Tài liệu Báo cáo khoa học: Characterization of human deoxyribonuclease I gene (DNASE1) promoters reveals the utilization of two transcription-starting exons and the involvement of Sp1 in its transcriptional regulation ppt

We first mapped the transcription start sites of DNASE1 in human pancreas and in the DNase I-producing human pan-creatic cancer cell line QGP-1, and revealed a novel site 12 kb upstream o

(DNASE1) promoters reveals the utilization of two

transcription-starting exons and the involvement of Sp1

in its transcriptional regulation

Yoshihiko Kominato1, Misuzu Ueki2, Reiko Iida3, Yasuyuki Kawai4, Tamiko Nakajima1, Chikako Makita1, Masako Itoi1, Yutaka Tajima1, Koichiro Kishi1and Toshihiro Yasuda2

1 Department of Legal Medicine and Medical Genetics, Gunma University, Japan

2 Division of Medical Genetics and Biochemistry, University of Fukui, Japan

3 Division of Legal Medicine, University of Fukui, Japan

4 Third Division of Internal Medicine, University of Fukui, Japan

Deoxyribonuclease I (DNase I, EC 3.1.21.1) is an

enzyme that preferentially attacks double-stranded

DNA by Ca2+- and Mg2+⁄ Mn2+-dependent

endo-nucleolytic cleavage to produce oligonucleotides with

5¢-phosphoryl- and 3¢-hydroxy termini [1,2] DNase I

is considered to play a major role in the digestion of

dietary DNA, because in vertebrates it is secreted by

exocrine⁄ endocrine glands such as the pancreas and

parotid gland into the alimentary tract [3–5] However,

the presence of the enzyme in mammalian tissues other than the digestive organs [6–8] suggested that it might have other function(s) in vivo; endogenous DNase I has been regarded as a candidate endonuclease respon-sible for internucleosomal DNA degradation during apoptosis [9] Furthermore, Napirei et al have shown that extracellular (serum) DNase I participates in the chromatin breakdown of necrotic cells, achieved by its diffusion from the extracellular fluid into the

Keywords

alternative splicing; deoxyribonuclease I;

genes; promoter; Sp1

Correspondence

T Yasuda, Division of Medical Genetics and

Biochemistry, Faculty of Medical Sciences,

University of Fukui, Eiheiji, Fukui 910-1193,

Japan

Fax: +81 776 61 8149

Tel: +81 776 61 8287

E-mail: tyasuda@fmsrsa.fukui-med.ac.jp

Database

The nucleotide sequences reported here

have been submitted to the GenBank/

EMBL/DDBJ Data Bank with accession

numbers AB188151 and AB188152

(Received 22 March 2006, revised 8 May

2006, accepted 15 May 2006)

doi:10.1111/j.1742-4658.2006.05320.x

Levels of deoxyribonuclease I (DNase I) activity in vivo have been shown

to be altered by physiological and⁄ or pathological processes However, no information is available on the regulation of DNase I gene (DNASE1) expression in vivo or in vitro We first mapped the transcription start sites

of DNASE1 in human pancreas and in the DNase I-producing human pan-creatic cancer cell line QGP-1, and revealed a novel site 12 kb upstream

of exon 1, which was previously believed to be the single transcription-starting exon This initiation site marks an alternative transcription-starting exon, designated 1a Exons 1 and 1a were used simultaneously as transcription-starting exons in pancreas and QGP-1 cells Promoter assay, EMSA and chromatin immunoprecipitation analysis with QGP-1 cells showed the pro-moter region of exon 1a in which the Sp1 transcription factor is specifically involved in promoter activity This is the first to be identified as a tran-scription factor responsible for gene expression of vertebrate DNase I genes Furthermore, RT-PCR analysis indicated alternative splicing of human DNASE1 pre-mRNA in pancreas and QGP-1 cells Only two tran-scripts among eight alternative splicing products identified can be transla-ted to produce intact DNase I protein These results suggest that human DNASE1 expression is regulated through the use of alternative promoter and alternative splicing

Abbreviations

AMI, acute myocardial infarction; ChIP, chromatin immunoprecipitation; DNase I, deoxyribonuclease I; DNASE1, DNase I gene; EMSA, electrophoretic mobility shift assay; PCI, percutaneous coronary intervention; SRED, single radial enzyme diffusion.

cytoplasm and nucleus of such cells [10] In a similar

context, DNase I has been postulated to be responsible

for the removal of DNA from nuclear antigens at sites

of high cell turnover and necrosis, and thus for the

prevention of systemic lupus erythematosus [11,12]

Recently, we demonstrated that an abrupt elevation of

serum DNase I activity occurs within  3 h of the

onset of symptoms in patients with acute myocardial

infarction (AMI) and that DNase I activity in serum

then exhibits a marked time-dependent decline within

12 h, returning to basal levels within 24 h [13]

More-over, percutaneous coronary intervention (PCI), which

is performed to treat patients with stable angina

pec-toris, offers an in vivo model of mild myocardial

ische-mia in humans Irrespective of a lack of alteration in

levels of other conventional cardiac markers such as

creatine kinase isoenzyme MB and cardiac troponin T,

serum DNase I levels rose significantly from basal

lev-els by 3 h after completion of the PCI procedure,

returning to basal levels by 12–24 h, in a manner

sim-ilar to in AMI [14] These findings permitted us to

sug-gest that myocardial ischemia rather than injury

induces such elevation in serum DNase I activity

However, the mechanisms for the elevation of serum

DNase I activity induced by ischemia during AMI or

PCI remain to be elucidated Delineation of the

molecular basis for our observations is essential to

evaluate the elevated DNase I activity in the sera of

patients with AMI and to validate the use of serum

DNase I activity as a new diagnostic marker for the

early detection of AMI To elucidate the molecular

basis of this phenomenon, it is important to

under-stand the regulatory mechanism of the human DNase I

gene (DNASE1) expression

Previous molecular–genetic studies have shown that

the human DNase I gene consists of at least nine

exons spanning > 3.2 kb of genomic DNA at

chromo-some 16p13.3; exon 2 includes the translation initiation

codon (ATG) and the first exon is believed to include

only the 5¢-UTR of the mRNA [15,16] However,

com-parison of the sequence of the 5¢-UTR of human

pan-creatic DNASE1 mRNA reported previously [15]

with the human genomic sequence showed that the

5¢-terminal 15 nucleotides of the 5¢-UTR are not found

in the human genomic sequence [16], whereas the 3¢

segment of 143 nucleotides in the 5¢-UTR matches the

genomic sequence Thus, the transcription initiation

site of exon 1 has not yet been definitively identified

To our knowledge, no information is available on the

regulation of vertebrate DNASE1 expression in vivo or

in vitro, including characterization of the promoter

region of the gene and the associated transcriptional

factors Therefore, delineation of the transcriptional

regulation of human DNASE1 may provide clues to the mechanisms underlying ischemia-induced elevation

of DNase I activity in vivo

Here, we describe the identification of a novel tran-scription starting exon, designated as exon 1a, in human DNASE1 and characterization of the promoter regions of the gene; Sp1 transcription factor plays an important role in promoter activity in the 5¢-upstream region of DNASE1 exon 1a

Results

Mapping of transcription initiation sites in the human DNASE1 gene

To identify the transcription initiation sites of DNASE1 in pancreas, 5¢-RACE based on RNA ligase-mediated and oligo-capping RACE [17] was performed with cDNA synthesized from pancreas Agarose gel electrophoresis of the 5¢-RACE products showed a slow-migrating major band and several faster-migra-ting faint bands Alternative splicing occurs in many genes, and so DNA fragments were purified from the major band and cloned into a sequencing vector DNA sequences were determined for seven transformant clones Four clones contained a 476-bp 5¢-RACE DNA product, of which the 3¢ 380 bp were identical to exons 1–3 from the 5¢-end of the reverse primer DN+231 to the 5¢-end of the region where the sequence of exon 1 reported previously [15] matched with the sequence deposited in GenBank (Accession

no AC006111) (Fig 1A) The sequence of the 5¢ por-tion beyond that point was identical to a 96-bp region

of the genomic DNA, indicating that the entire sequence of exon 1 comprises 243 bp

Human DNase I activity is mainly demonstrable in pancreas, alimentary tract and pituitary [3,18], and most serum DNase I appears to be produced from those tissues Moreover, our survey of DNase I-produ-cing cell lines showed a high level of DNase I activity and its transcripts in QGP-1 cells, which were estab-lished by Kaku et al [19] from a human pancreatic islet cell carcinoma (possibly D cells) as a carcinoem-bryonic antigen-secreting cell line A similar 5¢-RACE was performed with cDNA synthesized from total RNA of QGP-1 cells The DNA sequences of the 5¢-RACE products were determined for five transform-ant clones Four clones contained a 417-bp 5¢-RACE DNA product that appeared to be a hybrid between exon 1 and the upstream genomic DNA of DNASE1: the sequence of the 3¢ portion in those products was identical to that of exons 1–3 from the 5¢-end of the reverse primer DN+231 to a position +156 relative

B

Fig 1 The nucleotide sequence of the 5¢-flanking region in the human DNase I gene (A) The sequence located between positions )100 and +250 relative to the 5¢-end in DNASE1 exon 1 +1 above the sequence indicates the 5¢-end of exon 1 Open circles indicate locations of 5¢-ends

of the DNASE1 transcripts, determined by 5¢-RACE using cDNA obtained from pan-creas The star indicates the 5¢-end of the DNASE1 mRNA reported previously [3] Exon 1 (nucleotides +1 to +243) is indica-ted by uppercase letters within rectangles The vertical line between positions +155 and +156 represents the splicing junction where exon 1a is ligated to the part of exon 1 between positions +156 and +243 (B) The sequence located between posi-tions )200 and +150 relative to the tran-scription start site in human DNASE1 exon 1a Upstream counting is done from +1 of exon 1 )11871 above the sequence indicates the 5¢-end of exon 1a The num-bers in parentheses demonstrate the posi-tion of the corresponding nucleotides relative to the transcription start site in exon 1a Open circles indicate locations of 5¢-ends of the DNASE1 transcripts, deter-mined by 5¢-RACE using cDNA obtained from QGP-1 cells Exon 1a (nucleotides )11871 to )11770) is indicated by upper-case letters within rectangles Several puta-tive transcription factor binding sites were found using TRANSFAC software and are indi-cated by overbars The underlines repre-sent the locations of the two

oligonucleotide probes used for further ana-lysis The position and identity of mutations

at )11944 to )11941 are indicated in oligo-nucleotide m91,50 and reporter construct )197HmSp.

to the beginning of exon 1, as shown in Fig 1A.

Beyond that point, however, the sequence of the 5¢

portion showed 100% identity with that of genomic

DNA  12 kb upstream of the DNASE1 exon 1

(Fig 1B) More interestingly, the products lacked the

sequence between positions +1 and +155 in exon 1

This comparison with the upstream genomic sequence

of DNASE1 allowed us to demonstrate the presence of

an alternative exon, which we named exon 1a The

donor splice site between exon 1a and the subsequent

intron had GT, whereas the acceptor site between the

subsequent intron and the 5¢-end at position +156 in

exon 1 had AG Therefore, the splice sites seem to be

compatible with a splicing junction

Confirmation of utilization of exon 1a as

a transcription initiation exon and occurrence

of alternative splicing

To examine whether exon 1a is used as a transcription

starting exon in QGP-1 cells and pancreas, and to

con-firm the splicing junction between exon 1 and the upstream DNA, RT-PCR was carried out using a pri-mer specific for exon 1a and a reverse pripri-mer comple-mentary to the sequence in exon 8 DNA fragments of different sizes were amplified from the RNA of QGP-1 cells and pancreas (Fig 2A) Determination of the nucleotide sequences of the RT-PCR products revealed

Exons DN-110

DN-144

DN+89

895 bp

877 bp

929 bp

789 bp

764 bp

A

C

H

I

J

841 bp

D

755 bp

F

806 bp

E

730 bp

G

881 bp

B

DN+385 DN+721 DN-105

DN-9 +1

A

B

Fig 2 RT-PCR analysis to detect the transcription-starting exon in

DNASE1 of QGP-1 cells and pancreas (A) RT-PCR analysis Total

RNA prepared from QGP-1 cells or pancreas was

reverse-tran-scribed with random primer, and the resulting single-strand cDNA

was used as a template for PCR analysis The DNASE1

amplifica-tion was performed using either distinct starting exon-specific

pri-mer DN )110 (left) or DN)144 (right) and a common reverse primer

DN+785 complementary to exon 8 of the DNASE1 PCR products

were electrophoresed through a 1.5% agarose gel and stained with

ethidium bromide The amplified fragments were named A–J A

1 kb Plus DNA Ladder was used as a molecular size marker (B)

Splicing patterns of the amplified fragments A–J Nucleotide

sequences of these fragments were determined and then

com-pared Schematically represented DNASE1 was aligned with the

RT-PCR products amplified, using a set of each starting

exon-speci-fic primer (DN )110 or DN)144) and the DN+785 primer, which are

represented by arrows Open boxes represent the DNASE1 exons,

and a vertical broken line indicates the splicing junction between

exon 1a and a portion of exon 1 The thick straight lines represent

the intron sequence +1 indicates the position of the transcription

start site of exon 1 Dashed v-shaped lines in RT-PCR amplified

fragments A–J indicate regions that are removed by splicing,

whereas a dashed line in exon 7 of fragment C represents a

dele-tion of 39 bp The thick lines indicate intron 6 of 75 bp in fragments

B and C, 3¢ portion of 35 bp in intron 5 in fragments C and D, and

3¢ portion of 25 bp in intron 6 in fragments F and I The number at

the right of each RT-PCR product represents the length of the

prod-uct The scheme also shows the location of the reverse primer

DN+89 used in another starting exon-specific PCR and quantitative

real-time RT-PCR, the locations of the primers DN )105 and DN)9

used in ChIP assays, the locations of the primers DN+232 and

DN+254 utilized in 5¢-RACE, and the locations of the primers

DN+385 and DN+721 used in 3¢-RACE.

that the sequence between positions +156 and +243

in exon 1 was linked with that of exon 1a  12 kb

upstream of exon 1 in both QGP-1 cells and pancreas

(Fig 2B) In addition, another RT-PCR with a primer

corresponding to the 5¢-terminus of exon 1 and the

same reverse primer, followed by sequence

determin-ation, showed that DNA fragments derived from

exon 1, including the DNASE1 full-length and splicing

variants, were amplified in QGP-1 cells and pancreas

Therefore, these results allow us to conclude that both

exon 1a and exon 1 are used simultaneously as

tran-scription-starting exons in QGP-1 cells and pancreas

The RT-PCR products of different sizes observed in

QGP-1 cells and pancreas seem to arise from

alternat-ive splicing The complex patterns of alternatalternat-ively

spliced products are represented schematically in

Fig 2B Although the splicing patterns of the

DNASE1 transcripts were complex, the transcripts

could be classified into two types: first, the full-length

transcripts A and H and second, transcripts B–G, I

and J that all lack exon 3 Only two DNASE1

tran-scripts, those corresponding to the amplified products

A and H, can be translated to produce intact DNase I

protein By contrast, 3¢-RACE using total RNA from

QGP-1 cells and pancreas as a template gave a single

band with a sequence identical to that reported

previ-ously [16], in addition to the observation of a specific cleavage⁄ polyadenylation site located in the 3¢-flanking region of the gene at position 142 downstream of the stop codon Thus, DNASE1 splicing variants appear

to share a common 3¢-UTR

Quantitative real-time RT-PCR was performed using each distinct starting exon-specific primer and

a common reverse primer complementary to exon 2

to determine the relative abundance by comparing the copy number of transcripts containing exon 1a with the number starting from exon 1 The abun-dance of transcripts starting from exon 1 was 10-fold higher than that of the transcripts starting from exon 1a in pancreas, whereas it was half in QGP-1 cells

To examine whether transcription starting from both exon 1a and exon 1 results in the production of DNase I enzyme, we transfected the expression plasmids ex-pDN1a and ex-pDN1, containing the sequences cor-responding to the DNASE1 cDNAs starting from either exon 1a or exon 1, respectively, into COS-7 cells and then determined the levels of DNase I activity secreted into the medium of the cells transfected with each plas-mid DNase I activity could be seen in cells transfected with either expression vector, although they differed

in levels of expressed DNase I activity (Fig 3) The

ex-pDN1

Relative DNase I activities

in the supernatant of COS-7 cells 0

ex-pDN

+1

ex-pDN1a

ATG

+847 TGA

coding region

coding region

(+1) (+156)

(-11868) (-11770)

−5

+1 +847

exon 1

+1 +847

exon 1a

coding region

(+243)

Fig 3 Demonstration of the DNase I activities expressed by the DNASE1 expression constructs containing distinct 5¢-UTR regions of DNASE1 mRNA in COS-7 cells DNase I expression vectors, ex-pDN1 and ex-pDN1a, containing the entire 5¢-UTR region of each transcripts derived from exons 1 and 1a, respectively, were constructed and transfected into COS-7 cells Constructs are shown in the left-hand panel The numbers over the diagrams indicate the position of the corresponding nucleotides relative to the translation start site, and the numbers

in parentheses below the diagram show the position of the corresponding nucleotides relative to the transcription start site in exon 1 In the expression vector ex-pDN, the Kozak sequence just upstream from the coding region of DNASE1 is contained and indicated by a closed box In the expression vector ex-pDN1, the gray box represents the whole sequence of exon 1 In the expression vector ex-pDN1a, the sequence between +4 and +102 relative to the transcription start site of exon 1a, indicated by the open box, is ligated with the part of exon 1 between positions +156 and +243 The resulting DNase I activity in the medium secreted from each transfected cells was normal-ized by coexpressed b-galactosidase activity, and is shown in the right panel The mean values and standard deviations were calculated from five independent experiments The activity of the expression plasmid ex-pDN was assigned an arbitrary value of 1.0 The DNase I activities

of the cells transfected with ex-pDN1 were statistically significantly lower than those of cells transfected with ex-pDN or ex-pDN1a (P < 0.05).

findings indicate that transcripts starting from either

exon 1a or exon 1 are translated to produce intact

DNase I protein

Characterization of the promoter region of

exons 1a and 1 in the human DNase I gene

Because 5¢-RACE analysis identified two

transcription-starting exons used in DNASE1, we characterized the

promoters that regulate transcription of the DNASE1

messages containing exons 1 or 1a To examine

pro-moter activity in the 5¢-flanking region of exon 1 in

DNASE1, we first obtained the )1386M construct by

introducing the )1386 to +268 sequence of DNASE1

into the promoterless pGL3–basic vector upstream of

the luciferase coding sequence The reporter plasmid

was transfected into QGP-1 cells, followed by assay of

luciferase activities (Fig 4A) pGL3–promoter vector

containing the SV40 promoter and pGL3–basic vector

without the promoter sequence were used as positive

and negative controls, respectively The relative

lucif-erase activity of the )1386M construct was at least

eightfold higher than that of pGL3–basic vector and

was half that of pGL3–promoter vector These findings

demonstrate the promoter activity of the 5¢-flanking

region of exon 1 in DNASE1 Deletion of the

upstream end of the 5¢-flanking region of exon 1 from

position)1386 to )231, )116, or )78 did not result in

any significant change However, deletion of the

sequence from position )78 to )54 resulted in the loss

of 50% of the luciferase activity These results imply

that the )78 to )55 region is required for DNASE1

proximal promoter activity in the 5¢-flanking region of

exon 1

Similarly, we obtained the )2081H construct by

introducing the )13952 to )11781 sequence of

DNASE1exon 1a relative to the transcription start site

of exon 1 into the promoterless pGL3–basic vector

upstream of the luciferase coding sequence, followed

by transient transfection into QGP-1 cells The relative

luciferase activity of the)2081H construct was at least

14-fold higher than that of the pGL3–basic vector and

was not inferior to that of the pGL3–promoter vector

(Fig 4B), indicating promoter activity of the

5¢-flank-ing region of exon 1a in DNASE1 These results are

consistent with the finding that DNASE1 utilizes two

transcription starting exons in QGP-1 cells Deletion

of the upstream end of the 5¢-flanking region of

exon 1a from position)13952 to )11965 did not result

in any significant change However, deletion of the

sequence from position )11965 to )11944 elicited an

approximate twofold increase in luciferase activity,

indicating that negative regulatory elements are present

in the)11965 to )11944 region Furthermore, deletion

of the upstream end from position )11944 to )11931 resulted in a fivefold decrease in luciferase activity, suggesting that elements important for distal promoter function are contained within the deleted region Inspection of the sequence between )73 and )60 upstream of the transcription start site of exon 1a revealed a putative binding site for Sp1 transcription factor and related proteins, as shown in Fig 1B To evaluate whether the Sp1-binding site in the DNASE1 distal promoter is crucial for expression, a mutated binding site was introduced into the )197H construct, resulting in the loss of 80% of luciferase activity (Fig 4B) The data show that the Sp1 site is import-ant for the DNASE1 distal promoter function involved in transcription from exon 1a To demon-strate whether the sequence between )11944 and )11931 bound Sp1 transcription factor, EMSA was carried out using nuclear extracts prepared from QGP-1 cells (Fig 5) The oligonucleotide 90,51 probe produced a major up-shifted band when the probe was incubated with the nuclear extracts (lanes 1 and 7) Formation of the up-shifted complex, indicated by the arrow, was decreased by the addition of compet-ing unlabeled self oligonucleotide or Sp1 oligonucleo-tide (lanes 2 and 4), but not by addition of oligonucleotide m90,51 containing the same mutation

of the Sp1 site in )197HmSp construct as well as oligonucleotide mSp1 with a mutated Sp1-binding site (lanes 3 and 5) Consistently, formation of the DNA– protein complex was significantly reduced when the oligonucleotide m90,51 probe was incubated with the nuclear extract (lane 6) These observations suggest that an like protein binds to the putative Sp1-binding site between )73 and )64 relative to the tran-scription start site of exon 1a To investigate whether Sp1 itself binds to the oligonucleotide 90,51 probe, a supershift assay was performed Although the tran-scription factor pancreatic duodenal homeobox-1 pro-tein (PDX-1) binds to the sequence C(C⁄ T) and can heterodimerize with PBX [20], an anti-PDX-1 IgG failed to supershift the DNA–protein complex (lane 8) However, an anti-Sp1 IgG supershifted the DNA– protein complex in association with reduction in the amount of the complex (lane 9), suggesting that Sp1 binds to the putative site in the region from )90 to )51 relative to the transcription start site of exon 1a Because the Sp1-binding site might be recognized by other members of the Sp1 transcription factor family, including Sp2, Sp3, and Sp4 [21], a chromatin immu-noprecipitation (ChIP) assay was carried out with antibodies against Sp1 and Sp1-related proteins The precipitated DNA was subjected to PCR with specific

primers for the endogenous DNASE1 distal promoter

region Analysis revealed that Sp1 binds strongly to

the DNASE1 promoter, whereas neither Sp2 nor Sp4

bind to the promoter (Fig 6) However, Sp3 seemed

to bind weakly to the promoter, although EMSA

failed to show a supershifted band with anti-Sp3 IgG

(data not shown) These results provide direct

evi-dence for Sp1 binding to the DNASE1 distal

promo-ter The data demonstrate that Sp1 is involved in the

DNASE1 distal promoter function in transcription

from exon 1a

Discussion

In previous studies, we were the first to determine the

genomic structure of the mammalian DNase I gene;

the human gene is located on chromosome 16p13.3, is

 3 kb long and contains nine exons interrupted by

eight introns [16] Subsequently, the mouse [22], rat

[23] and bovine [24] genes have been shown to be

sim-ilar to the human gene In this study, we investigated

the upstream region of the human gene 5¢-RACE ana-lysis of QGP-1 cells and pancreas revealed the presence

of a novel exon, named exon 1a, 12 kb upstream of the original exon 1 in human DNASE1 Furthermore, because the full-length 5¢-UTR of the transcript from exon 1 was determined, the 5¢ boundary of exon 1 could be confirmed and was not in agreement with ear-lier studies [15] RT-PCR analysis and promoter assay

of the 5¢-upstream regions of both exons 1 and 1a clar-ified that human DNASE1 utilizes two transcription-starting exons simultaneously for expression of the gene Accordingly, the gene organization of human DNASE1 should be corrected to 10 exons interrupted

by nine introns spanning  15 kb of genomic DNA Although the DNA region corresponding to exon 1a

in human DNASE1 could not be found on inspection

of the rat and mouse databases, further investigation might identify an alternative transcription-starting exon in DNASE1 of mammals other than humans Multiple promoters and transcription initiation sites are frequently used to create diversity and flexibility in

−78 +1 +268

−1053 −231

luc

−1.0

−2.0

kb

Relative luciferase activities

in QGP-1 cells

luc

luc luc luc

luc luc luc luc luc

1

pGL3-Basic Vector

−1053M

−231M

−137M

−78M

−54M

−33M

−28M

−17M

+1M

luc

−116M

luc

−1386M

−138

A

−13202 −12395

luc

−1.0

−2.0

kb

Relative luciferase activities

in QGPI cells

luc

luc luc luc luc

luc luc luc luc luc

2

pGL3-Basic Vector

−1995H

−2081H

−1331H

−524H

−197HmS

−94H

−73H

−60H

−45H

−14H

luc

−197H

−11871

−11781

Fig 4 Summary of relative luciferase activities of the reporter constructs containing the different length of 5¢ upstream sequence of the human DNASE1 exon 1 (A) or exon 1a (B) The different length of 5¢ upstream sequence of the human DNASE1 exon 1 or exon 1a (horizon-tal bars) were inserted into the upstream of the firefly luciferase coding sequence of pGL3–basic vector The numerals over the diagrams are the nucleotide positions relative to the transcription starting site of exon 1 Constructs are shown in the left-hand panel; construct names are given at the left of the bar and the locations of the inserted fragment are shown The circle represents the CCCC fi AGAG substitution

at )11944 and )11941 introduced in reporter construct )197HmSp Each construct as depicted on the left was transiently transfected into QGP-1 cells One microgram of firefly luciferase reporter construct and 0.01 lg of SV40 ⁄ Renilla luciferase were used for each analysis The cells were harvested for firefly and Renilla luciferase after culture for 38 h The obtained firefly luciferase activity was normalized, which is shown in the right-hand panel Mean values and standard deviations were calculated from more than three independent experiments The activity of pGL3–promoter vector containing the SV40 promoter was arbitrary, given the value of 1.0.

the regulation of gene expression [25] The level of

transcription initiation can vary between alternative

promoters, the turnover or translation efficiency of

mRNA isoforms with different leader exons can differ, and alternative promoter usage can lead to the genera-tion of protein isoforms differing in amino acid sequence Human DNASE1 pre-mRNA is transcribed from different transcription start sites, exons 1 and 1a, resulting in generation of two kinds of gene transcript However, only the sequences of the 5¢-UTR are differ-ent between them, because these transcripts share all the other exons, exons 2–9, which contain the entire coding region Thus, use of alternative promoters in DNASE1 results in no generation of protein isoforms The 5¢-UTR of eukaryotic mRNA influences the initi-ation step of protein synthesis and thereby in part determines the translational efficiency of the transcript [26] In fact, as shown in Fig 3, the translational effi-ciency of the transcript from exon 1 was about half

of that from exon 1a We used genetyx software (GENETYX Corp., Tokyo, Japan) to search for poss-ible secondary structure in the nucleotide sequence of the 5¢-UTR in the transcripts starting from exons 1 and 1a, and found that the 5¢-UTR of the transcript from exon 1 has a higher content of stem-loop struc-ture than does that from exon 1a Because stable stem–loop structures are known to cause significant suppression of translation, the distinctive secondary structures of the 5¢-UTRs in these transcripts could lead to differences in translation efficiency between them

Recently, rat DNase I pre-mRNA was reported to be alternatively spliced in the kidney, leading to the gen-eration of two types of transcript of 1.3 and 1.5 kb [27]; the former showed an internal deletion of a 132-bp segment present in exon 1, and these transcripts were produced from the same transcription starting exon Similarly, the human DNASE1 pre-mRNA undergoes extensive alternative splicing in pancreas and QGP-1 cells, resulting in the generation of normal transcripts and many alternative splicing variants initiated from both exons 1a and 1, as shown in Fig 2B Alternative splicing transcripts without exon 3 shift the reading frame and generate a stop codon in exon 4, resulting in

no production of active DNase I enzyme Furthermore, differences in splicing patterns between QGP-1 cells and pancreas (Fig 2A) suggest that splicing patterns of DNASE1 pre-mRNA could change in a cell- and⁄ or tissue-specific manner Therefore, alternative splicing in DNASE1 may potentially serve as a regulatory device

to modulate levels of the gene products

No information on the regulation of gene expression

in vertebrate DNase I gene, such as characterization of the promoter region and transcription factors respon-sible for gene expression, has been obtained so far In this study, characterization of the promoter region in

Fig 5 Sp1 specifically binds to the DNASE1 distal promoter at a

site between )91 and )51 relative to the transcription start site of

exon 1a EMSA was performed with nuclear extract from QGP-1

cells DNA–protein interaction was investigated using radiolabeled

probe 90, 51 (lanes 1–5, 7–8) or m 90, 51 (lane 6) in the presence

or absence of a 200-fold molar excess of competing unlabeled

oligonucleotides or antibodies as indicated The major shifted

com-plex is indicated by the arrow Oligonucleotides Sp1 and mSp1

con-tained the wild and mutant types of Sp1 site, respectively (lanes 4

and 5) The nuclear extract was preincubated with anti-PDX-1 (lane

8) or anti-Sp1 IgG (lane 9) A supershifted band with antibody to

Sp1 is indicated by the arrowhead.

Fig 6 ChIP assays of the Sp1-binding status at the endogenous

DNASE1 distal promoter in QGP-1 cells with anti-Sp1, -Sp2, -Sp3,

and -Sp4 IgG The amplified DNASE1 distal promoter sequences in

the input and bound fractions are shown The PCR products of

97 bp were electrophoresed through a 2% agarose gel and stained

with ethidium bromide.

the human DNASE1 gene was performed with QGP-1

cells; using promoter deletion analysis we could survey

the upstream regions responsible for transcription from

exons 1 and 1a (Fig 4), in the latter of which a

poten-tial Sp1 site in the promoter contributes to its basal

activity EMSA experiments (Fig 5), ChIP assay

(Fig 6) and the introduction of mutations in the

reporter gene construct revealed that binding of Sp1 to

the upstream region of DNASE1 exon 1a controlled

the basal distal promoter activity This is the first to

be identified as a transcription factor responsible for

gene expression of vertebrate DNase I genes Sp1 is

the founding member of a growing family of

transcrip-tion factors that bind and act through GC boxes,

being important in the transcription of many cellular

genes Sp1 is abundantly distributed in most cell types,

irrespective of differences in levels of its expression

among different cell types [28] Shiokawa and Tanuma

used polyA+RNA dot blot hybridization to show that

DNASE1 is expressed in a wide variety of human

tis-sues [6], being compatible with direct involvement of

Sp1 in expression of the gene Furthermore, we

dem-onstrated in preliminary studies that upregulation of

distal DNASE1 promoter activity in QGP-1 cells by

hypoxia might depend on the Sp1-binding site

(Kominato et al., unpublished data) Because levels of

DNase I activity are known to differ greatly among

human tissues [3,7], it is likely that additional control

mechanisms such as alternative splicing and⁄ or other

post-transcriptional regulation are responsible for

tis-sue-specific distribution of DNase I enzyme activity

Experimental procedures

Cells

Human pancreatic cancer cell line QGP-1 (JCRB0183) was

obtained from Health Science Research Resources Bank

(Osaka, Japan) The cells were grown in RPMI-1640

con-taining 10% fetal bovine serum (Invitrogen Corp.,

Carls-bad, CA), 50 UÆmL)1 penicillin and 50 lgÆmL)1

streptomycin

5¢- and 3¢-RACE analysis

5¢-RACE was performed using the 5¢-GeneRacer kit

(Invi-trogen Corp.) according to the manufacturer’s instruction

Total RNAs isolated from QGP-1 cells using the acid

guanidine thiocyanate⁄ acid phenol method [29] and human

pancreas (BD Biosciences Clontech, Palo Alto, CA) were

employed for the RACEs Five micrograms of each total

RNA was treated with bovine intestinal phosphatase,

fol-lowed by incubation with tobacco acid pyrophosphatase

and ligation with the GeneRacer RNA oligo cDNA was synthesized by using Superscript III and the DNASE1-specific primer DN+254, the sequence of which was 5¢-TAGGTGTCTGGTGCATCCTG-3¢ 5¢-Ends were PCR amplified from these cDNA templates with a primer to the GeneRacer RNA oligo and the DNASE1-specific primer DN+232, the sequence of which was 5¢-TGAGGTTGTC CAGCAGCTTC-3¢ 3¢-RACE was performed using the 3¢-RACE system (Invitrogen Corp.) according to the manu-facturer’s instruction Five micrograms of each total RNA was subjected to 3¢-RACE under the same conditions described previously [30] The sequences of the DNASE1-specific forward primers used were 5¢-GACACCTTCAA CCGAGAGCC-3¢ (DN+385) and 5¢-ATGCTGCTCCG AGGCGCCGT-3¢ (DN+721) Conditions for these amplif-ications were 95
C for 9 min, 40 cycles of 94
C for 1 min,

55
C for 1 min, 72
C for 2 min, followed by incubation

at 72
C for 10 min PCR amplifications were performed in

a 50 lL reaction mixture containing 10 pmol of each pri-mer, 1.25 units of AmpliTaq Gold (Applied Biosystems, Foster City, CA), 1.5 mm MgCl2, 150 lm dNTP, and 1· buffer After cloning of the PCR-amplified products into a pCR2.1 plasmid vector (Invitrogen Corp.), the nucleotide sequences of the amplified fragments were determined, using the BigDye Terminator v3.1 Cycle Sequencing Kit (Applied Biosystems) with both M13 forward and reverse primers The sequencing run was performed on a Genetic Analyzer (model 310, Applied Biosystems) and all the DNA sequences were confirmed by reading both strands The nucleotide sequences reported here have been submit-ted to the GenBank⁄ EMBL ⁄ DDBJ Data Bank with acces-sion numbers AB188151 and AB188152

RT-PCR cDNA was synthesized from total RNA (2 lg) of QGP-1 cells using random hexamers and Superscript III (Invitro-gen Corp.) One and two of 20 lL of the resulting single strand cDNA reaction were used as templates for RT-PCR and quantitative real-time RT-PCR, respectively Similarly,

2 lg of total RNA prepared from human pancreas was also reverse-transcribed with random primer, and the synthes-ized cDNA was utilsynthes-ized as template in RT-PCR Starting exon-specific amplification of the DNASE1 mRNA was performed using each distinct starting exon-specific primer and the reverse primer DN+785, of which the sequence was 5¢-GCCATAGGCAGCCTGGAAGT-3¢, complement-ary to exon 8 of DNASE1 The sequences of the starting exon-specific primers were 5¢-GCCTTGAAGTGCTTCTTC AGAGAC-3¢ (DN)144) and 5¢-GCACAACACAGGGAA GCTTGG-3¢ (DN)110), corresponding to the sequence in exons 1 and 1a of the DNASE1, respectively Another start-ing exon-specific amplification of the DNASE1 message was performed using each distinct starting exon-specific primer

and the reverse primer DN+89, of which the sequence was

5¢-ATGTTGAAGGCTGCGATCTTCAG-3¢,

complement-ary to exon 2 of the DNASE1 Conditions for these

amplifi-cations were 95
C for 9 min, 30 cycles of 94
C for 1 min,

60
C for 1 min, and 72
C for 2 min, followed by

incuba-tion at 72
C for 10 min Determination of the nucleotide

sequences of the amplified fragments were described above

Plasmids

The different length of 5¢-upstream sequence of exons 1 and

1a in the DNASE1 were PCR-amplified using

sequence-spe-cific primers corresponding to the sequence deposited in the

GenBank with Accession no AC006111 and subcloned into

a firefly luciferase reporter vector, the pGL3–basic vector

(Promega, Madison, WI) Nomenclature used for the

var-ious reporter constructs is based on the nature of the

inser-ted fragments Letter symbol M reflects the restriction

enzyme cleavage site of MlnI at +268 relative to the

tran-scription start site of exon 1 in the DNSAE1, letter symbol

H reflects the restriction enzyme cleavage site of HindIII at

+90 relative to the transcription start site of exon 1a,

whereas numerals indicate the endpoints of the primers used

for PCR For example,)197H construct contains the

frag-ment bordered with PCR primer sequence starting at)197

relative to the transcription start site at one end and HindIII

site at the other

The DNASE1 expression plasmid ex-pDN was constructed

by preparing a PCR-amplified fragment containing the entire

coding sequence of the human DNASE1 message by using an

Expanded High Fidelity PCR system (Roche) with the total

RNA from QGP-1 cells and a primer set of the forward

pri-mer DN-5R and the reverse pripri-mer DN+854X (sequences

5¢-CGGAATTCTCAGGATGAGGGGCATGAAG-3¢ and

5¢-CGCTCGAGGCTGCTCACTTCAGCATCAC-3¢,

res-pectively) and directionally ligating the fragment into the

EcoRI⁄ XhoI site of pcDNA3.1(+) vector (Invitrogen Corp.)

The nucleotides underlined in the primer sequences above

represent EcoRI or XhoI sites The 5¢-UTRs of the DNASE1

messages were separately amplified by RT-PCR of total

QGP-1 RNA with a primer set of the common reverse primer

DN+89 and either of the starting exon-specific forward

primers DN)243R or DN)183R, the sequences of which

were

5¢-CGGAATTCACATTTGCCCCAGGGAAGGTC-3¢ and 5¢-CGGAATTCTCCTGCCCAGGACCCGAGG-5¢-CGGAATTCACATTTGCCCCAGGGAAGGTC-3¢,

respectively The DNASE1 expression plasmids ex-pDN1

and ex-pDN1a, containing the sequences corresponding to

the 5¢-UTRs of the transcripts from exons 1 and 1a,

respect-ively, were constructed by use of the overlap extension

method [31] with those PCR-amplified fragments and the

plasmid ex-pDN The plasmids pD(1) and pD(1a) were

con-structed by cloning the PCR-amplified fragments obtained

with either of the starting exon-specific forward primers

DN)144 or DN)110 and the common reverse primer

DN+89, respectively, into a pCR2.1 plasmid vector

For all the constructs, sequencing was performed over the entire region of the inserted sequences Plasmid DNA was purified by using HiSpeed Plasmid Kit (QIAGEN GmbH, Hilden, Germany)

Transfection and luciferase assay Transient transfection experiments into QGP-1 cells were performed with Lipofectamine Plus reagent (Invitrogen Corp.); 1 lg of firefly luciferase reporter and 0.01 lg of pRL-SV40 Renilla luciferase reporter (Promega) were used for each analysis QGP-1 cells were split, 18–24 h prior to transfection, into a six-well tissue culture plate (Becton Dickinson Labware, Franklin Lakes, NJ) at 1· 105ÆmL)1

At the time of transfection, cells were washed once with Opti-MEM I-reduced serum medium (Invitrogen Corp.) containing neither fetal bovine serum nor l-glutamine Plasmid DNA was suspended in 100 lL of Opti-MEM I-reduced serum medium, followed by mixture with

6 lL of Lipofectamine Plus reagent at room temperature for 10 min Four microliters of Lipofectamine Plus reagent were diluted in 100 lL of Opti-MEM I-reduced serum med-ium The two solutions were combined at room tempera-ture for 15 min, followed by the addition of 0.8 mL of Opti-MEM I-reduced serum medium The mixture was then overlaid onto the cells The cells were incubated for 3 h prior to addition of 1 mL of Opti-MEM I-reduced serum medium containing 20% serum to the DNA-containing medium, followed by incubation for 14 h Subsequently, those cells were cultured at 37
C under either normoxic or hypoxic conditions prior to the measurement of the activit-ies of firefly and Renilla luciferase Cell lysis and luciferase assays were performed using the Dual Luciferase Reporter Assay System (Promega) Light emission was measured by Wallac 1420 ARVOMX (Perkin–Elmer) The values were obtained in relative light units Variations in transfection efficiency were normalized to the activities of Renilla lucif-erase expressed from cotransfected pRL-SV40 Renilla luciferase reporter vector COS-7 cells were transiently cotransfected by using Lipofectamine Plus reagent (Invitro-gen Corp.) with 2 lg of the DNASE1 expression vectors and 0.6 lg of the pSV–b-galactosidase vector (Promega), followed by assay of DNase I and b-galactosidase activities, according to a previously described method [32]

Preparation of nuclear extracts and EMSA The nuclear extract and probe were prepared as reported previously [33], and the sequence of oligonucleotide 90,51 probe corresponds to nucleotides )90 to )51 relative to the transcription start site of exon 1a in the DNASE1 EMSA was carried out according to the method des-cribed previously [33] The different double-stranded oligo-nucleotides were obtained by annealing two chemically synthesized strands The double-stranded Sp1 and mSp1