synthetic secretin for reducing symptoms of autism. J Am Acad Child Adolesc Psychiatry 41:1315–1321[CrossRef][Medline]
  42. Martin ER, Menold MM, Wolpert CM, Bass MP, Donnelly SL, Ravan SA, Zimmerman A, Gilbert JR, Vance JM, Maddox LO, Wright HH, Abramson RK, DeLong GR, Cuccaro ML, Pericak-Vance MA 2000 Analysis of linkage disequilibrium in gamma-aminobutyric acid receptor subunit genes in autistic disorder. Am J Med Genet 96:43–48[CrossRef][Medline]
  43. Yamagata T, Aradhya S, Mori M, Inoue K, Momoi M, Nelson D 2002 The human secretin gene: fine structure in 11p15.5 and sequence variation in patients with autism. Genomics 80:185[CrossRef][Medline]
  44. A genomewide screen for autism: strong evidence for linkage to chromosomes 2q, 7q, and 16p 2001 Am J Hum Genet 69:570–581
  45. Philippe A, Martinez M, Guilloud-Bataille M, Gillberg C, Rastam M, Sponheim E, Coleman M, Zappella M, Aschauer H, Van Maldergem L, Penet C, Feingold J, Brice A, Leboyer M, van Malldergerme L 1999 Genome-wide scan for autism susceptibility genes. Paris Autism Research International Sibpair Study. Hum Mol Genet 8:805–812[Abstract/Free Full Text]
  46. Gutknecht L 2001 Full-genome scans with autistic disorder: a review. Behav Genet 31:113–123[CrossRef][Medline]
  47. Amir RE, Van dV, I, Wan M, Tran CQ, Francke U, Zoghbi HY 1999 Rett syndrome is caused by mutations in X-linked MECP2, encoding methyl-CpG-binding protein 2. Nat Genet 23:185–188[CrossRef][Medline]
  48. Ngan ES, Cheng PK, Leung PC, Chow BK 1999 Steroidogenic factor-1 interacts with a gonadotrope-specific element within the first exon of the human gonadotropin-releasing hormone receptor gene to mediate go-nadotrope-specific expression. Endocrinology 140:2452–2462[Abstract/Free Full Text]
  49. Cheng CK, Yeung CM, Chow BK, Leung PC 2002 Characterization of a new upstream GnRH receptor promoter in human ovarian granulosa-luteal cells. Mol Endocrinol 16:1552–1564[Abstract/Free Full Text]
  50. Lee KA, Bindereif A, Green MR 1988 A small-scale procedure for preparation of nuclear extracts that support efficient transcription and pre-mRNA splicing. Gene Anal Tech 5:22–31[CrossRef][Medline]
  51. Baek SJ, Horowitz JM, Eling TE 2001 Molecular cloning and characterization of human nonsteroidal anti-inflammatory drug-activated gene promoter. Basal transcription is mediated by Sp1 and Sp3. J Biol Chem 276:33384–33392[Abstract/Free Full Text]
  52. Rice P, Longden I, Bleasby A 2000 EMBOSS: the E